The Effectiveness of Health Care Teams

in the National Health Service

Report

Carol S. Borrill, Jean Carletta,Angela J. Carter, Jeremy F. Dawson, Simon Garrod,

Anne Rees, Ann Richards,David Shapiro and Michael A. West

Aston Centre for Health Service Organization Research,

Aston Business School, University of Aston

Human Communications Research Centre,

Universities of Glasgow and Edinburgh

Psychological Therapies Research Centre, University of Leeds

Contents______________________________________________

__

Key Findings

Acknowledgements

Health Care Team Effectiveness Project: Summary

Chapter 1 Teamwork, Communication and Effectiveness in Health Care: A Review Page 1

Chapter 2 Primary Health Care TeamResearch Methods and Sample Details Page 25

Chapter 3 Primary Health Care Team Results from Surveyand External Ratings Page 44

Chapter 4 Qualitative Research: Developing Objectives andEffectiveness Measures for Primary Health Care Teams Page 57

Chapter 5 Community Mental Health TeamsResearch Methods and Sample Details Page 78

Chapter 6 Community Mental Health TeamsResults from Survey and External Ratings Page 103

Chapter 7 Community Mental Health TeamsResults from Qualitative Research Page 121

Chapter 8 Secondary Health Care TeamsResearch Methods and Sample Details Page 141

Chapter 9 Secondary Care Teams Ratings Page 157

Chapter 10 Meetings and CommunicationResearch Methods Page 172

Chapter 11 Analysis of Communication in PHCT Teams Page 182

Chapter 12 Analysis of Communication in CHMT's Page 197

Chapter 13 Conclusions and Recommendations Page 215

Appendix I Survey Instruments/Rating Measures/Interview Schedules

Appendix II Knowing the way: Effectiveness in Primary Health Care

Appendix III Developing Effectiveness Measures for Primary Health Care Teams

Appendix IV Training Programme – Tools and Techniques for AssessingPerformance

Bibliography

Acknowledgements________________________________________________________

Liaison Officers: Liz MeerabeauSue LongsdateJohn Wilkinson

Advisory Group Members: Debbie MellorsNHS Executive

Sarah ConnorsNHS Executive

Jim FordNHS Executive

Bonnie SibbaldNHS Executive

Eileen RobertsonNHS Executive

Sheila RobertsDepartment of Health

Terry BreughaUniversity of Leicester

Anne NettonUniversity of Kent

Thelma SackmanNHS Executive

Research Team:

Dr Carol Borrill January 1997 - December 1999Aston Business SchoolAston UniversityBirmingham

Sam Bedlingham June 1997 - December 1999City UniversityLondon

Jean Carletta January 1997 - December 1999Human CommunicationResearch CentreEdinburgh

Christine Carmichael June 1997 - February 1998Institute of Work PsychologySheffield UniversitySheffield

Angela Carter January 1998 - December 1999Institute of Work PsychologySheffield UniversitySheffield

Jeremy Dawson July 1999 - December 1999Aston Business SchoolAston UniversityBirmingham

Simon Garrod January 1997 - December 1999Human Communications Research CentreGlasgow UniversityGlasgow

Heidi Frazer-Krauss January 1997 - June 1997Medical SchoolGlasgow UniversityGlasgow

Anne Rees January 1997 - June 1997Psychological Therapies Research CentreLeeds UniversityLeeds

Anne Richards January 1997 - December 1999Psychological Therapies Research CentreLeeds UniversityLeeds

Carein Todd April 1997 - May 1998Institute of Work PsychologySheffield UniversitySheffield

David Shapiro April 1997 - May 1998Psychological Therapies Research CentreLeeds UniversityLeeds

Michael West January 1997 - December 1999Aston Business SchoolAston UniversityBirmingham

David Woods January 1998 - June 1999Institute of Work PsychologySheffield University

______________________________________________

Summary______________________________________________

________

A primary prescription that policy makers and practitioners have offered for meeting

the challenges facing the National Health Service is the development of

multidisciplinary team working. The importance of team working in health care has

been emphasised in numerous reports and policy documents on the National Health

Service. One particularly emphasised the importance of team working if health and

social care for people are to be of the highest quality and efficiency:

"The best and most cost-effective outcomes for patients and clients are

achieved when professionals work together, learn together, engage in clinical

audit of outcomes together, and generate innovation to ensure progress in

practice and service."

Over the last thirty years this has proved very difficult to achieve in practice because

of the barriers between professional groupings such as doctors and nurses. Other

factors such as gender issues also influence team working. For example, G.P.s are

predominantly men while the rest of the primary care service population is

predominantly women; community mental health psychiatrists are predominantly

men, whereas the rest of the population of community mental health teams is

predominantly women, and in hospital settings the ranks of consultants continue to

be largely made up of men. Other factors which impede the creation of effective

multidisciplinary teams include multiple lines of management, perceived status

differentials between different professional groups, and lack of organisational

systems and structures for supporting and managing teams.

The Health Care Team Effectiveness Project was commissioned by the Department

of Health. The overall aim of the research described here was to determine whether

and how multidisciplinary team working contributes to quality, efficiency and

innovation in health care in the NHS.

The objectives of the research were to establish:

• which team member characteristics such as age, gender, occupational group,

experience, qualifications, and team size, influence how well the teams work

together;

• how team working processes, such as participation, reflexivity, communication,

decision-making and leadership contribute to the effectiveness of teams,

particularly the quality of health care and the development of innovative practice;

The research programme was carried out over a three year period by a team of

researchers based at the universities of Aston, Edinburgh, Glasgow, Leeds and

Sheffield. During the course of the study information on team working was gathered

from some 400 health care teams. This involved consulting over 7,000 NHS

personnel and a large number of NHS clients. Five national workshops were held

with key representatives from primary health care and community health care. A wide

range of research methods was used, including questionnaire surveys, telephone

interviews, in-depth interviews, observation, focus groups and video and audio tape

recordings of meetings

The research was carried out in two stages: quantitative data collection from 100

primary health care teams (PHCTs), 113 community health care teams (CMHTs) and

193 secondary health care teams (SHCTs), and in-depth work with a sub-sample of

teams.

Key findings

Effectiveness

Quality of teamworking is powerfully related to effectiveness of health care teams:

Ø The clearer the team's objectives

Ø The higher the level of participation in the team

Ø The higher the level of commitment to quality

Ø The higher the level of support of innovation

…. the more effective are health care teams across virtually all domains of

functioning

Innovation

Quality of teamworking is powerfully related to innovation of health care teams:

• The clearer the team's objectives

• The higher the level of participation in the team

• The higher the level of commitment to quality

• The higher the level of support of innovation

….. the more innovative are health care teams across virtually all domains of

functioning

Mental Health

Those working in teams have much better mental health than those working in looser

groups or working individually. The benefits appear to be due to:

• Greater role clarity

• Better peer support

Those working in teams are also buffered from the negative effects of organizational

climate and conflict.

The better the functioning of team with respect to…

• Clarity of objectives

• Levels of participation

• Commitment to quality

• Support for innovation

… the better the mental health of team members across all domains of health care.

Organisational performance

There is a significant and negative relationship between the percentage of staff

working in teams and the mortality in these hospitals, taking account of both local

health needs and hospital size. Where more employees work in teams the death

rate is significantly lower (calculated on the basis of the Sunday Times Mortality

Index, Dr Foster; deaths within 30 days of emergency surgery and deaths after

admission for hip fracture)1.

Retention and turnover

Within health care, those working in well functioning teams are more likely to stay

working in their settings than those working in poorly functioning teams.

Leadership

In Community Mental Health and Primary Health Care, where there is no clear

leader/co-ordinator or where there is conflict over leadership team objectives are

unclear, and there are….

Ø Low levels of participation

Ø Low commitment to quality

Ø Low support for innovation

Ø Poor team member mental health

Ø Low levels of effectiveness and innovation

Communication

Communication, integration and regular meetings in PHC and CMC health care

teams are associated with higher levels of effectiveness and innovation, yet the

quality of meetings (particularly in Primary Health Care) is often poor.

Professional diversity

Diversity of professional groups in Primary Health Care is clearly linked to levels of

team innovation. In newly formed Community Mental Health Teams, this relationship

does not appear. The same findings emerged from research carried out with 85

breast cancer care teams2.

1 This finding is based on research recently completed by the research team at the Aston Centre forHealth Services Organisation Research (further details available from West or Borrill).

2 This finding is based on research recently completed by the research team at the Aston Centre forHealth Services Organisational Research (further details available from West or Borrill).

Conclusions

• Systematic and revolutionary organizational change is necessary if the positive

results of this research are to be implemented in practice.

• NHS organizations have to developed as team-based, rather than hierarchical.

• Structure, culture, work design, HRM and management have to accommodate

and enable rather than impede team-based working.

• NHS employees should be trained in the KSAs for working in teams.

• NHS managers should be trained to manage team-based organizations.

Chapter 1

Teamwork, Communication and Effectiveness in HealthCare:

A Review

The challenges of organising health care in the modern United Kingdom context are

considerable. There are continual improvements in medical technologies, greater

levels of knowledge and awareness amongst patient populations and increasing

demands for the variety of sources of health care available within the National Health

Service. The provision of free health care at the point of delivery to the population

has become one of the most important issues in the national political agenda in the

early part of the twenty-first century. At the same time the National Health Service

has become a massively complex institution characterised by large organisations,

repeated restructurings, and subject to a wide range of political and economic

pressures. The response of the government has been to promise a huge increase in

spending on the NHS; a key question to be answered in relation to this political

agenda is how can we organise health care and achieve good, fair and cost effective

services for the whole population. This report focuses on determining whether, and if

so, how teamworking can help.

In this first chapter we review the research evidence about the potential benefits of

teamworking and the factors that influence the effectiveness of teams, focusing

particularly upon their use in health care settings. We draw on empirical evidence

from research conducted in the United Kingdom, mainland Europe, North America

and Australia. The literature on team composition and the processes which influence

team performance is briefly reviewed with particular emphasis on communication,

decision-making and problem-solving. We then explore the influences of

organisational context and leadership, before presenting the theoretical model which

guided the research programme described in this report.

First we consider what a ‘team’ means. The activity of a group of people working

co-operatively to achieve shared goals is basic to our species (Baumeister & Leary,

1995). The current enthusiasm for teamworking in health care reflects a deeper,

perhaps unconscious, recognition that this way of working offers the promise of

greater progress than can be achieved through individual endeavour. Mohrman,

Cohen, and Mohrman (1995) define a team as:

“a group of individuals who work together to produce products or deliver

services for which they are mutually accountable. Team members share

goals and are mutually held accountable for meeting them, they are

interdependent in their accomplishment, and they affect the results

through their interactions with one another. Because the team is held

collectively accountable, the work of integrating with one another is

included among the responsibilities of each member".

Benefits of teamwork

The belief that teamwork is the most effective way of delivering products and

services has gained increasing ascendancy within diverse organisational settings

(Guzzo & Shea, 1992; West, 1996). As organisations have grown in size and

become structurally more complex, the need for teams of people to work together in

co-ordinated ways to achieve objectives that contribute to the overall aims of

organisations has become increasingly urgent. Mohrman et al. (1995) offer ten

reasons for implementing team-based working in organisations:

• Teams are the best way to enact the strategy of organisations, because of the

need for consistency between organisational environment, strategy and design

(Galbraith, Lawler, & Associates, 1993).

• Teams enable organisations to speedily develop and deliver services cost

effectively, while retaining high quality.

• Teams enable organisations to learn (and retain learning) more effectively

(Senge, 1990).

• Cross-functional teams promote improved quality of services (Deming, 1986;

Juran, 1989).

• Cross-functional teams can undertake effective process re-engineering

(Davenport, 1993).

• Time is saved if activities, formerly performed sequentially by individuals, can be

performed concurrently by people working in teams (Myer, 1993).

• Innovation is promoted within team-based organisations because of cross-

fertilisation of ideas (Senge, 1990; West & Pillinger, 1995).

• Flat organisations can be monitored, co-ordinated and directed more effectively if

the functional unit is the team rather than the individual (Galbraith, 1993, 1994).

• As organisations have grown more complex, so too have their information

processing requirements; teams can integrate and link in ways individuals cannot

(Lawrence and Lorsch, 1969, Galbraith, 1993, 1994).

This approach to the delivery of services and products is not simply a managerial fad,

since there is substantial empirical evidence that the introduction of teamwork can

lead to increased effectiveness in the delivery of both quantity and quality of goods or

services (Guzzo & Shea, 1992; Weldon & Weingart, 1993).

Macy and lzumi (1993) conducted an analysis of 131 organisational change studies

in order to determine their effectiveness. Those interventions with the greatest

effects on organisational performance and 'the bottom-line' were team-related

interventions. They also reduced turnover and absenteeism more than did other

interventions, showing that team oriented practices can have broad positive effects in

organisations. Other research by Kahleberg & Moody (1994), who studied over 700

work establishments, found that those in which teamwork was developed were more

effective in their performance than those in which teams were not used. Finally,

Applebaum and Batt (1994) offer similar evidence. They reviewed the results of a

dozen surveys of organisational practices, as well as 185 case studies of innovative

management practices. They too found compelling evidence that teams contribute

to improved organisational effectiveness, particularly increasing efficiency and

quality.

Teamwork in health care

The importance of teamworking in health care has been emphasised in numerous

reports and policy documents on the National Health Service (NHS). One (NHSME,

1993) particularly emphasised the importance of teamworking if health and social

care for people were to be of the highest quality and efficiency:

"The best and most cost-effective outcomes for patients and clients are

achieved when professionals work together, learn together, engage in

clinical audit of outcomes together, and generate innovation to ensure

progress in practice and service."

Some limited research has suggested the positive effects of multidisciplinary

teamworking in health care. However, there are many difficulties inherent in

comparing evaluation studies, which include teams having different objectives and

organisation patterns, studies variously controlling for other concurrent changes in

local services and the pre-existing variations in services and cultures (Jackson,

Gater, Goldberg, Tantam, Loftus & Taylor, 1993).

In terms of the delivery of care, teams have been reported to reduce hospitalisation

time and costs, improve service provision, enhance patient satisfaction, staff

motivation and team innovation. We review the literature relevant to each of these

outcomes below.

Reduced hospitalisation and costs

Sommers and colleagues (2000) compared primary health care teams with physician

care across 18 private practices, and concluded that primary health care teams

lowered hospitalisation rates and reduced physician visits while maintaining function

for elderly patients with chronic illness and functional deficits. Significant cost savings

were born from reduced hospitalisation, which more than accounted for the costs of

setting up the team and making regular home visits. Jones (1992) also reported that

families who received primary health team care had fewer hospitalisations, fewer

operations, less physician visits for illness and more physician visits for health

supervision than control families. A similar pattern emerged for terminally ill patients,

where their increased utilisation of home care services more than offset savings in

hospital costs, such that there were average savings of 18% in hospital costs

(Hughes, Cummings, Weaver, Manheim, Brown & Conrad, 1992).

In another study in the U.S., Eggert and colleagues (1991) concluded that a team

focussed case management system generated similar benefits for elderly, chronically

ill patients. The team approach reduced total health care expenditures by 13.6%,

when compared to an individualised case management system. The team combined

earlier discharge, more timely nursing home placement and better-organised home

support and care, to reduce patient hospitalisation by 26%. Similarly, the cost

increases in ambulatory and nursing home care were offset by fewer and shorter stay

hospital admissions and reduced home care utilisation. For patients with dementia in

this study, the team model of case management reduced overall costs even further,

by 41% (Zimmer, Eggert & Chiverton, 1990). At the end of the 27-month study, there

were more team than control patients living at home and fewer in nursing homes. An

audit of the case managers' records highlighted more intense management activity in

the team group, where patients were referred more frequently for medical evaluation,

respite and day care. Team case managers had smaller caseloads, made more

home visits and had more case conferences. Teams were more familiar with local

community resources and were reported as being more responsive to patient crises.

The team approach was reported to offer greater intensity of case management,

which resulted in more efficient care provision in hospitals and home health services.

Improved service provision

Primary care teams appear to produce better detection, treatment, follow-up and

outcome in hypertension (Adorian, Silverberg, Tomer & Wamosher, 1990).

Specifically, nurses in England reported that working together in primary health care

teams reduced duplication, streamlined patient care and enabled specialist skills to

be used more cost-effectively (Ross, Rink & Furne, 2000).

Jansson, Isacsson and Lindholm (1992) analysed the records of general practitioners

and district carers over 6 years in Sweden. Care teams (GP, district nurse, assistant

nurse) were introduced into one region but were absent in another comparative

region. The care teams reported a large rise in the overall number of patient contacts

and in the proportion of the population who accessed the district nurse. Concurrently,

there was a reduction in emergency visits, which they attributed to better accessibility

and continuity of care in the teams.

Jackson and colleagues (1993) reported a similar pattern twelve months after the

introduction of a community mental health team in England. They reported a

threefold increase in the rate of inception to care, a doubling in the prevalence of

treated psychiatric disorder and a reduction in demand on the hospital’s outpatient

services. It was suggested that the team was making specialist care more available

to patients with severe mental illness who would not have previously received care

from mental health services. The team also provided care in a timelier manner that

was accessible and continuous.

Enhanced patient satisfaction

Hughes and colleagues (1992) compared the provision of hospital-based team home

care and customary care for 171 terminally ill patients in a large U.S. Department of

Veterans Affairs hospital. They noted increased access to home care services and

improved patient and carer satisfaction with hospital-based team home care. Both

patients and caregivers of the team expressed significantly higher levels of

satisfaction with continuous and comprehensive care at one month, and they

continued to express higher levels of satisfaction at six months. The team program

maintained patients at home for significantly more days than the control group, who

were kept in hospital in general wards for longer. Patients of the team received

almost twice as many home visits as the control group and visited the clinic

significantly fewer times.

Increased satisfaction by patients who had access to a primary health care team was

reported to include a higher mean number of social activities, fewer symptoms and

slightly improved overall health. These differences were noted in comparison to

patients who only had access to a physician (Sommers et. al., 2000).

Staff motivation

Primary care teamworking has been reported to improve staff motivation (Wood,

Farrow, & Elliott, 1994). In a study in Spain, Peiro, Gouzalez-Roma & Romos (1992)

showed relationships between work team processes, role clarity, job satisfaction and

leader behaviours. Effectiveness of teamwork was also related to job satisfaction

and mental health of team members. Sommers and colleagues (2000) suggested

that lower rates of hospitalisation for patients of primary health care teams were more

likely to be found in teams where individual members were most satisfied with their

working relationships.

Innovation

Teamwork is reputed to promote innovation in organisations including those in the

health care sector. In order to promote organisational innovation, policy makers and

practitioners are increasingly asking for clarification of the factors that determine

innovation in teams. Many input and process variables have been demonstrated to

predict innovation in teams.

In relation to inputs, there is some evidence that heterogeneity of team composition

is related to team innovation (Hoffman & Maier, 1961; McGrath, 1984; Jackson,

1996). West and Anderson (1996) carried out a longitudinal study of the functioning

of top management teams in 27 hospitals and examined relationships between team

and organisational factors and team innovation. Their results suggested that team

processes best predicted the overall level of team innovation, while the proportion of

innovative team members predicted the rated radicalness of innovations introduced.

Specifically, West and Wallace (1991) found that team collaboration, commitment to

the team and tolerance of diversity were positively related to team innovativeness.

By what means are these various benefits of teamworking in health care realised?

Partly at least through their composition and through effective team processes such

as communication, decision-making and problem-solving. We therefore briefly

review research in these areas before turning to consider the influence of the

organisations within which teams function.

Team composition and Processes

There is considerable agreement that heterogeneity of skills in teams performing

complex tasks is good for effectiveness (e.g., Campion et. al., 1994; Guzzo &

Dickson, 1996; Jackson, 1996; Millikan & Martins, 1996; Maznevski 1994).

Heterogeneity of skills and knowledge automatically implies that each team member

will bring a different knowledge perspective to the problem, a necessary ingredient

for creative solutions (Sternberg & Lubart, 1990; West, 1997).

However, teams that are diverse in task-related attributes are often diverse in

individual attributes. Variation in individual characteristics can trigger stereotypes

and prejudice (Jackson, 1996) which, via interteam conflict (Tajfel, 1978; Tajfel &

Turner, 1979; Hogg & Abrams, 1988), can affect team processes and outcomes. As

an example, Alexander, Lichtenstein and D’Aunno (1996) found that individuals in

multidisciplinary treatment teams in U.S. Department of Veterans Affairs hospitals,

who were members of larger and more heterogeneous teams, reported poor team

functioning. Physicians and social workers assessed team functioning more

positively than did nurses. The greater the diversity of individual characteristics of

team tenure, age and occupation within teams, the more negatively did team

members assessed team functioning.

Gender

Gender is an important influence on communication within teams. Not only are men

consistently more assertive in public situations and confrontations (Kimble, Marsh &

Kiska, 1984; Mathison & Tucker, 1982), but also communication expectations differ

for men and women. Sex-role stereotypes prescribe passive, submissive and

expressive communication for women while men are expected to be active,

controlling and less expressive communicators (LaFrance & Mayo, 1978).

Punishment for violation of expectations (Jussim, 1986; Jussim, Coleman & Lerch,

1987; Jackson, Sullivan & Lodge, 1993) may influence both the perceptions of

women in teams and their willingness to participate in team communication. Such

considerations are vitally important in health care teams where women dominate in

number, but men predominate in the highest status positions (in the present

research, GPs and psychiatrists, for example).

In support, Alexander, Lichtenstein and D’Aunno (1996) reported that the greater the

gender diversity, the more positive were team members’ assessment of how

cohesively and harmoniously teams operated. Their research suggested that mixed

gender teams included different orientations to work, namely a female focus on

workplace processes and relationships and a male focus on tasks and outcomes.

Team roles

It is important that teams have the appropriate mix of clearly defined team roles.

Jansson, Isacsson and Lindholm (1992) analysed the records of general practitioners

and district carers over a six-year period across 2 districts in Sweden following the

introduction of care teams into one region. They found that through the independent

roles of nurses and doctors were retained in the primary health care teams, all team

members interacted with the population in various situations, including home visits

and complemented each other across different competencies.

Team affective tone

Another important, but more controversial approach to understanding work team

processes and effectiveness, is offered by research on team affective tone. George

(1990) suggests that if members of a team experience similar kinds of affective

states at work (either negative or positive), then affect is meaningful not only in terms

of their individual experiences, but also at a team level. A number of studies have

demonstrated a significant relationship between team affective tone and behaviour

such as absenteeism (George, 1989, 1990, 1995). George proposes that teams that

are interested, strong, excited, enthusiastic, proud, alert, inspired, determined,

attentive and active, enable cognitive flexibility, creativity and effectiveness (George,

1996). However, she argues that team affective tone may not exist for all teams, so

it cannot be assumed a priori that it is a relevant construct for every team. George

(1996) sees team affective tone and team mental models as having a reciprocal

influence. So in a team with a negative affective tone, members would have different

cognitive processes from those in a team with a positive affective tone, which then

may influence team effectiveness.

There is some evidence that team mental models play an important role in team

decision-making (Klimoski & Mohammed, 1994), impacting on aspects of team

decision-making such as problem definition, speed and flexibility, alternative

evaluation and implementation (Walsh & Fahey, 1986: Walsh, Henderson &

Deighton, 1988). A team that has a high negative affective tone may tend to be more

rigid when making decisions. The nature and outcomes of team decision-making are

therefore likely to be affected by the interaction between team affective tone and

team mental models.

Communication

The study of communication in teams has a long history in social psychology, but

recent reviews by Guzzo & Dickenson (1996) and Guzzo and Shea (1992) reveal the

paucity of thorough industrial and organisational research in this area. Blakar (1985)

proposes five pre-conditions for effective communication in teams. Team members

must have shared social reality within which the exchange of messages can take

place, including a shared language base and perception. Team members must be

able to “decentre”, to take the perspective of others into account in relation to both

their affective and cognitive position (Redmond 1989, 1992). Team members must

be motivated to communicate. There must be “negotiated and endorsed contracts of

behaviour” (i.e. agreement among team members about how interactions take place).

Finally, the team must attribute communication difficulties appropriately, so if one of

the other preconditions is not being met, the team is able to correctly identify the

problem and develop a solution.

Several research studies in England have highlighted interprofessional

communication problems within primary health care teams. West and Field (1995)

and Field and West (1995) interviewed 96 members of primary health care teams

and described factors that impacted upon teamworking and communication in health

care. Structured time for decision-making, team cohesiveness and team-building all

influenced communication within teams. They highlighted the failure of health care

teams to set aside time for regular meetings to define objectives, clarify roles,

apportion tasks, encourage participation and handle change. Other reasons for poor

communication included differences in status, power, educational background,

assertiveness of members of the team, and the assumption that the doctors would be

the leaders (see also West & Pillinger, 1995; West & Slater, 1996).

Communication difficulties between different professional groups have been

highlighted particularly. Bond, et. al., (1985) surveyed 161 pairs of General

Practitioners (GPs) and health visitors, and 148 pairs of GPs and district nurses who

had patients in common. They reported low levels of communication and

collaboration between GPs and community nursing staff and suggested that GPs had

a very poor understanding of the health visitor's role. Similarly, McClure (1984)

describes low levels of communication in a survey of 48 health visitors and 45 district

nurses attached to general practices. Community nurses reported that

communication with practice staff was usually only about specific immediate patient

issues rather than team objectives, strategies, processes and performance review.

Health visitors were noted to be similarly unenthusiastic about progress in teamwork.

Ross, Rink and Furne (2000) found that health visitors perceived teams as less

effective. They suggested that health visitors were comparatively more defensive

about the benefits of changing role boundaries and considered themselves less able

to contribute to the teams as currently constituted. Cant and Killoran (1993) reached

similar conclusions, based on their research study with 928 practice nurses, 682

health visitors and 679 district nurses. They argued that joint professional training

and the instigation of regular team meetings were necessary to promote good

communication.

Cott (1997) used a social network analysis of 93 health care workers across 3

multidisciplinary long-term care teams to explore communication processes within

teams. She concluded that higher status multi-professional members communicated

most openly and worked fairly autonomously across loosely structured tasks, with

low levels of authority. In contrast, hierarchical nursing sub-teams did not report high

levels of information sharing.

West and Slater (1996) reported that much of the potential benefit of teamwork was

not being realised, with less than one in four health care teams building effective

communication and teamworking practices (see also West & Poulton, 1997). In a

similar vein, the Audit Commission report in 1992 drew attention to a major gap

between the rhetoric and reality:

"Separate lines of control, different payment systems leading to suspicion

over motives, diverse objectives, professional barriers and perceived

inequalities in status, all play a part in limiting the potential of multi-

professional, multi-agency teamwork. . . for those working under such

circumstances efficient teamwork remains elusive" (Audit Commission,

1992).

A number of researchers in different countries have highlighted the impact of

communication problems on patients across different types of teams. Nievaard

(1987) interviewed 112 nurses and 298 patients across 6 medical and surgical wards

of 2 general hospitals in the Netherlands. The study demonstrated the phenomenon

of problem shifting, where communication problems within the team were transferred

onto patients. It was reported that for hospital teams with a good communication

climate, nurses perceived patients as more attractive and interesting and less

dependent. However, if nurses viewed relationships with doctors, managers and

nurses in the team as problematic, their images of patients tended to be more

negative (unattractive, non-cooperative, dependent) and they did not want to

increase their contacts with patients.

Yeatts and Seward (2000) reported similar findings when they compared 3 self-

managed work teams in a medium size U.S. rural nursing home. They concluded that

enhanced communication between team members positively affected the service to

residents. Observations of a high performing team’s meetings showed that team

members had a high level of respect for each other, listened to each other, and were

not afraid to disagree when they held different views. Team members sought and

valued approval from each other, and they assisted each other to complete tasks.

Several studies have demonstrated how individual perceptions about teamwork and

roles can influence communication in teams. Dreachslin, Hunt & Sprainer (2000)

developed a grounded theory of the role that race plays in the self-perceived

communication effectiveness of nursing care teams in the U.S. They concluded that

racially diverse team members evaluated team communication according to different

perspectives and alternative realities.

When team members develop belief systems that are consistent with their

perspective and incongruent with other vantage points, differences in

perspective can result in alternative realities. Alternative realities encourage

participants to attribute causality differently which in turn fuels team conflict and

miscommunication by diminishing the team’s ability to reach a common

understanding of both the source of the conflict and the optimal path to its

resolution through effective communication (p. 1408).

Black participants were more likely to suggest that race exacerbated team conflict

and miscommunication, whereas white participants attributed problems to role and

status in the team. Further, different emphases and responsibility for communication

were acknowledged amongst the diversity of races, ethnicities, ages and genders.

Social isolation, selective perception and stereotypes also served to reinforce these

differences and deepen communication problems. Fewer occasions for social

interaction reduced opportunities to develop shared beliefs and a common social

reality across racial groups. The researchers therefore suggested that team

members be encouraged to understand different perspectives and appreciate

alternative realities, in order to lessen social isolation and reduce selective

perceptions and stereotyping behaviours.

Freeman, Miller and Ross (2000) also developed a grounded theory about

collaborative practice at the levels of the organisation, group and individual. They

conducted case studies of 6 teams working in a variety of specialist healthcare

services (diabetes, medical ward, primary healthcare, neuro-rehabilitation unit, child

development assessment, community mental health) and concluded that the

meanings different professionals ascribed to teamwork shaped how they

communicated and what they communicated about. When there was a lack of

congruence about aspects of teamwork, communication could potentially be

compromised. Individual perceptions determined the level of role understanding

considered necessary, and the value assigned to others’ contributions. Differences in

the understanding and valuing of team roles and levels of team learning exacerbated

underlying resentments, undermined professional esteem and created conflict.

Individual perceptions also influenced communication regarding tasks and about

sharing professional knowledge and ideas.

Decision making

Effective decision-making processes are central to team performance. Several

studies have reported the positive benefits of participative decision making in health

care teams. Yeatts and Seward (2000) compared 3 self-managed work teams in a

medium size U.S. rural nursing home. Team members of highly performing teams

reported that their ability to participate in work related decisions greatly increased

their job satisfaction and desire to come to work. These team members adopted a

consensus model of decision making, in which they clarified the problem, considered

alternatives, weighed the strengths and weaknesses of each alternative, and

selected the best option. Following their participation in making decisions, team

members reported an enhanced self-image and self-confidence, and they described

more positive interactions amongst themselves and with residents.

In contrast, Cott (1997) suggested that team members may not be equally

empowered to participate in decision making. Using a social network analysis of 93

health care workers across 3 multidisciplinary long-term care teams, she reported

that the highest status nurses and the core multidisciplinary professionals

participated most in decision making and problem solving activities. In comparison,

the lower status nursing sub-team primarily planned and assisted each other with

their more mechanistic tasks.

Problem solving

Team problem solving improves when members examine their definitions of a

situation to ensure they are solving the "right" problem (see for example, Bottger &

Yetton, 1987; Hirokawa, 1990; Landsberger, 1955; Maier, 1970; Schwenk, 1988). In

contrast, teams that detect problems too slowly or misdiagnose them often are

ineffective. Attributing problems to the wrong causes, or not communicating about

potential consequences, often undermine team effectiveness, especially when team

members fail to reflect on the possibility of error (Schwenk, 1984; Staw & Ross,

1989).

Teams that engage in more extensive scanning and discussion of their environments

perform better than those which do not identify problems (Ancona & Caldwell, 1988;

Main, 1989; Billings, Milburn & Schaalman, 1980). Tjosvold (1985; 1990) linked the

open exploration of opposing opinions within teams with effectiveness. Maier and

colleagues also suggested that cognitive stimulation produced novel ideas, and that

team effectiveness could be improved if teams were encouraged to be "problem

minded" rather than "solution minded" (Maier & Solem, 1962; see also Maier, 1950,

1970). Effectiveness was improved when teams questioned current approaches or

considered other aspects of problems (Maier, 1952). Similarly, Hackman & Morris

(1975) found that additional process discussions facilitated the quality of team

performance. The judged creativity of team decisions was related to the number of

comments made about performance strategy. When teams produced alternative

solutions to a problem, or separated and recombined problem solving strategies,

enhanced productivity was reported (Maier, 1970).

Teams that have to make complex decisions report that planning enhances their

performance (Hackman, Brousseau & Weiss, 1976; Smith, Locke & Barry, 1990).

However, when the environment becomes more uncertain, problem identification is

more difficult (Hedburg, Nystrom & Starbuck, 1976; Kiesler & Sproull, 1982).

Ineffective teams tend to deny, distort or hide problems (Stein, 1996). In some

teams, the identification of problems is discouraged as problems are regarded as

threats to morale, or a source of conflict (Janis, 1982; Miceli & Near, 1985; Smircich,

1983).

Thus far we have reviewed the benefits (and potential difficulties) of teamworking in

health care organisations - but the fact that teamworking takes place within

organisations is often ignored in the zeal to promote team effectiveness.

Accordingly, we now turn to address what is currently known about the influence of

their organisations upon teams.

Organisational context

Recent research suggests the broader context within which teams work has an

influence on their performance. Indeed the major change in emphasis in research on

teams in the last 15 years has been the shift from discussion of intrateam processes

to the impact of organisational context on teams. The organisation within which a

health care team functions can influence team effectiveness in a variety of powerful

ways. Researchers, such as Hackman (1990) and Tannenbaum, Beard and Salas

(1992) have suggested that the following are among the contextual factors that

influence team effectiveness:

• Team and organisational rewards

• Team objectives and performance feedback

• Training and technical assistance

• Physical work conditions

• Organisational climate

• Inter-team relationships

• Contracts and management structures

• Team size

These factors will be discussed further, in turn.

Team and organisational rewards

It has long been known in the social sciences that rewards are important for

improving performance. Reward systems, such as public recognition, preferred work

assignments and money enhance motivation and performance, particularly when the

rewards are contingent upon task achievement (Hackman, 1990; Sundstrom et al.,

1990; Vroom, 1964). However, team performance is most effective when rewards

are administered to the team as a whole and not to individuals, and when they

provide incentives for collaboration and communication rather than individualised

work (Hackman, 1990). This reinforces individuals working together as a team.

Gladstein (1984) found that in sales teams, pay and recognition affected the leader’s

behaviour and the way the team structured itself. Yet, NHS management directly

undermines teamwork in primary health care when they provide bonus systems to

GPs as independent contractors, despite the whole team contributing to the final

outcome.

Clear team objectives and performance feedback

In healthcare environments, team members need information about local health

needs and services, and national policies and guidelines, in order to set objectives

and target their activities appropriately. Further, feedback on team performance is

important for setting realistic goals and fostering high team commitment (Lathom,

Erez & Locke, 1988). Job satisfaction requires accurate feedback from both the task

and other team members (Drory & Shamir, 1988). However, team feedback can be

difficult to provide to teams with either long cycles of work or one-off projects

(Sundstrom et. al., 1990).

Training and technical assistance

Hackman (1990) argued that training and technical assistance is required for teams

to function successfully. Knowledge and training about team functioning is needed to

supplement team members’ own technical and medical skills and knowledge

(Poulton & West, 1993; Poulton & West, 1994a, 1994b; Poulton & West, 1997).

Limited empirical evidence suggests training is correlated with both self-reported

effectiveness (Gladstein, 1984) and managers’ judgements of effectiveness

(Campion et. al., 1993) in teams.

Physical Work Conditions

Physical conditions are another situational constraint that affect the relationship

between performance dimensions and team effectiveness. For example, a health

care team whose members are dispersed across sites, will find decision making

more difficult and ineffective than a team whose members share the same physical

location.

Organisational Climate

The climate of the organisation - how it is perceived and experienced by those who

work within it - will also influence the effectiveness of teams (Allen, 1996). Where the

climate is one characterised by high control, low autonomy for employees, lack of

concern for employee welfare and limited commitment to training, it is unlikely

teamworking will thrive (Markiewicz & West, 1997).

The extra commitment and effort demanded in team-based organisations requires

organisational commitment to the skill development, well-being and support of

employees (Mohrman, Cohen & Mohrman, 1995). Competition and intrigue can

further undermine team based working in health care, since teamwork depends on

shared objectives, participative safety, constructive controversy and support (West,

1990; West & Anderson, 1996). Ross, Rink and Furne (2000) reported that team

members’ willingness to work in teams was limited by the lack of a common set of

values about the benefits of teamwork. They recommended the need for clear

objectives, leadership, commitment and wide organisational ownership as precursors

for working in teams.

Professional subcultures also influence team effectiveness. Kinnunen (1990) used

an anthropological approach to distinguish different subcultures between medical,

nursing and management staff in a large primary health care organisation in Finland.

These three professional groups described different relationships to formal power

structures, which influenced their group behaviour, leadership style, administrative

orientation, decision-making preferences and patient interactions. In general, doctors

and managers shared basic assumptions about work that were paternalistic,

proactive, dominant and emphasised loyalty to authorities. In contrast, nurses

stressed participation, delegation, traditions and symbiotic harmony in work relations.

Inter-team relationships

In a comprehensive study of team-based organisations involving both questionnaire

and case study methods, Mohrman et. al. (1995) demonstrated that inter-team

competition is a major threat for team-based working. Teams that compete may

develop greater commitment to the team’s success than the organisation’s success.

Thus the health care team may focus on increasing the financial benefits to their

team at the expense of the wider National Health Service. Teams competing against,

rather than supporting each other may withhold vital information or fail to offer

valuable support in the process of trying to achieve team goals, without reference to

the wider goals of the organisation. Thus, health care teams may fail to pass on

information about former patients to other teams, focusing their efforts on their own

team’s immediate demands.

Ross, Rink and Furne (2000) reported a lack of focus on patient care in their

evaluation of primary care nursing teams in England. Nurses perceived that current

organisational change promoting teamwork was concerned with structure,

professional and organisational issues rather than with patient care. Some nurses

were concerned that moves towards integrated nursing were primarily motivated to

cut costs.

Contracts and Management Structures

Other relevant aspects of the organisational environment in health include the

independent contractor status of GPs and different management structures. There

are very few organisations where one or more senior team members work as

independent contractors and the rest of the team work within a variety of

organisations. Even the most sophisticated management practices, in environments

such as the oil and gas industry, are struggling with notions of how to operate joint

venture systems - whereas health care teams must deal with these issues constantly

but without the training and support given to teams in these other sectors.

Team size

The size of the team is also important, since bigger teams experience much greater

strains on effective communication. In most other sectors, teams tend to be divided

once they reach 12 or 13 members. But primary and secondary health care teams

(for example) can be 20, 30, 40 or more members in size. These ‘teams’ would be

more correctly termed ‘organisations’. In and of itself, this would not be a problem, if

those who run such organisations are adequately trained to manage large

operations. They require knowledge of the management of culture, power, conflict,

spans of control, strategies, innovation and above all, people. Yet primary health

care team leaders are rarely given such training (West, 1994). It is to the topic of

leadership that we now turn.

Leadership

There is considerable research evidence that leaders affect team performance (e.g.

Brewer, Wilson & Beck 1994; Komaki, Desselles & Bowman, 1989) and evidence of

the relationship between leadership style and team effectiveness. Eden (1990)

examined the effects of platoon leaders’ expectations on team performance. His work

with the Israeli Defence Forces showed that those platoons which trained under

leaders with high expectations, performed better on physical and cognitive tests.

Podsakoff and Todor (1985) investigated the relationship between team members’

perceptions of leader reward and punishment behaviours and team cohesiveness,

drive and productivity. Results showed that both leader contingent reward and

punishment were positively related to team drive and productivity. Leader contingent

reward was also related to cohesiveness, while leader noncontingent punishment

behaviour was negatively related to team drive. Jacob and Singell (1993) examined

the effects of managers on the won-lost record of professional baseball teams over

two decades and found that leaders did influence team performance by exercising

tactical skills and improving the performance of team members. George and

Bettenhausen (1990) studied teams of sales associates reporting to a store manager

and found that the favourability of leader’s moods was negatively related to related to

employee turnover.

Primary health care team members in England rated their effectiveness more highly

when they had strong leadership and high involvement of all team members (Ross,

Rink & Furne, 2000). In nursing care teams, Dreachslin, Hunt and Sprainer (2000)

concluded that leadership mitigated the influence of race in self-perceived

communication effectiveness. Participants’ comments supported the theme that

team leaders who encouraged discussion about differences enhanced perceived

team effectiveness. They suggested that leaders provided a unifying force through

validating the alternative realities and appreciating the different perspectives of team

members, thus moderating the potentially negative effects of racial diversity on team

processes.

Developing Teams in Organisations

To what extent is it possible to develop team working to ensure higher levels of

effectiveness? Tannenbaum, Salas, & Cannon-Bowers (1996) have reviewed

research in this area and related results to a comprehensive model of team which

integrates interventions (Tannenbaum, Beard and Salas, 1992). They describe a

number of intervention types include team member selection and teambuilding:

Team member selection

Although organisations tend to use quite sophisticated methods for

selecting employees for individual jobs, they rarely use systematic

methods for selecting for teams. But systematic selection methods can

help identify people with greater skill levels. There is strong evidence

that a team composed of skilled and motivated people will be more

effective than other teams (Tziner, 1988). Selection interventions could

improve team effectiveness by increasing the professional or skill

diversity of health care team members, thereby increasing the range of

competencies in the team.

Teambuilding

Some teambuilding interventions focus on role clarification, some on

interpersonal relationships or conflict resolution issues, while others take

more of a general problem-solving approach (Tannenbaum, Salas &

Cannon-Bowers, 1996). Team norms, attitudes, climate and power

distribution can be affected by teambuilding approaches. Many team

processes, including communication, decision-making and mutual role

understanding, are often direct targets of team building interventions.

Weldon and Weingart (1993) describe the importance of planning in teams for

achieving team goals, and suggest that team members are characteristically slow to

respond to changes in their tasks or their environments that make their strategies

ineffective or their goals obsolete. They propose five ways of supporting team work.

Goals should be set for all dimensions of performance that contribute to the overall

effectiveness of the team; feedback should be provided on the team's progress

towards its goal; the physical environment of the team should remove barriers to

effective interaction (consider the difficulties faced by members of a dispersed health

care team); team members should be encouraged to plan carefully how their

contributions can be identified and co-ordinated to achieve the team goal; and team

members should be helped to manage failure, which can damage the subsequent

effectiveness of the team.

Pritchard, Jones, Roth, Stuebing and Ekeberg (1988) tested some of these ideas by

measuring the effects of team feedback, goal setting and incentives on productivity.

Five organisational units in the military were studied. One, a maintenance section,

repaired a variety of electronic equipment used for aircraft communications. The

other four sections together made up a material storage and distribution branch.

Productivity baselines were established before each team received new "treatments"

(i.e., performance feedback eight months after the study began, goal setting five

months later, and incentives a further five months later) to determine the incremental

effects of these "treatments". First, the level of performance of the teams was

measured over a period of eight months and then information on their performance

was given to each unit for five months. The teams next set clear targets in addition to

the performance feedback, and their performance was measured for another five

months. Feedback was in the form of computer-generated reports, given monthly to

the personnel of each unit. Finally, incentives were offered for high performance, in

the form of time off from work. Using these approaches, the average increase over

baseline productivity was 50% for feedback, 75% for goal setting and 76% for

incentives. The results showed a major increase in productivity among the teams,

though the unique contribution of each component of the intervention is difficult to

estimate accurately. Both goal setting and feedback had powerful effects on

performance.

Transition of organisations to teamworking

One of the most exciting developments in the field is the new emphasis upon the

development of team-based working in organisations (Mohrman, et. al., 1995;

Markiewicz & West, 1996, 2001). This reflects a concern amongst practitioners with

how team-based working can be effectively introduced into organisations. Mohrman

et. al., studied 25 teams in four companies using a grounded research methodology,

involving managers and internal customers. In the second phase of their research

they surveyed 178 teams across seven corporations, involving team members,

managers and customers. In this way, they developed a five stage design sequence

for the transition to a team-based organisation:

1. Identifying work teams and the nature of the task

This involves process analysis to determine essential work activities that have to be

conducted and integrated to produce products or services; deliberations analysis

which identifies dialogues about issues that have to be repeatedly resolved in order

to provide shared direction and enable people to complete their tasks; and task

interdependence analysis which determines where and to what extent individuals and

teams have to rely on each other to complete their tasks.

2. Specifying integration needs

In order to integrate across multiple teams and components of business units,

Mohrman et al recommend management teams, representative integrating teams

(where an overall co-ordinating team had representatives from each of those teams

collectively involved in producing a product or service), individual integrating roles,

and improvement teams.

3. Clarifying management structure and roles

This stage involves putting as much self-management responsibility into the teams

as possible; involving team members in determining how leadership tasks will be

performed and by whom; using lateral mechanisms for cross-team and organisation-

wide integration so that teams participate in that integration; and creating

management roles which link teams to the organisational strategy and ensure they

are responsive to the organisational and wider environmental context.

4. Designing integration processes

The research evidence suggests that team-based organisations should set clear

directions in the organisation, (for example by defining, communicating and

operationalising a strategy at all levels, aligning goals, assigning rewards in

accordance with organisational goals, and planning collectively); managing

information distribution and communication; and developing an appropriate decision

making strategy (by clarifying decision making authority, and appropriately involving

organisational contributors).

5. Developing performance management processes

Finally, the model suggests the need to manage performance - defining, rewarding

and reviewing performance and involving internal and external customers, and team

members. Mohrman et. al., report that the more people were rewarded for individual

performance, the worse team performance was. The more people were rewarded for

team performance, the better was the team and the business unit’s performance and

the more process improvements the team and the business unit instituted.

Conclusions and Research Mode

A review of the literature reveals that progress has been made in understanding the

factors that influence the ability of people to work effectively together in teams.

However, in the health care domain progress is still patchy and only a few studies are

constructed on firm theoretical bases. Progress is further inhibited by the added

difficulty of operationalising the concept of effectiveness. In the research described

in this report we attempted to build our research on a well-accepted theoretical base

and to engage a large number of health care teams in the research endeavour.

Moreover, we were charged with grasping the nettle of effectiveness in health care

and developing robust and sufficiently broad measures of this difficult concept.

Finally, the research team, drawn from a wide range of epistemological backgrounds

and theoretical orientations, determined to employ diverse, powerful and innovative

research methods to answer the question of what factors influence the effectiveness

of health care teams. The starting point for the research was a model of the factors

influencing team effectiveness and which distinguishes between at least three major

domains of effectiveness. Theoretical approaches to understanding teams at work

have been dominated by the input-process-output structure, mainly because of its

categorical simplicity and utility (see Figure 1 below) (West, Borrill, & Unsworth,

1998). This is the model used to guide the research described in this report.

Figure 1: Input, process, output model of team effectiveness

INPUTS GROUP PROCESSES OUTPUTS

Domain

Health CareEnvironment

Organisationalcontext

Team task

Team composition

Leadership

Clarity of objectives

Participation

Task orientation

Support for innovation

Reflexivity

Decision making

Communication/integration

Effectiveness - self andexternally rated

Clinicaloutcomes/quality ofhealth care

Innovation - self andexternally rated

Cost effectiveness

Team member mentalhealth

Team member turnover

Inputs

Teams work within a domain such as primary care, secondary care or community

mental health. They also work in a health care environment that may be more or less

deprived. The team works for and within an organisation; thus it will be affected by

the interaction with the surrounding organisational context. A team has a task that

potentially impacts upon team processes and effectiveness (the management of

immunisation for children under five years; intensive care nursing; or care of the

elderly with mental health problems). The team consists of a collection of individuals

- who represent the group’s composition – varying in professional background,

gender, age, personality etc. Finally, the team exists within a wider society that will

affect the teams’ fundamental beliefs and value systems, i.e., the cultural context.

Processes

Processes within teams enable them to achieve their goals. A fundamental

requirement for effectiveness is that teams have clear objectives to which their

members are committed. Other processes include participation in decision-making,

emphases on quality, and support for innovation. Another fundamental process is

the extent of coordination and integration of team members’ work (Worchel, Wood, &

Simpson, 1992). And of course, leadership and communication are likely to be

important to team effectiveness. Another potentially important process variable is

reflexivity or the extent to which team members collectively reflect on the objectives,

strategies, processes and environment of the team and make changes appropriately

and accordingly.

Outputs

Six principle outputs can be distinguished: overall effectiveness, clinical outcomes,

team

member mental health, innovation, team member turnover, and cost effectiveness.

In the research programme described in this report we explore the relationships

between inputs and processes; inputs and outputs; and processes and outputs in

390 UK NHS teams, during the course of which we consulted with over 7,000 NHS

personnel and with a large number of NHS clients.

Using this model, we determined to explore the extent to which team working was

associated with better quality health care for patients and to identify the factors

associated with effective teamwork.

Chapter 2

Primary Health Care TeamResearch Methods and Sample Details

The research with Primary Health Care Teams had two stages. The first was a

questionnaire survey, and related data collection methods, involving large numbers

of teams conducted in order to gather data on team inputs, processes and outcomes.

The second stage involved intensive examination of a sub-set of teams to explore in

more depth targeted team processes and outputs.

Quantitative Methods

An overview of the methods used is given in Figure 2.1.

Figure 2.1: Details of the three samples and research methods

Sample size Survey data Additionalquestionnaires/Telephoneinterviews

External ratings

100 teams1156 respondents

Team compositionTeam functioningTeam effectivenessTeam innovationMember stress

Team composition

Team meetingsTeam managementDecision making

Team effectivenessTeam innovation

The Sample

The research design required data to be gathered from 100 Primary Health Care

teams (PHCTs) varying across a number of dimensions, including size (number of

team members, number of GPs, list size); Jarman index; location (urban, rural, inner

city), and geographical location. Databases of GP practices were accessed from 19

Health Authorities and 300 teams were randomly selected.

Letters explaining the objectives of the research and inviting teams to participate in

the research, together with an information sheet were sent to the senior GP partner,

senior health visitor and practice manager/senior receptionist in each practice. A

reply slip was included, which also solicited additional information about the team

(fund holding status, frequency and type of meetings, Jarman index, number of GPs,

list size etc.)3

The initial letter was followed up with a telephone call to the practice manager/senior

receptionist at all 300 practices. If teams had already indicated a willingness to

participate, practical arrangements for questionnaire distribution were made.

Researchers requested the name of a contact person in the team to enable

continued effective liaison. The contact person was telephoned at a later date to

determine whether the team was willing to take part in the research. Teams that did

not return a reply slip were also telephoned and provided with additional information.

Further follow-up telephone calls were made until the team made a decision about

participation in the research (some PHCTs were contacted six or seven times before

a decision was made). When teams agreed to collaborate in the research,

questionnaires were sent to the contact person for distribution to team members.

After three months the response rate from 10 teams was below 30% and 23 had not

returned any questionnaires. These teams were dropped from the sample and

replaced with 7 teams based in an inner city area, and 7 from a rural location,

resulting in a final sample of 100 teams. The total response rate was 55.8%.

Response rates for teams ranged from 21.4% to 100%, with a mean of 57.6%.

Data Collection Methods

Data on team functioning and effectiveness were collected using three methods: self

report questionnaires completed by individual team members; self report and

telephone interview surveys with the team contacts; and external ratings from

primary health care representatives and health authority staff.

1156 respondents from 100 PHCTs completed questionnaires on their perceptions of

team functioning and team effectiveness. Of these, 85% were female; 15% were

GPs; 14.2 % practice nurses; 23% trust nurses (health visitors, district nurses,

3 Copies of interview schedules, questionnaires and all data collection instruments areavailable from the first author of this report.

midwives); 33.5% administrative staff, 7.3% managers and 4.3% professions allied to

medicine (PAMs).

Team contacts from 77 PHCTs provided information on team context, team

composition and team processes in a self-report questionnaire survey, and 100

provided information via a telephone interview. This enabled a reliability check on the

data for 77 of the teams.

Questionnaires completed by individual team members

This questionnaire was in four sections (a copy of the primary health care

questionnaire is included in Appendix I).

Section 1: Team working

This contained seven measures of team working. Four of these were drawn from the

Team Climate Inventory (Anderson & West, 1994,1998) that is based on a well-

developed theoretical model of team functioning (West, 1990). The four measures

assess levels of:

• team participation

• clarity of and commitment to team objectives

• emphasis on quality

• support for innovation.

Three other measures were included:

• reflexivity – the extent to which team members reflect upon their team

objectives, strategies and processes and make changes accordingly (West,

1996; Swift & West, 1998).

• team innovation – the extent to which the team has introduced innovations in

objectives, work strategies, processes and relationships

Respondents were also asked to write descriptions of the major changes or

innovations introduced by the team in their work in the previous 12 months.

Section 2: Effectiveness

This included 21 measures of primary health care team effectiveness adapted from

Poulton and West (1999). There are three underlying dimensions:

• team working

• patient orientation

• organisational efficiency

Section 3: Team member stress

This included a measure of psychological stress, the GHQ-12 (Goldberg, 1972;

Goldberg & Williams, 1991). The GHQ-12 is widely used as a screening tool for

detecting minor psychiatric disorder in the general population, and in occupational

mental health research. It covers feelings of strain, depression, inability to cope,

anxiety based on insomnia, lack of confidence and other psychological problems.

Within a Department of Health-funded study of the mental health of the NHS

workforce, the GHQ-12 showed good validity against a psychiatric interview (Hardy,

Shapiro, Haynes, & Rick, 1999).

Section 4: Biographical information

This section included questions on biographical and team characteristics (e.g. age,

gender, ethnic origin, job title, employer, team composition, team leader).

Additional Practice Information – Survey

This was completed by the contact person in the PHCT (usually the practice

manager). It included questions on: team context (relationships with external

agencies such as health authorities and trusts); type of primary health care practice

(fundholding, non-fundholding, dispensing) quality of premises; team composition

number in each occupational group, grade, hours worked, time working in the team);

staff development; and team processes (communication and decision making in

meetings).

Additional Practice Information - Telephone Interview Schedule

The contact person in the team (usually the practice manager) responded to the

telephone interviews. The focus of the questions was on decision-making and

communication in the team: specifically who was involved in making operational,

strategic and clinical decisions in the team, how these decisions were communicated

in the team and what mechanisms were in place within the team to promote

communication (memo systems, message books, informal meetings, email).

Information was also gathered on the services and clinics provided by the team.

External ratings – team effectiveness and innovation

Health Authorities employ staff to provide support to primary health care teams, a

role involving working closely with a wide range of teams. Contact was made with

staff in this role at each of the Health Authorities where the teams in the sample were

located. They were asked to provide ratings of effectiveness and innovation for all of

the teams from their area that were participating in the research. External ratings of

effectiveness were obtained for 84 teams. They were rated on the same 21

effectiveness dimensions included in the primary health care team questionnaire.

Examples of the dimensions include:

Ø The extent to which teams made efficient use of the practice budget

Ø The extent to which teams previewed and adjusted skills in line with

the identified health care needs of the practice population

External ratings – changes introduced by teams

Three representatives from Primary Care rated the changes or innovations

introduced by the teams (reported in the questionnaire for individual team members).

One of the raters was employed by a Local Health Authority and had responsibility

for developing team working in primary care. Another was employed by a community

trust, also in a role which supported primary care teams. The third rater was a part

time general practitioner who had been involved in development and research

projects in primary care. They rated teams on four dimensions (West & Anderson,

1996):

• magnitude - how great would be the consequences of changes introduced

• radicalness - to what extent the status quo would change as a consequence

• novelty - how new in general were the changes

• impact - to what extent changes would improve PHCT effectiveness.

Using the ICC (2) (Shrief & Fleiss, 1979) the inter rater agreement was calculated for

each dimension: Magnitude - 0.663, Radicalness - 0.630, Novelty - 0.539, Impact -

0.779.

Sample Details

In this section we describe characteristics of the primary health care team sample

that participated in the questionnaire and interview component of the research

programme. Following the model used to guide this research (see page 1) we

describe four categories of inputs:

• team task - Indicated by the size of the practice population and its

fundholding status

• team composition - team size, ratio of part-time team members, gender mix,

number of occupational groups represented in the team

• health care environment - the Jarman Index (an index of social deprivation),

its location (city, urban, urban/rural or rural)

• organisational context - the NHS Region within which the team is located.

We also describe the relationships between these four domains of inputs. The

reader will also find an account of the frequency and content of meetings held in the

teams, and of the team members’ perceptions of leadership in the team in

subsequent chapters.

Team Task

Practice population or ‘list’ size

The practice population or ‘list’ size ranged from 1500 to 21,850. The mean size was

6,902 patients with a standard deviation of 4,692 (see Figure 2.2).

Figure 2.2: Percentages of primary health care teams with patient populations or ‘listsizes’ of various sizes

Fundholding status

Forty one percent of teams in the sample had fundholding status.

0 %

5 %

1 0 %

1 5 %

2 0 %

2 5 %

3 0 %

<25

00

2501

-

5000

5001

-

7500

7501

-

1000

0

1000

1-

1500

0

1500

1-

2000

0

>20

000

List Size

List Size

Location

Four location categories were used: city, urban, urban/rural and rural. Sixty-five

percent of teams were located in urban areas other than cities, 20% in cities, 6% in

areas described as both urban and rural, and the remaining 8% were in rural

locations.

Health care environment

Jarman index

The Jarman index is a measure of social deprivation. The higher the score the

greater the health needs of the practice populations served by the team.

Scores in the sample included in this research programme ranged from 0% to

100. The mean Jarman score was 15.52 and the standard deviation 22.72.

Figure 2.3 shows the percentage of teams with each category of Jarman

score.

Figure 2.3: Team Location

Figure 2.4: Jarman index

0

5

10

15

20

25

0%0%

1 - 10%

1 - 10%

Frequency

0%

10%

20%

30%

40%

50%

60%

70%

City Urban Urban/Rural Rural

Whole time equivalents

Using hours worked to calculate team size enabled an estimate to be made of the

number of ‘whole time equivalents’. This statistic shows that team size varied from

1.49 to 31.9 members. The mean size was 9.35 with a standard deviation of 6.75.

The size of teams working in so-called ‘single handed GP practices’, ranged from

1.88 to 16.13, with a mean of 7.48 and a standard deviation of 3.74.

Number of GPs

The number of GPs in the teams ranged from 1 to 11. The mean number of GPs

was 3.7 and the standard deviation was 2.4 (Figure 2.5).

Whole time equivalent GPs

The range of whole time equivalent GPs was from one toten. The mean was 3.16 and the standard deviation 2.0.

Figure 2.5: Number of GPs in the primary health care teams

Gender

The majority of team members (85.5%) of the Primary Health Care team sample

were female. The break down for gender by occupational group is shown in Figure

2.7 and this reveals that the only imbalance in favour of men is in the highest status

group – GPs. Otherwise, primary health care is a domain in which women form the

vast majority of the workforce. Primary health care is largely in the hands of women

in the UK. And of course, this has important implications for our understanding of its

0 %

5 %

10%

15%

20%

25%

30%

1 2 3 4 to 5 6 to 8 9 to11

Team Size (no of GP's)

functioning, issues of team processes and, given the anomalous preponderance of

male GPs, of leadership issues.

Fig 2.7: Distribution of gender by occupational group in primary health care teams

GPs Practice

Nurses

Trust

Nurses

Admin

staff

Managers PAMs Other

Male 99 5 16 10 24 8 1

Female 75 160 249 378 61 42 4

Fig 2.8: The distribution of mean age across the occupational groups

Mean Age Standard Deviation

GPs 41.2 7.6

Practice Nurses 42.1 7.6

Trust Nurses 42.2 8.3

Admin/clerical 43.7 10.2

Managers 44.3 8.4

PAMs 41.6 11.0

Others 40.8 8.2

Age

The distribution of age across the occupational group is shown in Figure 2.8.

Occupational Groups

The majority of teams comprised: GPs, practice nurses,administrative staff, district, health visitors and one ormore practice managers. Less than 15% of teams did notinclude trust nurses, and only 6% of teams had nomanager. Twenty four percent of teams included ‘other’types of staff (counsellors, community psychiatricnurses, physiotherapist etc).

Proportion of occupational groups

The proportion of each of the main occupational groups in the teams is shown in

Figure 2.9.

Figure 2.9: Proportion of occupational groups in the sample

A different picture of team composition emerges when hours worked is taken into

account and team membership calculated using 'whole time equivalent' figures. This

shows that largest grouping is GPs, with only 10.7% of the input to the team being

provided by trust nurses.

2.10: Proportion of occupational groups in the sample

GPs

Practice Nurses

Trust Nurses

Admin/clerical

ManagersPAM's Others

GPs

Practice NursesTrust

Nurses

Managers

Admin/clerical

PAM's Others

Organisational Context

Figure 2.11 Regional variations

Location Number of teams

South East and London 18

Midlands 4

East Anglia 4

Nottinghamshire 18

South Yorkshire 43

West Yorkshire 13

Teams were distributed across six regions with thedistribution shown in Figure 2.11.

Relationships between Input Factors

In this section we describe the relationships between aspects of team task, team

composition, health care environment and organisational context. As we might

predict, there are some important and significant relationships between them.

• The number of team members and the number ofpatients on the team’s list were positively correlated(0.85), with an average of 291 patients on a team’slist per member of staff. This ratio did not varyacross location (city, urban, urban/rural, rural),Jarman index, or average of number of hoursworked by team members.

• There were no significant differences in the composition of the teams

between different types of locations (city, urban, urban/rural, rural). Nor were

there differences in the composition of fundholding and non-fundholding

practices.

• There was a significant relationship between Jarman score and number of

managers; teams with fewer managers had a higher Jarman score.

• Teams with a higher Jarman index also had significantly more ‘other’ types of

staff in the team. This may reflect the fact that the range of services required

is much greater in socially deprived than socially enriched areas.

• There was a higher proportion of Professions Allied to Medicine (PAMs) in

teams with a larger list size.

Jarman index / location

73 teams provided a Jarman index score of more than 0%. The mean score was

15.52 and the standard deviation 22.72. These were distributed across locations as

follows:

City – 18

Urban – 39

Urban / rural - 5

Rural – 6

Unclassified - 5

• The Jarman index for city practices was significantly higher (mean = 32.6%)

than for urban practices (mean = 11.9%)4

Qualitative Research Methods

Research in the second stage of the research programme explored in depth, and

using a variety of consultation and qualitative research methods, all issues of team

functioning and effectiveness. The methods used are shown in Fig 2.12.

4 The data collected as part of this research can be subjected to much further analysis andinformation extraction. The researchers are committed to working with others to ensure themaximum exploitation of this hard won data set. If there are analyses readers wish to conductthe researchers would urge them to contact the first author of this report.

Figure 2.12: Consultation and Qualitative Research Methods used for Primary HealthCare Teams

Analysing PHCteam processes

Development of PHCobjectives

Development of PHCeffectiveness measures

§ Audio and videorecording of twomeetings for eachof twelve teams

§ Series of four nationalworkshops withdomain relevantexperts from primarycare.

§ Work with 12 teams tovalidate objectives.

§ In-depth work with two PHCTsto develop measures.

§ Training and dissemination toten PHCTs.

§ Training team facilitators.

Video and audio recordings of team processes

All teams involved in the questionnaire and interview components of the research

programme were invited to participate in the next stage of the research. This involved

analysis via video and audio recording of two of their team meetings. Teams were

selected randomly for this element of the research. Twelve teams volunteered.

We selected meetings that were multidisciplinary in composition and that involved

decision-making (as opposed to information dissemination only). This is because

understanding team working in this context demands that we observe professionals

from different backgrounds working together dynamically, and integrating their

different perspectives to initiate action and change. Multi-disciplinary meetings were

those in which a range of disciplines (doctors, nurses, health visitors, practice

managers, etc.) was represented and participated. For the most part, primary health

care teams allowed us to observe the practice business meetings, in which the day-

to-day running of the practice was discussed. In one team, the GP partners made all

decisions affecting the practice. In this case we recorded the partners’ meeting.

Wherever possible, we recorded two meetings of the same type for each team.

Dates for meeting recordings were at the discretion of the practice, so the two

meetings recorded were not always in sequence. Researchers requested that

recorded meetings should be held in their usual locations, with their usual meeting

protocols (agendas, minutes, chairing procedures, etc.), and that attendance should

be the same as if the meeting was not being recorded. The researcher who managed

the recording equipment made herself as unobtrusive as possible. Meeting size

ranged from three people to twenty-five.

Audio recording was done with two omni-directional PZM tabletop microphones

linked to different channels of a high quality audiotape recorder; the microphones

were set up so as to maximise channel differentiation but to be unobtrusive enough

that participants would not move them. A single static video camera on a tripod was

trained to record the gross movements of as many of the participants as possible;

this record was used only to aid speaker identification during transcription. Before

each meeting was opened all participants introduced themselves and their

occupation and upon the basis of this each was allocated a speaker number.

Therefore the first person to introduce him or herself became speaker 1, the second

speaker 2 and so on.

Meetings were transcribed from the audiotapes by an audio typist who had not

attended the meeting. Audio typists transcribed complete contributions in order,

according to when they began, labelling each contribution by speaker number, but

did not code finer timing information. Speaker identification was facilitated both by

the video recording and by a seating plan drawn up during the meeting by the person

recording the meeting. A contribution was defined as a period of speech from one

individual in which the only major pauses coincided with silence from the other

speakers, so that the pause was likely to be caused by the speaker thinking and not

by the speaker listening to someone else's contribution. Under this definition,

speakers cannot follow themselves in the speaking order. Overlapped speech was

transcribed, with the extent of the overlap roughly marked. Infrequently, parts of the

meetings were omitted because they were so badly overlapped that we could not

track individual contributions. After transcription, the transcripts were completely

anonymised taking out all staff, patient, place names, place and local authority

names or possible team or person identifiers.

An example transcription excerpt is given in Figure 2.13. Transcription proceeds one

contribution per row. Column one contains the speaker number. Column two

contains the words said, with coding information in a different font, and column three

contains any notes which the transcriber wished to make (for instance, about people

entering or leaving the room).5

5 In previous work using these methods on four to twelve person meetings, transcribers wereable to agree very reliably who made any one contribution; using the kappa statistic, K=.93, k= 2, N = 230, with an average of 2% and a maximum of 6% non-backchannel contributionsleft as unidentified.

Figure 2.13: An example of the layout of the transcription format used for PHCTmeeting transcription

1 Shall I open /4 the meeting

4 Yep, let’s get on with it.

3 My apologies I am going to have to leave beforethe end. I have an appointment in Place 1.

1 Are you skiving off?

Group laughter

Because one of the factors of interest in our study is how well teams communicate

across disciplines, our analysis relies on a classification of meeting participants by

occupation. For ease of reference, categories are identified by colour as well as

number. For primary health care teams, Figure 2.14. shows the categories used.

Figure 2.14: Categories used for PHCT meeting participants

1 GPs

2 practice managers

3 practice nursing staff, including nurse practitioners

4 attached staff (mostly health visitors, midwives and district nurses)

administrative staff (mostly secretaries and receptionists)

6 miscellaneous (visitors, resident caretakers, medical students)

Development of Performance Measures for PHCTs

In the broader organisational literatures on team effectiveness, a widely adopted

approach is the Productivity Measurement and Enhancement System (ProMES)

based on research by Naylor, Pritchard & Ilgen (1980) (see also Pritchard, 1995).

Effectiveness criteria are established in group discussions with team members and

managers. The variables are then “psychologically scaled” to a common

effectiveness scale. Based on group consensus about expected levels of

effectiveness, which are given a zero value, maximum effectiveness levels (set at

+100), and minimum levels (-100) are set. Each variable is also weighted in terms of

its perceived contribution to the overall effectiveness of the team or organisation.

The system is then used to set objectives, develop indicators, monitor and improve

performance and give feedback to the team (Pritchard, 1995). The approach has

been spectacularly successful in many settings (Pritchard, 1995) and is promising for

primary health care, because of the sophistication of the approach, its theoretical

robustness and practical utility in complex contexts.

The ProMES was implemented in three main stages:

1. Core objectives for primary health care teams were developed using the

constituency approach and ProMES in four national workshops with

representatives from Primary Care.

2. Usable ProMES effectiveness measures were developed and applied within

primary health care teams.

3. Primary health care team members and trust representatives were trained to

develop and implement effectiveness measures using ProMES in primary

health care teams.

An initial ‘stakeholder analysis’ identified 13 stakeholders in primary health care.

These included:

Ø GPs

Ø Health Visitors

Ø District Nurses

Ø Practice Nurses

Ø Midwives

Ø Administrative staff

Ø Department of Health

Ø NHS Executives

Ø Patients

Ø Health Authority

Ø Researchers

Ø PAMs

Ø CPN

Advice was sought from contacts in primary health care about key experts who could

represent the views of each stakeholder group, and about whether the initial list of

stakeholders was sufficiently comprehensive. The experts suggested by the contacts

were sent information about the research programme, invited to attend the four one-

day workshops, and asked to suggest additional or alternative key experts who could

also make a contribution. In addition, professionals who were currently engaged in

clinical practice in primary health care teams were invited. The majority of those

contacted were able to commit themselves to attending two or three of the

workshops. A full list of those attending and their institutional affiliations is given in

Appendix II.

During the workshops focus group methods were used. Delegates were divided into

three working groups. These were designed so that (a) a range of stakeholder views

was represented, and (b) one or two or group members had attended most or all of

the workshops and so could share with new members the learning and experience

from previous workshops. Each group worked with a trained facilitator, and a note

taker recorded the group discussion and the decisions made.

Workshop 1

Objective: to develop objectives for primary health care.

The delegates were presented with a set of objectives for primary health care

developed by the researchers (based on the work of Poulton & West, 1994) and then

worked together to discard, add or refine objectives. The revised objectives were

discussed with members of four primary health care teams (who endorsed their

relevance and value), and combined into a single list.

Workshops 2 and 3

Objective: to develop measures of effectiveness in relation to the primary health care.

Delegates were presented with the refined and agreed objectives for primary health

care. Each group worked on developing effectiveness measures for objectives.

Workshop 4

Objective: to plan the implementation of effectiveness measures in primary healthcare.

In the fourth Workshop, participants critically appraised the objectives and measures

developed, and considered how they could be applied in practice by PHC teams and

others. This session was used to plan the implementation and evaluation of the

effectiveness measures in practising teams. Seven core objectives, with associated

sub-objectives, were identified and agreed by the professionals attending the

workshops and 19 effectiveness measures were developed (see Appendix II).

Work implementing performance measures was carried out in two phases. In the

first phase, we carried out in-depth work with two primary health care teams to

develop performance measures, based on the objectives and measures developed in

the four national workshops, and used these to provide feedback on team

performance. Details of this work are provided in Appendix III. We worked with one

team over a period of 15 months, and with the second for a period of 8 months. A

design team was established in each PHCT that included at least one representative

from each of the occupational groups in the team. In one-hour workshops held every

fortnight, ProMES was used to develop performance measures specific and

appropriate to each team. Teams carried out further development work between

meetings such as gathering data and consulting colleagues.

In the second phase, PHCT representatives and trust employees attended a

‘ProMES in Primary Health Care’ training programme. etters inviting representatives

from PHCTs to attend the ProMES training were sent to 60 PHCTs that had

participated in the first stage of the research (all teams with a response rate of 50%

and above). Letters were sent to the practice manager, senior health visitor and

senior partner. Follow-up phone calls were made to the teams, but representatives

from only two attended the training. The other participants were service

representatives, employed by community trusts, to support and develop primary care

team working.

The training programme included an overview of the ProMES approach; the

development of performance measures; guidance on running ProMES workshops in

PHCTs; training in how to collect and use performance information. The programme

for the training is outlined in Appendix IV. After the training, three follow-up

workshops were held with participants. The purpose of these was to provide support

to those using ProMES with primary health care teams, and to critically review the

measures developed in the training. Feedback from the participants about the

success of their interventions in teams has been positive and suggests there is real

benefit to all aspects of primary health care team functioning from employing this

approach. At the same time, it is a demanding exercise that requires commitment by

team members to implement.

The results from stage 1 and 2 of the research programme are described in the

following two chapters. Details of the findings from the primary health care team

surveys and external ratings are outlined in Chapter 3, and the objectives and

effectiveness measures developed for primary care, outlines in Chapter 4.

.Chapter 3____________________________________________________________________________

Primary Health Care TeamResults from Survey and External Ratings

Summary of Findings

• Large PHC teams are rated as more effective and innovative by external raters.

• The greater the number of professional groups represented in the primary health

care teams, the more highly rated is the innovativeness of the team.

• The better the team processes and reflexivity, the more innovative they are rated

by external raters.

• The greater the number of team meetings, the higher the level of innovation in

primary health care teams.

• PHC teams with clear leaders have good team processes.

• Conflict over leadership leads to poor quality team working. However, teams

where leadership roles are shared are more innovative.

Introduction

The data analysis explored two main questions

• Is there an association between the composition of a primary health care teamand team processes?

• Is there an association between the composition and processes of the primaryhealth care team and the effectiveness of the team?

The team characteristics, team processes and measures ofteam effectiveness are summarised in Figure 3.1.

Figure 3.1: Team Inputs, Processes and Effectiveness Variables

Characteristics Processes Effectiveness

IndividualAgeGenderEthnicityGradeTime in jobTime in team

TeamOccupational groupsSize (number of members)Number of GP'sHours workedGrade MixTraining

Team contextList sizeLocation (urban, rural, city)Jarman indexFundholding statusDispensingPurchasingQuality of premisesExternal contactsRelationships with HA

Team processesParticipationInnovationObjectivesEmphasis on qualityReflexivityNumber meetingsTypes of meetingsFrequency of meetingsDecision makingLeadershipIntegration andcommunication in the group

Team ratingsOrganisationTeam workingPatient focusInnovation

External ratings (innovation)MagnitudeRadicalnessNoveltyImpact

External ratings (effectiveness)OrganisationTeam workingPatient focus

Types InnovationsQuality of careExternal collaborationResponsibility of healthUse of resourcesProfessional developmentTeam satisfactionResponsivenessStress (GHQ 12)

Team inputs

Information about the team members’ ages, gender, ethnicity, grade, professional

group, employer, tenure and team leadership was collected from each team member.

Information was also gathered on team size, hours worked, qualifications, training,

list size, practice location (urban/rural/city), Jarman index, fundholding status, and

whether the practice was purchasing and/or dispensing.

Team Processes

Individual team members rated team processes on six dimensions: participation;

support for innovation; clarity of team objectives; emphasis on quality; reflexivity; and

integration. The variables participation, support for innovation, clarity of team

objectives and emphasis on quality were very highly correlated and were combined

to form one variable describing team processes. Information about decision-making

processes, communication, number and types of meetings, who attended meetings,

and how the team was managed was collected from practice managers. The

information on team meetings was categorised according to who contributed to

operational, strategic and clinical decisions. In addition, a new variable

‘interdependence’ was developed which assessed the extent to which there were

mechanisms within the team to encourage interdisciplinary communication.

Team effectiveness

This was assessed using information from a variety of sources. Team members

rated their teams’ effectiveness on three dimensions: team working, organisational

efficiency and patient orientation. Team members also rated their teams’

innovativeness and described the innovations implemented by the team in the

previous year. These reports were categorised to determine the types of innovations

implemented. External raters assessed the innovations reported by the teams on

four dimensions: magnitude; radicalness; novelty and impact on team effectiveness.

External ratings of team effectiveness were provided by Health Authority

representatives on two dimensions – clinical and organisational. Individual team

members also completed the GHQ-12 (a measure of mental health or psychological

stress). The measures of interest for this report are overall effectiveness,

effectiveness of patient-centred care (both externally rated and self-rated), overall

innovation (both externally rated and self-rated), number of innovations to do with

healthcare, and mental health measured by GHQ-12.

Results

The main method of analysis was multiple regression. For each dependent variable,

possible predictors were split into groups according to type of variable (e.g.

occupational group, team context), and stepwise regression was used to identify

those which might ultimately predict the dependent variable. The second stage of

each analysis involved entering all those identified into a further stepwise regression,

to find out which variables had significant effects independent of other predictors.

This way, process variables were predicted by team characteristics, and

effectiveness, innovation and mental health were predicted by team characteristics

and team processes.

Since there was often evidence of relationships between size and other variables,

this was always dealt with first. Where relationships were apparent, later analysis

revealed whether this was due to team size per se or another feature of having a

larger team.

Question 1 – Is there an association between the composition of a primary

health care team and team processes?

There was no evidence that PHC team size had an association with any team

process except frequency of meetings. Here we see that teams of 20 or less have,

on average, 2.6 meetings a month; teams of 20-30 members have 6.1 meetings a

month, and teams of over 30 have 6.5 meetings a month.

Other predictors of team processes (after the second stage of analysis) are shown in

Figure 3.2.

Figure 3.2: Relationships between Team Composition and Team Processes

Dependent variable Predictor variables β p R2

Team processes Proportion of managers 0.305 0.010Proportion of “other” staff 0.253 0.032No. of GPs (WTE)1 -0.244 0.036 0.192

Reflexivity None

Integration Proportion of managers -0.256 0.035 0.065

Number of meetings No. of practice nurses 0.418 0.001 0.175

Consensus on leadership Proportion of “other” staff 0.366 0.002No. of managers -0.290 0.012 0.237

1WTE = whole time equivalents

Patterns emerging here are mainly to do with the representation of managers and

“other” staff types in the teams (anyone other than GPs, nurses, admin/clerical staff,

managers and PAMs). Having a larger proportion of “other” staff in the team has a

positive effect on team processes and consensus on leadership. A larger number of

managers also has a positive effect on team processes, but has the opposite effect

on integration. There is a negative association between a larger number of managers

and agreement about who leads the team.

Question 2 – What affects the effectiveness and innovation of a primary

health care team?

Team size was positively associated with a number of dimensions of effectiveness

and innovation, as shown in Figure 3.3. Generally, larger teams were rated as more

effective by external raters and introduced more innovations overall, and specifically

in relation to patient care.

Figure 3.3: Relationship between team size and ratings of effectiveness, innovationand mental health

Variable Correlation p

Overall effectiveness (external) 0.284 0.012

Overall effectiveness (self-rated) 0.086 0.401

Effectiveness of patient care (external) 0.255 0.002

Effectiveness of patient care (self-rated) 0.125 0.222

Innovation (external) 0.403 < 0.001

Innovation (self-rated) 0.123 0.226

Number of innovations re: patient care 0.255 0.013

Mental health 0.056 0.585

Further analysis revealed that the relationships between team size and all the

innovation variables was curvilinear, with teams of sizes around 40 being the most

innovative. Notice that there were no relationships between team size and self-rated

effectiveness, innovation or mental health.

It is also interesting to note the associations between team size and the individual

items of the external effectiveness ratings, to see what aspects of effectiveness are

most related to team size. These are shown in Figure 3.4. Larger teams appear to

be more responsive to patients and are more likely to conduct clinical audit.

Figure 3.4: Relationship between team size and individual externally ratedeffectiveness items

Effectiveness item Correlation p

Provision of information about services 0.131 0.236

Implementing procedures for dealing with patients’comments, suggestions and complaints

0.269 0.015

Maintaining clinical competence in line with patientneeds

0.186 0.083

Auditing clinical practice 0.263 0.017

Setting protocols 0.373 0.001

Commitment to professional and personal development 0.273 0.012

Understanding and valuing roles of all members 0.003 0.981

Implementing a clear strategy for communication 0.132 0.203

Profiling health needs and targeted interventions 0.160 0.165

Reviewing and adjusting skill mix 0.209 0.068

Collaborating with other agencies 0.094 0.389

Making effective use of budget 0.126 0.265

Implementing recommendations of the PHC Charter 0.311 0.004

Concentration on achievement of The Health of theNation targets

0.258 0.024

The main reasons for larger teams being more effective appear to be their

effectiveness in setting protocols and implementing recommendations of the PHC

charter.

Results of the stepwise regression analyses of effectiveness on team characteristics

and processes are shown in Figure 3.5.

Figure 3.5: Relationships between Team Composition and Processes, and

Ratings of Effectiveness

Dependent variable Predictor variables β P R2

Overall effectiveness No. of admin. staff 0.400 0.003(external) Proportion of GPs -0.279 0.035 0.199

Overall effectiveness (self-rated)

None

Effectiveness of patient care(external)

Team size 0.357 0.010 0.127

Effectiveness of patient care Team processes 0.632 <0.001 0.400

(self-rated)

The relationship between self-rated effectiveness of patient care and team processes

is not entirely surprising, given that both variables were constructed from the

individual questionnaires sent out and hence this analysis is prone to common

method variance. The relationship between team size and externally rated

effectiveness of patient care is shown in Figure 3.6. This relationship is more reliable

and suggests that better patient care is delivered in larger primary health care teams

sizes, up to 30 to 40 members.

Figure 3.6: Relationship between team size and effectiveness of patient care(externally rated)

Team size

706050403020100

Effe

ctiv

enes

s of

pat

ient

car

e

6

5

4

3

2

1

Figure 3.7: Relationships between Team Composition and Ratings of Innovation

Dependent variable Predictor variables β P R2

Innovation – overall Professional diversity 0.308 0.002Reflexivity 0.318 0.001Team size 0.290 0.003 0.363

No. of healthcare innovations Professional diversity 0.263 0.024No. of practice nurses 0.299 0.011Team processes 0.342 0.005Lack of clear leadership 0.274 0.023 0.298

Innovation (self-rated) Reflexivity 0.384 0.018Team processes 0.315 0.050 0.454

It seems that professional diversity in the team, reflexivity (taking time out to review

objectives, strategies and processes) and team processes all have positive effects on

innovation. Some of these relationships are illustrated in figures 3.7 to 3.11.

Figure 3.8: Healthcare innovations and professional diversity

Figure 3.9: Overall innovation and professional diversity

1

1.5

2

2.5

3

3.5

4

4.5

5

6 or fewer 7 8 9 10 11 12 or more

Number of professions represented in team

Ove

rall

inno

vatio

n -

exte

rnal

rat

ing

0

1

2

3

4

5

6

7

< 7 7 8 9 10 11 12 > 12

Number of professions represented in team

Mea

n nu

mbe

r of

inno

vatio

ns r

e:

qual

ity o

f he

alth

care

Figure 3.10: Relationship between reflexivity and overall innovation

Reflexivity

6.05.55.04.54.03.53.02.5

Ove

rall

inno

vatio

n

2.0

1.5

1.0

.5

0.0

-.5

-1.0

-1.5

LFigure 3.11: Relationship between team processes and number of innovations inhealthcare

Team climate

4.24.03.83.63.43.23.02.82.62.4

Inno

vatio

ns in

hea

lthca

re

10

8

6

4

2

0

-2

Bearing in mind that a larger score represents poorer mental health, results show

that teams which are composed of members who are relatively similar in terms of

age, have a larger proportion of managers, and avoid conflict over who leads the

team, are characterised by better mental health for their members. This is shown in

Figure 3.12.

Figure 3.12: Relationships between Team Composition, Processes and TeamMembers’ Mental Health

Dependent variable Predictor variables β p R2

Mental health Proportion of managers -0.420 <0.001

Age diversity 0.363 0.001

Conflict over leadership 0.253 0.017 0.375

Frequency of team meetings

Frequency of meetings was also examined as an explanatory variable. Figure 3.13

shows that frequency of meetings in primary health care team predicted external

ratings of innovation. Coming together to discuss objectives, exchange information

and make decisions is likely to lead to the generation of ideas for new and improved

services and ways of working.

Figure 3.1: Frequency of PHC team meetings as a predictor of innovation

Dependent variable β p R2

Innovation – overall (external) 0.242 0.026 0.059

Number of innovations re: healthcare 0.198 0.072 0.039

This result is illustrated in Figure 3.14.

Figure 3.14: External Ratings of Overall Innovation and Number of Meetings

2

2.2

2.4

2.6

2.8

3

3.2

3.4

3.6

3.8

4

1 or less 1 to 4 4 to 6 6 to 10 More than 10

PHCT Meetings per month (average)

Ove

rall

inno

vatio

n

It was also shown that this effect is independent of both self-rated processes and

team size.

Leadership

Research evidence suggests that leadership is an important factor contributing to

team effectiveness. We therefore explored the contribution of leadership to team

effectiveness and innovation in primary care teams separately. We explored the

extent to which there was a clear leader in the PHC teams, and who was regarded as

the leader. Only a third of PHC teams reported having a single clear leader. Nearly

half reported having a number of people lead the team, which, in most contexts, is

likely to cause considerable confusion. The most frequently named leader of PHC

teams is the Practice Manager. Only a third of team members nominated a GP.

Clarity of leadership was examined as an explanatory variable. Figure 3.15 shows

that team processes were poorer where there was no clear leadership, (from either

one individual or several people), or where there was conflict over leadership.

Figure 3.15: Clarity of Leadership in the PHC team predicting processes

Dependent variable Predictor variables β p R2

TCI mean score Lack of clear leadership -0.311 0.001Conflict over leadership -0.294 0.002 0.180

Reflexivity Lack of clear leadership -0.366 <0.001Conflict over leadership -0.250 0.008 0.193

Integration All having leadership roles 0.214 0.035 0.046

Figure 3.16 shows that lack of clarity about leadership inthe teams also predicted lower levels of effectiveness asrated by the team. Figure 3.16 also shows thatinnovation, rated externally, was higher in teams withshared leadership, or in teams where more peoplereported that everyone had leadership roles. Of course, itmay be that shared leadership is itself an innovation withprimary health care. Teams which had no clearleadership or conflict over leadership were less likelythan others to reviews and modify their objectives,strategies and team processes.

Figure 3.16: Clarity of Leadership in the PHC team predicting externally ratedeffectiveness and innovation

Dependent variable Predictor variables β p R2

Overall effectiveness(external)

Lack of clear leadership -0.249 0.029 0.062

Overall effectiveness (self-rated)

Lack of clear leadership -0.299 0.003 0.089

Effectiveness of patientcare (external)

None

Effectiveness of patientcare (self-rated)

Lack of clear leadership -0.215 0.034 0.046

Innovation (external) Shared leadership 0.237 0.020All having leadership roles 0.216 0.033 0.087

Innovation (self-rated) Lack of clear leadership -0.365 <0.001 0.133

No. healthcare innovations None

Mental health Conflict over leadership 0.294 0.003 0.086

These effects, and those for externally rated effectiveness, are all entirely mediated

by group processes suggesting that the mechanism by which leadership influences

effectiveness is through developing good team processes, such as shared

objectives, participation, emphasis on quality and support for innovation.

We also find that there is less clarity of leadership in teams which have a greater

proportion of part time workers (r = 0.309, p = 0.016), and there is less likely to be a

single clear leader in teams with greater professional diversity (r = 0.309, p = 0.016).

Both of these support the finding in Figure 3.2 that there is less consensus on who

the team leader is in teams with a larger proportion of “other” staff types.

Overall, the findings from this stage of the research reveal a very clear picture

of the factors predicting the effectiveness and innovations of primary health

care teams: size, clear leadership, professional diversity and integration

through regular meetings are key factors in predicting PHC team

performance. Of course, it could be that teams that innovate and are effective

have the confidence to recruit members from diverse professional

backgrounds, and are required to meet more often because of the innovations

they introduce. Clear leadership may emerge as a consequence of innovation

and as a consequence of the cohesiveness arising from effective

performance. Such interpretations are feasible and need to be explored

empirically. However, the interpretation implied in our presentation of results

(inputs and process predict performance) is consistent with research into

effectiveness of teams from across a range of sectors and countries (Cohen &

Bailey, 1998, West, Borrill & Unsworth, 1999).

Chapter 4

Qualitative Research:

Developing Objectives and Effectiveness Measures

for Primary Health Care Teams

Summary of Findings

• Using the constituency approach seven core objectives were developed which

were judged relevant and covered all the main aspects of primary health care

activity.

• Measures which could be used to measure performance on each of the

objectives were developed in workshops with a range of primary health care

stakeholders.

• Primary health care teams used the ProMES approach to develop measures

which could be used to measure their performance against the objectives for

primary care.

• Primary health care teams were able to use the measures developed to get

feedback on their performance, and use this information to introduce

improvements in patient care.

Measuring Effectiveness in Health Care

There is little agreement in primary health care about what constitutes effectiveness.

One reason for this is that primary health care comprises a wide range of

stakeholders (health care professionals, trusts, health authorities, patients, carers,

voluntary groups) each with their own aims, objectives and priorities which influence

how effectiveness is conceptualised. In addition, there is considerable variation in

philosophies of care among the professionals groups within primary care (Toon,

1994), and different approaches and perspectives on what is judged to be high

quality of care (Maxwell, 1992). One consequence of this is that health care will be

judged as more or less effective depending upon the criteria adopted by the

particular stakeholder, or on the philosophy or care espoused by a professional

group.

To enable these differing priorities and perspectives within health care to be taken

into account the qualitative research carried out by the research team used the

constituency approach (Connolly 1990) to develop objectives, and the Productivity

Measurement and Enhancement System (ProMES) developed by Naylor, Pritchard

and Ilgen (1980) to develop effectiveness measures. There were two main stages to

the work: developing objectives and effectiveness measures in national workshops;

developing effectiveness measures with primary health care teams.

Stage 1 - National Workshops: Developing Objectives and EffectivenessMeasures for Primary Health Care

The aim of this stage of the qualitative research was:

• To develop a set of objectives for primary health care which was acceptable to all

perspectives in primary health care

• To develop effectiveness measures which were acceptable to all perspectives in

primary health care.

The constituency approach was used to develop objectives for primary health care in

four national workshops with representatives from primary health care (see chapter

2). These objectives were then validated in workshops with representatives from 12

primary health care teams. The objectives and sub-objectives developed as a result

of the workshops and consultations with primary health care team representatives

are shown in Fig 4.1.

Figure 4.1: Core Objectives for Primary Health Care teams

Promote, maintainand improve health

§ Provide high quality health care

§ Accurate identification of individual and population healthcare needs

§ Review and improve the effectiveness of health careprovision

§ Manage illness, injury and disease taking account ofagreed standards and evidence based practice

Enable personaland communityresponsibility forindividual health

§ Enable patients/clients to make informed decisions abouttheir own health.

§ Proactively encourage positive health behaviour

§ Implementation of health education and preventativecare programmes

Efficient use ofresources

§ Human resources – skills, knowledge, expertise, time

§ Physical resources – budgets, equipment, premisesContinuouspersonal andprofessionaldevelopment

§ Individual annual training plans which take account of theplans of the PHCT

§ Equal access to training/development resources

High team membercommitment, stressand satisfaction

§ Team working

§ Mechanisms for reviewing and acting upon staffdissatisfactions, conflicts and complaints

Responsiveness toclients andcommunity

§ Gather information and feedback from clients/communitystakeholders/opinion leaders

Collaboration andpartnership withother relevantorganisations

• Build external relationships with clear objectives and highlevels of participation, interaction and trust

The first aim of the national workshops was to get agreement on the objectives for

primary health care, and to develop a set of objectives that cover all aspects of team

activity. The work carried out by workshop participants, and the subsequent

amendments made as a result of the rating and discussions with PHCT

representatives, enabled this main objective to be achieved. Given the diversity of

views, agendas and perspectives in primary health care it was a major achievement

that by the end of the four workshops agreement had been reached.

The second aim of the workshops was to develop effectiveness measures for primary

health care. A preliminary set of effectiveness measures was developed during the

workshops that reflect the range of stakeholder perspectives. These can be

developed further and used by primary health care teams. The research team

carried out additional work on some of the preliminary measures, developing

indicators of team effectiveness that could be used to measure performance.

Effectiveness Measures Developed in the Workshops

Objective 1 - Promote, maintain and improve health

Quality of care

• Patient Charter taken into account

• health promotion activities carried out

• appropriate skill mix in the team to meet patient needs

• measure - % of appropriate consultations as % of total consultations

• measure - appropriate immunisation rates (without adverse incidents)

• measure - effective management and knowledge of chronic diseases (epilepsy,

diabetes, asthma)

• measure - quality of patient consultations

• measure - appropriate admissions to hospital

Accessibility of service

• appropriate number of surgeries offered and times (also flexibility)

• appropriate length of consultation (also flexibility)

• waiting times. Time taken to get routine and emergency appointments (with any

member of the PHCT)

• availability of non face to face contact i.e. telephone access

• clients seen consistent with the severity of their needs (e.g. emergencies seen

quickly)

Chronic disease management

• effective management and knowledge of incidence of critical diseases: E.g.

Epilepsy, Diabetes, Asthma

In terms of -

Diagnosis

Registers

Protocols

Interviews

Referral/use of other services

• appropriate referrals to other services. Such a measure could indicate a lack of

skills in the team or illness in the community

• number of effective or appropriate contacts with agencies such as palliative care,

social services etc. i.e. good network of services

• level of appropriate access to the right services. Quality of partnerships and

alliances in referrals is important here. This indicator might also be linked with

the range of skills in the team

• identification and reduction of health and social care “grey areas” e.g., when

health care professionals do social care activities

• appropriate waiting times for admission to hospital i.e., for treatment from other

agencies

• rates of emergency admissions/self referrals

Treatment

• use of evidence based treatment and prescribing protocols

• appropriate intra-team referral. The group felt this was possibly more important

than referral to other agencies

• low adverse complications incidence

• care delivery derived from plan of care. Having action plans helps evaluation of

goals

• the team produces R&D strategy (based on consensus)

• the team produces clinical audit and clinical supervision action plans (based on

consensus)

• care packages/episodes of care (rather than just number of contacts)

• progress towards Health of the Nation targets

Identification of health needs

• identification of health needs and the mechanisms to adjust efforts to match

these needs

• utilisation of external bodies to identify service plans and needs met

Data Collection: Practice level

• measure - types of information collected demographics/diseases/conditions/

activity levels)

• measure - accessibility of data collected to PHCT

• measure - PHCT contributing to data

• compare with national/regional data

Data collection: Local, regional and national sources

• Assess completeness of data set

Use of data for:

• daily planning

• longer term planning - strategy/direction

• identifying gaps in provision and skill mix

• measure - number of action taken/changes made

- up-take of training

- modification of skill mix

- review process

- formal service plans

• budget allocation consistent with priorities

• Identifying and utilising opinion leaders in the community and community

networks

Accountability

• meeting NHS care standards

• meeting NHS reporting requirements

• meeting requirements of other appropriate external agencies

• progress towards Health of the Nation targets

Objective 2 - Enable personal and community responsibility for individualhealth

• effective health education and preventative health care programs

• appropriate immunisation rates (without critical adverse incidents). What is

appropriate will vary in accordance with local needs)

• information to patients and health education

- includes information and knowledge and explanation for patients

- making it personalised - so patient is recognised as an individual

• increasing knowledge about health in the population

- i.e. with employers, teachers etc.

• patients educated to make appropriate self-referrals to members of the PHCT.

Where is the locus of control, within the team or with the patient?

• number of health problems revealed by screening

• provision and take up of preventative health care programmes

Objective 3 - Efficient use of resources

• monitor appointment management - DNAs

• protocols: new, renewed, rejected

• use of accommodation/equipment

• develop skills inventory and monitor use of skills

• measure - input costs: GP: practice size

• measure - initiatives developed to use time effectively and review process

• measure - balance between outputs and resources/monitor over time

• measure - % of time with patients

• review duplications of roles/effort

• existence of evidence based prescribing protocols for practice (and review of

these). There is a need to close the loop between cost effectiveness and clinical

effectiveness of prescribing

• existence of evidence based treatment protocol (including shared protocols and

reviews)

• use of clinical guidelines (not just medical - so incorporates everyone in team)

• planned clinical audit

• degree to which safety standards were being complied with (Baseline could be

minimum standards set by Health and Safety Executive)

• how effectively the PHCT computer systems are being used

Objective 4 - Continuous personal and professional development

Development of skills

• Regular development and learning needs to be considered at the level of the

individual, the team, the national governing bodies, and the professional bodies

that monitor health care professionals

• strategy plan for training and development

- long term and short term

- individual skills/job description

- match health needs/individual needs

- match to organisational objectives

- who contributes to developing the plan

- take account of each individual’s understanding of ‘development’

• measure - commitment to development of skills in the team

• equal access to/management of training budget

• research activities carried out - how funded, quality and quantity.

• utilisation of a full range of training methods (e.g., on the job, networking)

• opportunities for job exchange

• skill sharing - opportunities and time spent

• mechanism in place to evaluate the effectiveness of training and development

that is done, including publication of the existence of training opportunities, time

available, equipment, instructors, etc.

• job performance of staff should be assessed on a regular basis using an agreed

upon procedure

Checklist of activities which contribute to the development of skills:

• availability of peer support/mentoring (inside and outside)/advocacy

• individual career development plans

• reviews of development plans

• IIP in place

• staff appraisal - linked to short term and long term goals

• identification of training needs - and review of these

• training equally available across team

• feedback on training attended/portfolio maintained/accreditation

• protected time available for professional development

• access to resources to support training and development

• indicators for learning and training

Team member mental health/stress

• procedures to resolve conflict between patient/practitioner needs

• procedures for taking account of personal needs/family commitments

• staff allowed to be off sick

Objective 5 - High team member commitment, stress and satisfaction

• measure - staff turnover/absence

• measure - how valued staff feel/commitment/grievances

• mechanism in place for reporting satisfaction levels back to team members and

dealing with dissatisfaction when it becomes a problem

Team development

• team participate in team development activities?

• availability of social budget

• shared understanding of roles and values

• the team contribute to the annual report/business plan?

• the team has regular meetings

• procedure for coordinating sub-groups and whole team

• procedures for communication

• procedures to integrate roles/expectations across professional groups

• processes for critically evaluating and improving decision making

• equal opportunity for participation in decision making

Learning organisations

• measure - support for innovation

• measure - resulting changes

Objective 6 - Responsiveness to clients and community

• a commitment to client satisfaction within the team

• complaints procedures in place

• accessible information produced for patients

• patient choice re health care

• take account of patient perceptions of improved health and stress

• use questionnaires/surveys to assess patient satisfaction

• user involvement in decisions about their own health

• actions taken in response to patient suggestions for improvement not the same

as complaints - giving patients the opportunity to make comments without feeling

as if they are complaining.

• carefully listening to the client

• giving clients the information to make informed choices

• getting inputs on client needs from clients, community and opinion leaders,

groups representing clients

Objective 7 - Collaboration with other organisations

• measure - staff use of skills and resources available

• effective contacts with related agencies and groups outside the PHCT

Development Work Carried Out by the Research Team

The research team carried out additional work after the national workshops refining

some of the measures of effectiveness identified in the workshops. These are

described below.

Objective 1 - Promote, maintain and improve health

• The PHCT would have a monthly (or more frequent) staff meeting where a

sample of cases was reviewed. This review would include the appropriateness of

who saw the client, what procedures used, and whether that client was handled

appropriately in all aspects. The measure would be the percentage of cases

which were considered as being managed appropriately. This would also be the

basis for discussion of what improvements need to be made for those specific

clients and for clients in general.

• The task of developing a health needs analysis can be broken down into

definable steps, e.g. get information on how to do such an analysis, decide on a

plan for doing the analysis for that particular PHCT, gather the information, put

the information together into a form that the PHCT can use to make decisions.

Each of these steps would be given a time for completion. The indicator would

be the percentage of the analysis completed compared to the anticipated time for

completion.

Survey on client perceptions of health improvement after treatment. For

example, each client is given a questionnaire or a sample of clients are called by

phone and asked about improvements. Measure is the percentage of clients

improving. For the various specific targets given by agencies outside the PHCT

such as immunisation rates, develop a scoring system whereby each level of

meeting the objective gets a certain number of points, e.g. if the target

immunisation rate was 80%, actually doing 80% would give 100 points, 60%

immunised would be 20 points, 70% 80 points, 90% 130 points, etc. The

number of points would be based in the importance of that target. The index

would be the percentage of actual points earned compared to the maximum

possible points received if all targets were met.

• The percentage of required reports completed on time

• The number of required reports returned by agencies requesting corrections or

additional information. (This would be an index of the quality of the reports.)

Objective 3 - Effective use of Resources

• Number of new initiatives developed that are designed to help team members

use their time better. These initiatives should also be reviewed on a regular basis

to ensure they are still effective.

• Percent client contact time as a percentage of total time. This measure gets at

how much time is devoted to clients. It does not measure how well that time is

being spent. Other indicators are needed to address this issue. (Note that this

indicator is one where there is probably an optimal level between the extremes.

Too little time with clients may suggest too much administration time. Too much

time with clients may suggest too little administration time.)

• Percentage of staff turnover over time. High staff turnover leads to inefficient

resource utilisation because it takes time to teach procedures to new staff and

work is lost as a departing staff member leaves. This measure would also be an

indicator for the satisfaction of team members.

• Percentage of appointments which are unfilled or where the client did not come.

Objective 4 - Continuous personal and professional development

• Training and development. A list of training and development experiences for

each person on the team would be developed each year. For example,

attendance at a certain type of conference, training on a piece of office

equipment, learning a new procedure, etc. This list would be the development

plan for that person for that year. There would be two measures for training and

development. The first would be the percentage of team members who had the

written plan. The second measure would be the percentage of the development

plan items actually completed.

• Which team members are reviewed, given feedback, and have a formal, jointly

developed action plan for making improvements.

Objective 5 - High team member commitment, stress and satisfaction

• Measure overall satisfaction on a monthly or bi-weekly basis with a very brief

questionnaire that would take no more than 2 minutes to compete. Measure

would be the percentage of staff indicating Satisfied or Very Satisfied with their

jobs.

• Staff turnover is also a satisfaction measure. Note this measure under Effective

Management of Resources.

Objective 6 - Responsiveness to clients and community

• Establish a formal procedure where clients can make complaints including a

process for following up on these complaints. Measure is the number of such

complaints which were not concluded to the client’s satisfaction within one week.

Stage 2 - Effectiveness Measures Developed by Primary Health Care Teams

The ProMES approach is based on a theory of motivation which proposes that effort

is maximised when there is a clear link between effort and outcomes, there is

agreement about what are valued outcomes, feedback is provided on performance

and the evaluation of performance is judged to be fair (Pritchard, Jones, Roth,

Stuebing & Ekeberg (1988). Research evidence shows that involving individuals in

the process of agreeing the valued outcomes from their work and developing

methods for assessing their performance has a greater impact on performance than

when these are imposed (Pritchard, 1995). The research team therefore carried out

ProMES work with primary health care teams so they had the opportunity to develop

their own effectiveness measures.

There were two main aims for this work:

• To demonstrate that primary health care teams could develop effectiveness

measures using the ProMES approach

• To demonstrate that primary health care teams could use the measures

developed to get feedback on their performance.

Qualitative work using ProMES was carried out with two PHCTs (see Chapter 2).

There were four distinct stages to the work:

1. Establishing a 'design team', these were representatives from the team who were

primarily responsible for developing the measurement and feedback system.

2. Reaching agreement that the objectives developed in the constituency

workshops, were relevant and related to all the main activities of the

organisation/team.

3. Develop measures that could be used to assess the extent to which these

objectives are being achieved.

4. Using measures to gather information about how well the team was performing.

The researchers worked with the primary health care design teams over a period of

eighteen months, meeting for one hour once a fortnight. As a result of this work the

primary health care teams successfully developed effectiveness measures that they

could use to assess performance on all of the objectives for primary health care. A

major issue in primary health care is the considerable work pressures and demands

made on all members of the team. This is a major constraint on the time team

members have available to engage in activities which do not directly contribute to the

delivery primary health care team services. It was therefore a significant

achievement that teams were able to develop measures, and demonstrates what can

be achieved as result of a relatively small investment of time.

The measures developed by the teams are listed below. Information on how to use

the measures is provided in Appendix III.

Objective 1: Promote, maintain and improve health

Measure 1 - Review of quality in case management

Percentage of cases judged to be managed appropriately on the most relevant

quality dimensions.

Measure 2 - Young People’s Sexual Health

Percentage unwanted teenage pregnancies in a 6-month period

Percentage of teenagers prescribed the morning after pill in a 6-month period

Percentage of teenagers requesting pregnancy tests in a 6-month period

Measure 3 - Young People’s Health - Alcohol and Drug Misuse

Number of teenagers attending A & E after drug overdose in a 3-month period

Number of teenagers attending A & E after excessive alcohol consumption in a 3-

month period.

Measure 4 - Patient access to consultations with a GP

The number of days that patients wait to see a GP of their choice

Measure 5 - Patient access to a quality consultation with GPs

Percentage of patients whose appointment with a GP is minutes duration in a

3-month period.

Measure 6 - Use of out of hours services by patient

Percentage reduction in the use of private out of hours services by patients in a 6-month

period.

Measure 7 - Patients have access to an appropriate health professional

Percentage of patients, in a 6 month period, who have contact with a

health professional from the team at a time and location most appropriate

to them and to the professional.

Measure 8 - Patients have access to a home visit from an appropriate health

professional.

Percentage of patients in a 6 month period who have a home visit from the

health professional judged by the patient and the health professional to be

most appropriate.

Objective 2: Enable personal and community responsibility for individualhealth

Measure 9 - Patients understand the role and function of the PHCT.

Number of patient requests, use health professionals’ time and PHCT services which

are inappropriate in a 3 month period.

Objective 3: Efficient Use of Resources

Measure 10 - Patients able to manage minor illness

Percentage of patients seen by health professionals in the team who had a minor

illness which could have been managed themselves.

Measure 11 - Patients/clients who do not attend for an appointment

Average percentage of total patients' appointments not kept in a week

(calculated over a 3-month period).

Measure 12 - Efficient use of administrative systems

Percentage of patients not attending appointments with health professionals in the team which

result from errors in the administrative system.

Measure 13 - Efficient use of GP resources in the team

Average number of patients seen by a GPs in a week

Objective 4: Continuous personal and professional development

Measure 14 - Team member access to training

Percentage of who are satisfied with the extent to which their training needs are assessed and

met in the previous year.

Objective 5: High team member commitment, stress and satisfaction

Measure 15 - Team member commitment and satisfaction

Percentage of staff in the team who feel committed and satisfied

Measure 16 - Team members use each other's skills, knowledge and expertise

appropriately

Percentage of team members who report that skills,

knowledge and expertise within the team are used

appropriately in 3-month period.

Measure 17 - Effective team working

Percentage of requests for help and information and referrals from other team

members which are inappropriate in a 3 month period.

Objective 6: Responsiveness to client and community

Measure 18 - Patients’ Experiences of the PHCT service (1)

Percentage of patients who report that their experiences of the PHCT services match

the range and standard agreed by the PHCT.

Measure 19 - Patients’ experiences of the PHCT services (2) (Using the existing

measure)

Percentage of patients whose experiences of the PHCT services meet the standard

set by the team.

Measuring Performance

The second aim of the qualitative work with primary health care teams was to

demonstrate that it was possible for primary health care teams to use effectiveness

measures to obtain feedback on performance. Both of the primary health care teams

were able to used effectiveness measures to gather feedback information. Below we

detail the procedure used by one of the teams to develop a measure of patient

satisfaction, gather feedback from patients and then make changes on the basis of

this feedback

Measure 19 = Percentage of patients who report that their experiences of the

PHCT services match the standard agreed by the PHCT.

The measure was developed by the design team as follows.

The team listed all of the services they provided (e.g. consultations with a health care

professional, clinics, district nursing and health visiting services) and also considered

features of the delivery of services which they believed would be associated with

patients satisfaction (e.g. short waiting times, prompt repeat prescriptions, phone

answered quickly, access to advice).

A questionnaire was developed which enabled patients to report their experiences of

the services and the features associated with satisfaction (see Appendix III, p…..).

The areas covered by the questionnaire were: waiting time to see a GP; waiting time

for the phone to be answered; waiting time for a repeat prescription; waiting time to

see a practice nurse; awareness of health visitor services, waiting times at health

visitor clinics; and waiting time for district nurse visits. Patients were also asked to

provide comments on how different services they had experienced could be

improved.

Patients were asked factual questions about their experiences, not for opinions.

For example:

The last time you wanted an appointment with any of the GPs, how soon did you get

one?

Same day [ ]

Next day [ ]

After 2 days [ ]

Longer_________

The last time you asked for a repeat prescription, how long did you have to wait to

get it?

1st time ___________ days. Not Applicable [ ]

2nd time___________ days

The team identified additional patient information that would help to understand the

information collected on patients’ experiences (age, gender, number of visits to the

surgery in the previous month).

Before distributing the questionnaires the team determined the standards they

wanted to achieve. For each question they decided what would be acceptable and

unacceptable responses, and the standard they would like to achieve. They

determined the percentage of patients they would expect to experience the service in

a particular way, the percentage that was unacceptable, and the percentage that

would be an ideal. For example, the team decided what percentage of patient they

would expect to see a GP on the same day, the next day, after 2 days, or after a

longer period of time, the percentages for each which was unacceptable and the

percentages that they would like to achieve.

Over a one week period all patients (or for children, their carers) attending the

surgery were asked to complete a questionnaire. 100 questionnaires were sent to

home addresses, and an additional 100 distributed via district nurses and health

visitors. The information from patients was collated and a mean score calculated for

each item on the survey.

The score for each item was then compared with the expected standard,

unacceptable standard and ideal standard, and the differences between the actual

mean and these percentages calculated. This provided the team with feedback

about the extent to which the experiences of patients matched the standards the

team were trying to achieve, where experiences fell below standards, and where they

were achieving the ideal standard.

Results from the patient satisfaction survey

Responses were received from 320 patients which provided a valuable source of

feedback on the services provided by the primary health care team. On many

aspects the reported experiences of patients matched or exceed that of the

standards set by the team. Where the reported experiences fell below the team's

standards the reasons for this were explored by the design team and changes made

to the provision of this service. For example, the survey revealed that 50% of

patients had waited for between 10 and 15 minutes to see the practice nurse after

their appointment with the GP. This was below the target set, 90% of patients

waiting a maximum of 5 minutes.

Two main reasons were identified for the longer waiting time: patients were not clear

about the procedure for seeing the practice nurse after their GP consultation; and

there were insufficient consultation rooms to accommodate the patients who needed

to see these nurses. Two changes were proposed to reduce the waiting time.

Firstly, an information slip explaining the procedure for seeing the practice nurse was

produced which GPs could give to patients when they referred them to the nurse.

Secondly, the use of consultation rooms was reviewed. An antenatal clinic, which

used two consulting rooms, was held at the same time as the morning surgery. It was

proposed that this clinic was run at a different time thus providing two additional

rooms the practice nurses could use for patient consultations.

Discussion

The aims of the qualitative research were to develop agreement among primary

health care professional about the objectives for primary health care, to develop

measures that would provide feedback on the extent to which effectiveness was

being achieved, and to demonstrate that primary health care teams could develop

and use effectiveness measures.

The national workshops brought together a wide range of primary care stakeholders;

representatives from district nursing, health visiting, general practice, practice

nursing, midwifery, mental health, professions allied to medicine, social services,

health authorities, the Department of Health, NHS Executive, NHS trusts, patient

actions groups and academia. During the course of the workshops these

stakeholders, who had differing aims, objectives, priorities and philosophies of care

were able to reach agreement about the objectives for primary health care. The

practitioners at the workshops judged these objectives relevant and useful, as did

members of primary health care teams who were consulted during the development

process.

Using the objectives developed in the workshops, the ProMES approach was used

with two primary health care teams to develop measures. This stage of the research

also involved working with multidisciplinary groups of health professionals who

developed a set of effectiveness measures that could be used to assess the team's

performance. These measures were used by the teams to get feedback on how

effectively they functioned, and, as illustrated in the example discussed above, this

feedback was used to improve the quality of care to patients and to use the

resources available to the team more efficiently.

The second stage of the research demonstrated that it is possible for primary health

care teams to develop and use effectiveness measures. Further work is required to

improve and refine the measures developed by the primary health care teams, and to

test their generalisability for primary health care teams in a range of settings.

The qualitative research has demonstrated that a combination of the constituency

approach and ProMES provides a practical method that can be used to help primary

health care teams clarify their objectives and to obtain feedback on the effectiveness

of the services provided. This will help health professionals to prioritise resources

and to deliver high quality, cost-effective health care.

Chapter 5

Community Mental Health TeamsResearch Methods and Sample Details

Introduction

NHS secondary mental health care is delivered primarily through multidisciplinary

community mental health teams (CMHTs). These face many challenges. They are

tasked with complex statutory and professional responsibilities (Peck & Parker,

1998). The demands of a primary care-led NHS often conflict with the policy

imperatives of the sensitive area of risk management relating to severe mental health

problems (Onyett, 1995). In addition, the voice of service users gains strength,

adding to workload and pressures. Team members are employed within two very

different bureaucracies; those of health and social care, and come from diverse

professional backgrounds. However, the development of joint commissioning

approaches between health and local authority social services requires them to

function as integrated teams (Hannegan, 1999). Their constituent professions may

jibe at the adjustments this requires (Mistral & Velleman, 1997), for which their

training may not prepare them well.

The current policy agenda is increasingly outcomes-focussed. Accordingly, CMHTs

are required to monitor their performance (Bhugra, Bridges, & Thompson, 1995) and

effectiveness, as a strong commitment to monitoring and evaluation is considered

essential for adequate management of CMHT services (Carter Evans, Crosby,

Prendeergast & De Sousa Butterworth, 1997). The competition for resources

amongst elements of health and social care provision requires that each provide data

to demonstrate the value of its contribution. More positively, effectiveness measures

may also bring some clarity to teams' efforts to chart their own progress towards

meeting diverse expectations.

The organisation of CMHTs is central to their functioning (Bhugra, et. al., 1995;

Onyett, 1997). Their core rationale is to bring together a range of professions in

order to deliver more effective care co-ordination than could be achieved without an

integrated, multidisciplinary team. Achieving that integration is by definition an

organisational task (Onyett, 1995; Pincu, Zarin & West, 1996), requiring that the team

be more than the sum of its diverse constituent members acting individually. This

task is rendered considerably more challenging by the need for multi-agency working

across the health-social services divide (Department of Health, 1995).

User and carer perspectives are increasingly important. The National Health Service

Patients' Charter for Mental Health Services (Department of Health, 1997) sets out

rights and expected standards of service for users and potential users of these

services. It aims to ensure that the NHS 'listens and acts upon people's views and

needs'. A continuing push for users and carers to be involved in decisions relating

to mental health care (Faulkner, 1997), and also to be included at the level of

planning and developing services, presents a further challenge to teams which

deliver integrated care within the Care Programme Approach (Department of Health,

1990).

There is a growing international literature on CMHTs as a mode of delivery of mental

health care. Within the UK, a notable source of this has been the Sainsbury Centre

for Mental Health. The key issues with emerge from the research literature include

the following, for each of which a representative citation is provided:

• The many managerial, professional and clinical barriers to effective

multidisciplinary teamwork (Peck & Norman, 1999).

• The importance of integrated operational management of CMHTs (Onyett,

1997).

• Leadership, integration and agency as key precursors of effectiveness

(Grusky, 1995).

• The threats to effectiveness arising when resource constraints lead teams to

over-emphasise control and efficiency at the expense of creative thinking and

innovation (Drolen, 1990).

• The mismatch between current training arrangements and current and future

service needs (Sainsbury Centre for Mental Health, 1997).

• The specific leadership skills required by CMHTs, in which training is

necessary (Reed, 1995; Sluyter, 1995.

• The highly demanding nature of CMHT work (Prosser, Johnson, Kuipers,

Szmukler, Bebbington & Thornicroft, 1996).

• Detriments to morale and effectiveness from excessive workload (King,

LeBas & Spooner, 2000).

As the Health Team Effectiveness research programme was nearing completion, the

National Service Framework for Mental Health (NSF; Department of Health, 1999)

was published. This seeks to establish national standards for mental health care.

Within the NSF, national support for local action includes workforce planning,

education and training. This aims to enable mental health services to ensure that

their workforce is sufficient and skilled, well led and supported, to deliver high quality

mental health care. A Workforce Action Team (WAT) has been established to

provide national leadership in developing and taking forward the workforce action

plan. We have identified within the WAT interim report (dated April 2000) several

themes that a study of mental health team-working can usefully address:

• Education and training: What are the training requirements for effective

teamworking and how might these be met?

• Recruitment and retention: What are the salient features of team

composition? What factors are associated with staff turnover, and how might

retention be improved?

• Leadership: How does this impact on quality of care? How can it be best

developed?

• Primary care: What characteristics of primary health care teams are

conducive to high-quality mental health care?

• Professionally non-affiliated staff: What can the contributions of support

workers tell us about the potential for further development of non-affiliated

staff?

Before offering answers to some of these questions, we describe the methods used

in our research.

The research with Community Mental Health Teams (CMHTs) had two stages. The

first was a questionnaire survey, and related data collection methods, involving large

numbers of teams to gather data on team inputs, processes and outputs. The

second stage involved intensive examination of a sub-set of teams to explore in more

depth targeted team processes and outputs. An overviews of the methodology for

stage 1 is given in Figure 5.1.

Figure 5.1: Details of CMHT research methods stage 1

Additional questionnaires/

Sample size Survey data Telephoneinterviews

Externalratings

113 teams

1443respondents

Team compositionTeam functioningTeam effectivenessTeam innovationStress

Team compositionTeam meetingsTeam managementDecision makingClinical systemsmanagement

Teameffectiveness

Teaminnovation

Summary of Research Methods

A. National workshop to derive CMHT effectiveness criteria

B. 113 Community Mental Health Teams

§ Survey of all team members

§ Questionnaires or telephone interviews with team leaders

§ External ratings of team effectiveness

§ External ratings of innovations introduced by the teams

C. 10 Community Mental Health Teams

§ Videotaping and analysis of team meetings

§ Caseload analysis and client selection

§ Interviews with practitioners on two occasions, 6 months apart

§ Use of HoNOS to record client outcomes

§ User and Carer Service Satisfaction Questionnaires

Quantitative methods

The Sample

The research design required data to be gathered from 100 CMHTs. Initially, chief

executives of 101 community mental health trusts in 4 regions, Northern and

Yorkshire, North West, Trent, and North Thames, were approached, to inform them

of the study and to encourage participation of all CMHTs managed by that trust. The

aim was to limit the geographical spread while accessing representative CMHT’s, in

terms of different socio-economic locations, skill mix and client base. Three months

after the first mailshot, follow-up letters were sent to all trusts not responding. Of the

101 approached, 81 responded: 11 had no community adult mental health services;

12 declined to participate and the remaining 58 provided names and contacts for all

CMHTs managed. The main reasons for not participating were either that (a)

caseloads were such that teams were too busy (3 trusts); or (b) the Trust was in the

process of reorganisation (7 trusts); or (c) the teams were already taking part in other

research (2 trusts).

With the CMHT names provided we made direct contact with 162 CMHT’s, inviting

participation in the study after consensus to participate had been achieved within

each team. The final number of participating teams was 113 from 45 trusts. Details

of the sample are shown in Table 1. At different stages of the access procedure, it

was open to Trusts or CMHTs to refuse to participate; the sample was therefore

made up of volunteering CMHTs. We performed a post hoc check on socio-

economic representativeness, which indicated that the whole range of deprivation

scores was represented (Mental Illness Needs Index (MINI) range 91.3 (low need) to

118.5 (high need), mean 103.3).

Data collection Methods

Data on team functioning and effectiveness were collected using three methods: self

report questionnaires completed by individual team members; self report or

telephone interviews with team leaders; and external ratings from community health

care representatives, social services and health authority staff.

The named contact for each of the 113 participating CMHTs provided a

comprehensive list of all team members, which included all personnel attending

regular team meetings. Survey questionnaires were sent to 1925 named individuals,

with returns from 1450 (75%). The return rates for professional groups were:

administrative staff 57%; community psychiatric nurses 82%; occupational therapists

83%; psychiatrists 55%; clinical psychologists 90%; social workers 53%; and support

workers 68%. Overall, 925 women (64%) were included in the sample, and the

mean age was 40 (SD 8.37).

Team leaders from 91 CMHTs provided information on team context, team

composition, team processes, and clinical management in a self-report questionnaire

survey.

Questionnaires completed by individual team members

This questionnaire was in four sections (the CMHT survey is included in Appendix I).

Section 1: Team working

This contained six measures of team working. Four of these were drawn from the

Team Climate Inventory (Anderson and West, 1994; 1998) that is based on a well-

developed theoretical model of team functioning (West, 1990). The four measures

assess levels of

§ team participation

§ clarity and commitment to team objectives

§ emphasis on quality

§ support for innovation

Two other measures were included:

• reflexivity, the extent to which team members reflect upon their team objectives,

strategies and processes and make changes accordingly (West, 1996; Swift &

West, 2000)

§ team innovation, the extent to which the team has introduced innovations in

objectives, work strategies, processes and relationships

Respondents were also asked to describe the major changes or innovations

introduced by the team in their work in the previous 12 months.

Section 2: Effectiveness

These included 27 measures of community mental health team effectiveness derived

at a stakeholder workshop (Rees, Stride, Shapiro, Richards & Borrill, in press;

Richards & Rees, 1998). Three underlying dimensions were evident:

§ team working

§ patient/client orientation

§ organisational efficiency

Section 3: Team member stress

This included a measure of psychological stress, the GHQ-12 (Goldberg, 1972). The

GHQ-12 is widely used as a screening tool for detecting minor psychiatric disorder in

the general population, and in occupational mental health research. It covers

feelings of strain, depression, inability to cope, anxiety based on insomnia, lack of

confidence and other psychological problems. Within a Department of Health funded

study of the mental health of the NHS workforce, the GHQ-12 showed good validity

against a psychiatric interview (Hardy, Shapiro, Haynes & Rick, 1999).

Section 4: Biographical information

This section included questions on biographical and team characteristics (e.g. age,

gender, ethnic origin, job title, employer, team composition, team leadership).

Additional Team Information – Survey

This was completed by the team leader or co-ordinator in 92 CMHTs, and combined

the PHCT data collection via survey and telephone interviews. Besides the

information on decision-making and communication systems, data were collected

relating to the clinical systems the CMHT implemented for dealing with referrals, both

emergency and routine, and for accessing inpatient beds.

External ratings – team effectiveness

Each of the 113 CMHTs in the survey sample was approached to nominate three

professionals external to the team, within the local Trust, Social Services, Health

Authority, or GP practices in their catchment area, in order to collect corroborative

data on team effectiveness. Thirty-three teams nominated up to 4 external judges

each. Judges’ ratings were made using the same 27 effectiveness dimensions that

team members had used to rate their team’s effectiveness.

External ratings – team innovation

Two experts known to the research team rated the descriptions of changes or

innovations introduced in each CMHT over the previous 12 months, and which team

members had described in their questionnaire responses. The changes were rated

on the following dimensions (West & Anderson, 1996):

§ magnitude, how great would be the consequences of changes introduced

§ radicalness, to what extent the status quo would change

§ novelty, how new in general were the changes

§ impact, to what extent changes would improve CMHT effectiveness

Sample Details

The aim of the research programme is to determine which team characteristics are

associated with good team functioning and team effectiveness. In this section, we

describe characteristics of the CMHT sample that participated in the survey

components of the research programme. Following the model used to guide this

research see Chapter 1) we describe four categories of inputs:

§ team characteristics – indicated by size, mean age, mean tenure, gender mix,

ratio of full-time members, length of time the team had been in existence.

§ team composition – indicated by the distribution of occupational groups within the

team.

§ team task – indicated by the MINI (high or low deprivation scores), how quickly

the CMHT saw emergency referrals, pooling of referrals, the use of a single

integrated set of client case notes, whether waiting lists were in operation for

client assessment.

§ team environment – indicated by how the CMHT was commissioned and the

English NHS region within which the team was located.

We also describe the relationships between these four domains of inputs. The

reader will also find an account of team members’ perceptions of leadership in the

team.

Team Characteristics

Number of team members

In terms of the number of individuals employed within each team, this ranged from 6

to 51. The mean size was 17.04 members, SD 7.99. Distribution of sample team

size is shown in Figure 5.2 below.

Age and gender distribution

Overall, 925 women (67%) were included in the sample. Figure 5.3 shows the

percentage of women in CMHTs. The mean age was 40, SD 8.37. Across the 113

CMHTs, only one CMHT was made up of only women. The age distribution appears

normal, but it is noteworthy that there are very few CMHT workers below 30 or above

50 years old. This age profile resembles that of qualified nurses.

Figure 5.2: Distribution of team size (number of teammembers) across the sample

0

5

10

15

20

25

30

35

up to 10 11 to 15 16 to 20 21 to 25 > 25

Figure 5.3: Percentage of women in CMHTs in thesample

0

5

10

15

20

25

30

35

80 to100%

70 to79%

60 to69%

50 to59%

25 to49%

% of teams

Tenure

All team members indicated how long they had been in the CMHT. Mean tenure

across teams was 37 months, SD 19 months. This relates to the short life of one

third of teams in the sample, which had been in existence for less than 2 years.

Whole time equivalents

Number of Teams

% of women

Number ofmembers

Taking account of hours worked to calculate team size in terms of whole time

equivalents shows that team size varied from 5.5 to 48.5. The mean size was 15.81

WTEs, SD 7.53. The small size of the mean difference between numbers of

members and WTEs suggests that the great majority of team members were

employed on a full-time basis. Twelve per cent of CMHTs in the sample were

comprised solely of full-time workers. Across the sample, mean percentage of full-

time workers was 77.49, SD 13.74.

Length of time CMHT in existence

CMHTs were formally introduced on a national basis in 1990, to provide integrated

care in the community for mentally ill people. In this sample, the length of time the

teams had been existence varied from 6 months to 7 years. For subsequent

analysis, these were categorised as less than 2 years (n = 31); from 2 to 5 years

(n = 36); and 5 or more years (n = 25).

Team composition

Figure 5.4 gives the breakdown of the sample by professional group and by gender

(n = 1363).

Figure 5.4: CMHT occupational groups by gender

WOMEN MEN

ADMIN 181 7

CPN 316 214

OT 92 14

PSYCHIATRY 27 55

COUNSELLING/PSYCHOLOGY 43 17

SOCIAL WORK 129 83

SUPPORT WORK 68 22

OTHER 62 32

TOTAL 918 444

As expected, the largest occupational group was nursing, making up 39% of the

sample. The next largest groups were social work (16%) and administrative staff

(14%). Occupational therapy (8%), support work (7%), psychiatry (6%) and

psychology/counselling (4%) were the smaller occupational groups. As compared

with the overall preponderance of women, who formed two-thirds of the respondents,

administrative staff were, unsurprisingly, even more predominantly female. Almost

90% of occupational therapists were women. In contrast, two-thirds of psychiatrists

were men. The gender mix of nursing and social work showed a modest

preponderance of women.

At the team level, multidisciplinary mix was as shown in Figure 5.5. Noteworthy here

is the fact that just 12% of teams included members from all disciplines (psychiatry,

social work, psychiatric nursing, clinical psychology, occupational therapy).

Figure 5.5: Multidisciplinary mix within CMHTs

0 5 10 15 20 25 30 35

Number of CMHTs

CPN plus one other Health

CPN Psychiatry and SW

CPN plus 2 other Health

CPN plus SW

All disciplines, no SW

All disciplines except psychology/OT

All disciplines except psychiatry

All disciplines

Team Task

Mental Illness Needs Index (MINI)

MINI scores for the Health Authority areas within which CMHTs’ populations were

based ranged from 91.3 (low need) to 118.5 (high deprivation), mean 103.34, SD

6.91. Figure 5.6 shows the distribution of CMHTs across Health Authorities with low

(30%), medium (40%) and high (30%) deprivation scores.

Figure 5.6: Teams in areas of high of high, medium and low need as indicatedby the MINI

low need

medium need

high need

CMHT response to emergency referrals

CMHTs indicated how quickly, on average, emergency referrals were seen. Elapsed

time before emergencies were seen ranged from ‘within the hour’ to ‘within two

weeks’. Across all teams, the mean wait for emergencies was 26 hours, SD = 44

hours.

Within-CMHT pooling of referrals

CMHTs provided information on how referrals, other than emergencies, were dealt

with. Five CMHTs did not pool referrals; 20 CMHTs pooled some referrals; and 63

CMHTs pooled all referrals.

Use of integrated case notes

In 40 CMHTs, each discipline kept client case notes separately; in 12 CMHTs, notes

were separate but available for reference by other disciplines; and in 34 CMHTs,

each client had one integrated set of case notes.

Assessment waiting list implementation

Fifty-three CMHTs indicated that they did not operate a waiting list prior to

assessment, and 36 CMHTs indicated that they did operate a waiting list.

CMHT Organisational context

NHS Region

CMHTs were sampled from four NHS regions in England. The participating regions

contributed 32, 26, 32 and 23 respectively. To safeguard the anonymity of the

participating teams, these regions are not identified in this report.

Local commissioning arrangements

There were three models of commissioning for these teams: 25 CMHTs were

commissioned by Health Services only; 39 by Health and Social Services jointly; and

33 by Health and Social Services separately. CMHTs are constituted on a multi-

agency basis between health and social services. Most CMHTs necessarily combine

staff working within the management structures of each of the two agencies, tasked

with meeting the objectives of both. However, each CMHT must function as a

coherent entity working towards mutually agreed objectives and following mutually

understood and functionally interdependent practices. Local commissioning

arrangements may impact on integrated team functioning.

CMHT constructed process and outcome variables

Apart from scales derived from survey items, for example, those from the Team

Processes Inventory, CMHTECQ and GHQ, others were computed to measure the

clarity of team leadership, within-team variation in relations to the clarity of team

leadership, the team’s efforts to communicate other than in a formal meeting

environment, and the team’s turnover.

Clarity of CMHT leadership

All team members provided information about the clarity of leadership in the CMHT.

Team means were aggregated from the single survey item: ‘ Does the team have a

single clear leader or co-ordinator?’ where ‘yes’ scored 1 and no entry was

scored ‘0’. Members of 13 CMHTs in the sample were unanimous in reporting that

their team had a single clear leader, while members of six CMHTs were unanimous

in declaring that the team had no single clear leader or co-ordinator. The aggregated

measure was treated as a process variable.

Within-team variation in relation to the clarity of team leadership

Blau’s index of variation was used to calculate the extent of disagreement within the

team about the clarity of leadership. Five variables with values ranging from 0 to 1

were constructed for the proportion of each team giving each of the five possible

responses. The resulting variable was treated as a process variable.

Internal Communication

In the Additional Team Information survey, team leaders were asked two

questions to indicate (a) how much team members had access to information

other than that conveyed in meetings, for example, with the use of memos,

whiteboards, newsletters; and (b) how much social activity team members

participated in together. Responses were combined to give a measure of the

CMHT’s intent to communicate both informally and socially, on a scale of 1

(poor) to 5 (high quality effort to communicate). Figure 5.7 shows how teams

varied on this dimension, which was treated as a team process variable.

Figure 5.7: CMHTs' Intent to Communicate

interdisciplinary communication

4 and above3 to 42 to 31 to 2

num

ber

of te

ams

50

40

30

20

10

0

Turnover

In the Additional Team Information survey, team leaders were asked to indicate how

many staff had left the team in the previous 12 months. Turnover was computed as

the percentage of staff in the team (size) who had left, and was treated as an

outcome variable.

Relationships between ‘input’ dimensions

Team size

Figure 5.8: Scatterplot of CMHT climate against team size

team size going by team members declared

6050403020100

self

repo

rt: c

ombi

ned

TC

I sca

les

4.5

4.0

3.5

3.0

2.5

2.0

Three CMHTs were larger than all others, and scatterplots (e.g. Figure 5.8)showed that on some dimensions of team functioning these were outliners.They were removed from all analysis involving team size. Team size wasclearly associated with the diversity of disciplines within the team: largerteams were more likely to have psychiatrists (r = .20, p < .05) andpsychologists (r = .25, p < .01). Larger teams also had a lower percentageof full-time staff (r = -.22, p < .05). Team size was associated with only oneaspect of the team’s task or organisational environment: larger CMHTs weremore likely to operate a waiting list for assessment,r = .29, p < .01.

Team average age

The average age of the CMHT was likely to be greater if social workers (r =.30, p < .01) and psychologists (r = .25, p < .01) were members. However,it was likely to be lower if the MINI score of deprivation was high (r = -.36, p< .001).

Team average tenure

Average job tenure was lower if there were social workers in the team (r = -.20, p <

.05). Predictably, average job tenure was strongly associated with the length of time

the team had been in existence (r = .51, p < .001).

Length of time CMHT in existence

Beyond the relationship with average job tenure in the team, the life of theteam was associated with whether it operated at separate sites (r = -.23, p <.05), that is, the longer the team had been in existence, the less likely was itthat staff were based at different locations.

Percentage of full-time workersWe saw above that larger teams were likely to have a higher percentage ofpart-time workers. Such part-time practitioners were likely to be psychiatrists(r = -.26, p < .01), psychologists (r = -.31, p < .01) and occupationaltherapists (r = -.28, p < .05). Although they described themselves as being‘part-time’, these disciplines may well have divided their time between theCMHT in question and other responsibilities.

Percentage of women in the team

A lower percentage of women in the team was associated with a higherdeprivation rating, R = .21, p < .05.

Further effects of team composition

When psychiatrists were in the team, it was less likely that the team would implement

a single, integrated set of case notes for each client (R = -.32, p < .01), but if an

occupational therapist was in the team, the opposite was the case (R = .31, p < .01).

Again, psychiatrists were more likely to be in the team if the deprivation rating was

high (R = .23, p < .01). If there were social workers in the team, emergencies were

likely to be seen more quickly (R = -.27, p < .05). The team was more likely to

operate a waiting list for assessment if there were social workers (R = .28, p = .01),

psychologists (R = .23, p < .05) or occupational therapists (R = .23, p < .05) in the

team.

Further effects of clinical system implementation

Where the team pooled referrals at a central point, they were also likely to operate a

single, integrated set of case notes for each client (R = .25, p < .05), although a

single referral point was also associated with lower deprivation scores (R = -.29, p <

.01). Where the deprivation score was higher, waiting lists for assessment were less

likely to be in place (R = -.26, p < .05).

Intensive analysis

Selection procedure

Teams rating themselves as highly effective or as highly ineffective, incomparison to the self-ratings of the full sample of 113 teams, were selected.This recruitment strategy was designed to maximise the power of between-team analyses to detect associations between intensive analysis variables andeffectiveness. We computed aggregate team scores on the CMHTteamworking questionnaire scales. These comprised the Team ClimateInventory (TCI) participation, support for innovation, and task orientationscales; the user orientation, use of resources, and internal process scalesfrom the service delivery effectiveness (SDE) items; and the 12-item GHQ.The standardised team scores on the TCI and the SDE were summed, and thestandard team score on the GHQ-12 subtracted from this total. Thisalgorithm yielded 14 teams above the 80th percentile and 14 below the 20thpercentile. To achieve a target sample size of 16 teams in this phase of thestudy, all 28 were invited to participate after the team had gained consensusamongst members. Ten volunteered, distributed across the 4 NHS regions asfollows: A, 3; B, 2; C, 1; and D, 4. MINI scores for the 10 teams covered awide range, 91.3 to 110.0, with a mean of 101.4. Three of the 10 had ratedtheir activity as effective, leaving 7 who rated their team as ineffective. Wefollowed up the 4 teams rated as effective which had not already responded,but failed to increase the number. To a considerable degree, therefore, thiswas a self-selected sample. The 10 teams comprised three self-rated as‘effective’ and seven self-rated as ‘ineffective’.

Representativeness of sub-sample CMHTs

Independent t-tests were used to compare group means on appropriate dimensions,

together with Levene’s test for equality of variances. This process included

comparisons for:

§ Team characteristics and composition: size; age; tenure; percentage of full-time

workers in the team; percentage of men in the team; length of time the team had

been in existence; professional mix.

§ Task environment: MINI; number of GP’s linked to the CMHT; ; whether the team

held a single, integrated set of case notes for each client; whether referrals were

pooled or not; use of a waiting list for emergency referrals.

§ Team processes: team processes scale; perceived clarity of leadership; intent to

communicate informally; amount of meeting time available; perceived quality of

relationship with GP’s, Health Trust, and Social Services.

§ Team effectiveness: levels of stress; self-report effectiveness; external judges’

ratings of performance; external judges’ ratings of team innovations; self-report

innovativeness; turnover.

For the most part, group means were similar, with t values ranging from .07 to 1.65.

Exceptions are shown in Figure 5.9. Levene’s tests for equality of variances did not

reach significance for any dimension.

Figure 5.9: Significant differences between survey sample and sub-sample

Surveysample mean

(SD)

Sub-sample mean(SD)

T value probability

Percentage offull-time staff

78.48 (13.27) 67.28 (15.09) 2.52 .01

Average tenureof staff (months)

38.66 (18.89) 24.62 (20.39) 2.23 .03

Psychologist inthe CMHT

0.34 (0.48) 0.80 (0.42) 3.26 .007

CMHT stresslevel (GHQ)

0.96 (0.16) 1.09 (0.20) 2.20 .03

External ratingsof CMHTeffectiveness*

3.59 (0.38) 3.25 (0.47) 2.07 .05

* The Ns on this dimension were 25 in the survey group, 8 in the sub-sample group

These comparisons showed that, in relation to most team characteristics, and most

aspects of task environment, team process and effectiveness, the sub-sample group

did appear to be reasonably representative of the full survey sample. However, sub-

sample teams had a lower percentage of full-time staff, their members were likely to

have been in the team for a shorter time, were more likely to have a psychologist or

counsellor in the team, their members experienced a higher level of stress, and

teams were rated lower on effectiveness by external judges. This last finding is not

surprising, given that more self-ratedly ineffective than effective teams had

volunteered to take part in this phase of the research. Chi-square tests showed that

the sub-sample teams were representatively distributed throughout the 4 NHS

regions, but were not representatively distributed in terms of local commissioning

arrangements, chi square = 6.33, p < .05, with 7 teams being jointly commissioned

by Health and Social Services, and 3 by the Health Service only.

Descriptives on the 10 CMHTs in the intensive phase

Figure 5.10 shows the characteristics of the ten teams. All had CPNs, all were

multidisciplinary, although only 3 of the 10 teams had access to psychiatry input

within the team.

Figure 5.10: Team characteristics of sub-sample CMHTs

TeamA

TeamB

TeamC

TeamD

TeamE

TeamF

TeamG

TeamH

Team I Team J

Size 16 12 15 15 22 12 18 17 46 12

Mean age 44.9(11.70)

37.1(7.43)

34.6(7.94)

41.2(7.22)

38.2(7.44)

40.4(8.05)

45.2(6.45)

41.9(8.02)

43.3(9.85)

41.0(8.43)

Mean tenure(months) 13.58 23.56 21.07 35.73 10.55 12.00 9.80 3.85 47.11 69.00Length of life ofCMHT

Less than2 years

2 to 5years

2 to 5years

Over 5years

Lessthan

2years

Lessthan

2years

Lessthan

2years

Lessthan

2years

2 to 5years

Over5 years

Psychiatry in No no no Yes no no No No yes yesOT in Yes yes yes Yes yes no Yes No yes yesPsychology in Yes yes yes Yes no no Yes Yes yes yesSW in Yes no no Yes yes yes Yes Yes yes yes% men 33 38 15 27 36 62 27 23 37 42% full-time 83 56 54 55 64 100 60 77 67 58MINI 97.4 110.0 110.0 91.3 96.6 94.8 97.4 102.6 - 104.2Linked GP’s 34 38 26 30 - - 24 35 120 35Commissioning H & S

jointlyH

onlyH

onlyH & Sjointly

H & Sjointly

H & Sjointly

H & Sjointly

H & Sjointly

H & Sjointly

H only

NHS region B A A B C B B C D A

It is apparent that there were differences between the teams on all structuraldimensions. For example, team 1 stood out as a large team compared withothers, and teams B and C were in Health Authorities which had a higherMINI score than others, indicating localities with higher deprivation. All teamshad a mix of disciplines, although team F had only social work input, over andabove the CPNs who were present in all teams. Percentages of male and full-time workers in these teams varied widely. Half of the teams had been inexistence for less than 2 years, and this partly accounted for shorter tenure ofstaff, although there is commonly high turnover of staff in the CMHTenvironment.

Audio and video recording of Community Mental Health Team meetings

The ten teams volunteering for the intensive stage research also agreed that their

meetings could be recorded; meeting sizes ranged from six to twenty five members.

The meetings were multi-disciplinary business meetings in which decisions were

made about the running of the team. This meeting was chosen by the teams as the

one to record because it is the main forum, outside clinical meetings, that are multi-

disciplinary meetings. Wherever possible, we recorded two meetings of the same

type for each team. Dates for meeting recording were at the discretion of the CMHT.

For details of recording procedures, equipment and transcription see Chapter 2.

Longitudinal data collection: clinical outcomes; use of resources; patient andcarer satisfaction

Two site visits six months apart were arranged with each participating CMHT. Before

the first, every practitioner in the team completed a caseload audit summary to define

as precisely as possible the population the CMHT was serving (Manchester Audit

Tool, recommended by the Kings Fund mental health team). This required a

breakdown by 19 classifications of the care worker’s entire current caseload in terms

of diagnosis, severity, complexity and chronicity. Simple guidelines for completing

the audit questionnaire were included. For the purposes of patient selection, an

individual breakdown was also completed by each practitioner, using either codes or

names. Stratified sampling was carried out by the research team using SPSS

random number generation. Stratification ensured that users selected were

representative of (a) the individual practitioner’s caseload; and (b) the team’s entire

caseload profile. Stratification was based on scores provided by practitioners for

each client for severity, chronicity and complexity. Scores were then summed. For

each of the 10 teams, 40 users were selected.

If practitioners indicated that clients were unable to complete a questionnaire,

because they were in an acute episode, or could not read, or did not read English

well, they were replaced by others on the key worker’s caseload matched for

stratification. Packs of information sheets, consent forms and return envelopes were

sent to each key worker for each of their selected clients (and carers where

appropriate). Practitioners were requested to discuss the research with clients and

hand them information and consent form, and for the purposes of confidentiality, to

ask them to send signed consent direct to the research team. Once the research

team had received signed consent from client or carer, Patient/Carer Service

Satisfaction questionnaires were sent out, to assess the effectiveness and

acceptability of the service received, and users’ quality of life. The questionnaire was

that developed by the Tameside & Glossop Rehabilitation team, who gave

permission for its use.

First site visit

Two researchers visited each team for a day to interview all practitioners with a

caseload. At the interview, practitioners answered questions about each of their

selected clients individually. Biographical data were collected: gender, age ethnic

origin, type of housing, and whether they had a formal carer. Questions relating to a

period of the previous 6 months covered use of resources, for example, types and

dosage of psychoactive medication, day hospital attendance, number of contacts

with CMHT professionals, time spent as an in-patient. Other questions covered

referral, diagnosis, CPA level, care plan targets, whether or not the client was on

section, and projected clinical outcomes for 6 months ahead. Practitioners were

provided with 6 HoNOS forms for each client, to be completed by the practitioner at

each contact over the following 6 months, or up to discharge.

Second site visit, after an interval of 6 months

Self-report key worker schedules were designed, following the model of the

practitioner interviews implemented at the first site visit. The schedules were sent

two weeks in advance of the visit to the site to collect completed schedules and

HoNOS questionnaires, and respond to queries. All team members completed the

survey questionnaire for a second time, so that change over time could be measured.

The costs of use of resources were computed by combining estimates of quantity

and cost per unit. Medication costs were based on the 1997 BNF. Contact costs

were based on Unit Costs of Health and Social Care (Netten & Dennett, 1997).

Elements included in the costs were medication, inpatient days, day hospital

attendance, respite care, day centre use, drop-in use, occupational therapy groups,

outpatient attendance and contacts with CMHT staff.

Data analytic strategy

Psychometric Considerations

The self-report measures of community mental health team process and

effectiveness used in this research were complex and novel. Accordingly, prior to

analysis to address the substantive questions listed above, preliminary psychometric

analysis considered the intercorrelations among these measures at the team level.

Specifically, the extent of specificity vs. redundancy in these measures had not

previously been assessed. To be useful in testing theories relating team processes to

effectiveness, the measures would have to demonstrate sufficient specificity that the

relationships among them not be most parsimoniously explained as reflecting a

single evaluative or morale factor. As shown in Figure 5.11, all intercorrelations were

significant and substantial, ranging from .53 to .89.

Figure 5.11: Team-level correlations among self-report measures of CMHT processand effectiveness

TCI:participation

TCI:support

forinnovation

TCI:clarity of

objectivesreflexivity innovation

CMHTEQ:external

requirements

CMHTEQ:internal

processes

TCI: support for innovation .858TCI: support for innovation .858TCI: clarity of objectives .633 .771TCI: clarity of objectives .633 .771Reflexivity .706 .816 .706Reflexivity .706 .816 .706innovation .531 .743 .662 .732innovation .531 .743 .662 .732SDE: external requirements .642 .701 .652 .566 .526SDE: external requirements .642 .701 .652 .566 .526SDE: internal processes .818 .889 .791 .744 .671 .842SDE: internal processes .818 .889 .791 .744 .671 .842SDE: monitoring/evidence .555 .676 .620 .578 .551 .878 .777SDE: monitoring/evidence .555 .676 .620 .578 .551 .878 .777

All correlations have N = 113, p < .01.

We therefore considered whether the effectiveness of a team as reported by its

members could be distinguished from its climate, also as reported by those same

team members. We entered team means on the four Team Processes Inventory

(TCI) scales and the three SDE scales in a factor analysis. This indicated that a

single factor was the most efficient way to describe the differences among the teams;

attempts to force a two-factor solution did not support a distinction between team

processes and self-reported effectiveness, as the scales with highest loadings on

each factor comprised a mixture of both TCI and SDE scales.

External ratings of effectiveness were available for 33 teams. For this subsample,

we considered whether external ratings of effectiveness were any more highly

correlated with self-reported effectiveness than with the TCI. There was no such

difference. External ratings of effectiveness were as highly correlated with team

members’ ratings of team processes, r = .64, as with their reports of team

effectiveness, r = .60.

The quality of innovations described by the members of all 113 teams was rated by

external judges. These ratings were no more strongly correlated with team

members’ TCI ratings, r = .44, than with their reports of the team’s effectiveness, r =

.39. Considering only the most relevant TCI scale, support for innovation was

correlated at r = .48 (indistinguishable from the overall TCI correlation of .44) with

external ratings of innovation. The fact that the TCI correlates rather more highly, at r

= .64, with external ratings of team effectiveness than with external ratings of

innovations, r = .44, is further evidence against specificity. Meanwhile, self-reported

innovation (considered an effectiveness dimension) correlated very similarly at r =

.55 with external ratings of innovation.

These analyses rather suggest that the self-report measures should be most

parsimoniously considered to reflect a single evaluative or morale factor, rather than

to tap specific aspects of team process or effectiveness. The measures may not,

therefore, be sufficiently precise to reveal subtle relationships between CMHT

process and effectiveness.

Sequencing of Multivariate Analyses

Analysis proceeded in two steps. First, the questions identified above were

addressed in sequence. For each dependent variable in turn, potential predictor

variables were entered in groups. For example, for each team process variable in

turn, team characteristics were entered as one set of predictors, before moving on to

a second analysis looking at team composition factors as predictors, a third analysis

with team task factors as predictors, and a fourth with organisational context factors

as predictors.

Secondly, for each class of dependent variables in turn, variables that had emerged

from the foregoing analyses as showing predictive relationships independent of the

other variables in their respective group were entered into new analyses including all

such variables across the groups. Analyses at this second step identified predictors

that were independent in their effects of other significant predictors across all classes

of predictor. Such predictors warrant closer attention; accordingly, this account of

our findings will emphasise this second phase of the analysis.

In the next chapter we describe the results of these analyses.

Chapter 6

Community Mental Health TeamsResults from Survey and External Ratings

Summary of Findings

• Teams whose members were more positive about team processes are rated

as more effective by external stakeholders

• Teams whose members agree as to how clearly the team leadership role is

defined are rated by external stakeholders as more effective

• Teams whose members describe their team processes positively perceive

their teams as more effective

• Teams whose members report clarity as to the leadership role perceive

their teams as more effective

• Teams that have been in existence for a relatively long time tend to

describe their team as more effective

• Larger CMH teams are rated as more innovative by external judges

• Teams who perceive their performance as highly reflexive are rated as

more innovative by external judges

• Teams who perceive their team processes and reflexive behaviour as

positive also see themselves as more innovative

• Teams who perceive their team processes as poor experience higher

levels of stress

• Teams with older members enjoy more stable membership, as do teams

including social workers

• Teams using integrated client case notes, and teams not including

psychiatrists, are clearer about the leadership role

Introduction

Our analysis was informed by the input-process-outcome model presented in

Chapter 1. This entailed predicting process variables from input variables, and

predicting outcome variables from both input and process variables.

As for the analysis of PHCTs, we addressed two main questions:

• Is there an association between the composition of a community mental

health team and team processes?

• Is there an association between the composition and processes of the

community mental health team and the effectiveness of the team?

The team characteristics, team processes and measures of team effectiveness are

summarised in Figure 6.1.

Figure 6.1: Team Inputs, Processes and Effectiveness Variables

Characteristics Processes Effectiveness

IndividualAgeGenderTime in jobTime in teamTime in team

TeamOccupational groupsSize (number of members)Number of GP'sHours workedMultidisciplinary mixGender mix

Team contextCommissioning typeMINI indexUse of integrated casenotesResponse time foremergenciesWaiting list in operationNHS RegionRelationship with SocialServicesRelationships with GP'sRelationship with Trust

Team processesParticipationInnovationObjectivesEmphasis on qualityReflexivityNumber of meetingsTypes of meetingsFrequency of meetingsPotential time for differentdisciplines to meetDecision makingLeadershipIntegration andcommunication in the group

Team ratingsOrganisationTeam workingPatient focusInnovation

External ratings (innovation)MagnitudeRadicalnessNoveltyImpact

External ratings (effectiveness)OrganisationTeam workingUser/carer focus

Types of InnovationsQuality of CareExternal collaborationResponsibility for healthUse of resourcesProfessional developmentTeam satisfactionResponsivenessStress (GHQ 12)Turnover

Team processes

Individual team members rated team processes on six dimensions: participation;

innovation; team-objectives; emphasis on quality; reflexivity; and interdependence.

Information about decision-making processes, communication, number and types of

meetings, who attended these meetings, and how the team was managed was

collected from Practice Managers. The information on team meetings was

categorised according to who contributed to operational, strategic and clinical

decisions. In addition a new variable ‘integration’ was developed which assessed the

extent to which there were mechanisms within the team to encourage inter

disciplinary communication and working.

Team Inputs

Information about the team members’ ages, gender, ethnicity, professional group,

tenure and team leadership were collected from each team member. Information

was also gathered on team size, hours worked, commissioning type, MINI index,

NHS Region, whether the CMHT made use of a single integrated set of client case

notes, how long the team took to respond to emergency referrals, whether they

operated a waiting list for assessment, and the team’s relationships with GP’s, Trust

and Social Services.

Team Processes

Individual team members rated team processes on six dimensions: participation;

support for innovation; clarity of team objectives; emphasis on quality; reflexivity; and

integration. The variables participation, support for innovation, clarity of team

objectives and emphasis on quality were very highly correlated and were combined

to form one variable describing team processes. Information about decision-making

processes, communication, number and types of meetings, and who attended

meetings was collected from team leaders. An index of the amount of time the

various disciplines in the team could potentially meet was computed. In addition, a

new variable ‘integration’ was developed which assessed the extent to which there

were mechanisms within the team to encourage interdisciplinary communication.

Team Effectiveness

As with PHCTs, this was assessed using information from a variety of sources.

Team members rated their teams’ effectiveness on the three dimensions of the

Community Mental Health Team Effectiveness Questionnaire (CMHTEQ; Rees,

Stride, Shapiro, Richards & Borrill, in press), developed within this project: team

working; organisational efficiency; and patient orientation. Team members also rated

their teams’ innovativeness and described the innovations implemented by the team

in the previous year. External raters assessed the innovations reported by the teams

on four dimensions: magnitude; radicalness; novelty and impact on team

effectiveness. External ratings of team effectiveness on the CMHTEQ were provided

by external experts nominated by the team and based in local GP practices, Social

Services, the Trust or Health Authority. Individual team members also completed the

GHQ-12 (a measure of mental health or psychological stress). The measures of

interest for this report are overall effectiveness of the team in delivery of services,

user-centred care and dealing with the demands of the parent organisation (both

externally rated and self rated), overall innovation (both externally rated and self

rated), team turnover, and mental health measured by the GHQ-12.

Results

The method of analysis was similar to that described for the PHCT data in Chapter 3.

Stepwise multiple regressions were carried out, with possible predictors of each

dependent variable being split into groups according to type of variable, to identify

those which might ultimately predict the dependent variable. As with the PHCT

analysis of Chapter 3, to reduce the complexity of the data set and to guard against

Type 1 errors arising from multiple statistical tests, we focused on a second-level

analysis combining predictors across the groups. Process variables were predicted

by team characteristics or “inputs”; and effectiveness, innovation, turnover and

mental health were conceptualised as “outcomes” predicted by team characteristics

(“inputs”) and by team processes. In a final stage of the analysis, both team

characteristics (inputs) and team processes were considered together as predictors

of the outcome variables (effectiveness, innovation, turnover and mental health).

We found that the CMHT data called for a different approach to the issue of team

size than we adopted for the PHCT data. It transpired that the relationship between

team size and other variables was largely due to 3 outlying teams (with more than 36

members). Accordingly, rather than considering team size first in all analyses as we

had done with the PHCT data, we excluded these teams from analyses including the

size variable, and thereafter treated team size in the same way as other team

characteristics.

Question 1 – Is there an association between the composition of a CMHT and team

processes?

Figure 6.2: Relationships Between Team Composition and Team Processes

Dependent variable Predictor variables β p R2

Consensus on leadership Team size -.250 .017Tenure -.215 .039 .097

Reflexivity Presence ofpsychiatrist(s)

-.225 .018 .042

Integration noneTeam processes nonePotential to meet noneNumber of meetings Tenure .212 .046 .045

Figure 6.2 presents the significant team composition predictors of each team

process variable. Larger teams, and teams whose members had been longer

in their jobs in the team, showed less consensus on the clarity of leadership in

the team. Teams including psychiatrists as members were less reflexive. The

tabled result was obtained after excluding the responses of psychiatrists

themselves, which had inflated the observed relationship. This, our preferred

analysis, is more conservative because it excludes the effect of the tendency

of psychiatrists themselves to rate the team as less reflexive. It may therefore

be interpreted as showing an association, albeit modest, between the

reflexivity ratings of non-psychiatrist team members and the presence of

psychiatrists within the team.

Question 2 – What Affects the Effectiveness and Innovativeness of a Community

Mental Health Team?

Figure 6.3: Relationships Between TeamComposition and Processes, and Ratings ofEffectiveness

Dependent variable Predictor variables β p R2

Overall effectiveness(external)

Team processes .643 <.001 .413

Consensus onleadership

-.311 .046 .487

Overall effectiveness (self-rated)

Team processes .861 <.001 .742

Consensus onleadership

.163 .003 .766

Length of time the teamhad been in existence

.121 .020 .780

As shown in Figure 6.3, two variables predicted, independently of one another, the

ratings of CMHT effectiveness given by their local stakeholders: team processes and

lack of consensus relating to team leadership. In other words, teams whose

members rated their team processes favourably, and teams whose members agreed

amongst themselves as to how clearly the leadership role was defined within the

team, were rated by external stakeholders as more effective. Figures 6.4 and 6.5

below show these predictive relationships graphically.

Figure 6.4: Impact of team processes onexternally-rated team effectiveness(data from 32 CMHTs)

Team Processes

5.04.54.03.53.02.52.0

5.0

4.5

4.0

3.5

3.0

2.5

2.0

Figure 6.5: Impact of clarity of leadership onexternally-rated team effectiveness (data from32 CMHTs)

clarity of team leadership

1.21.0.8.6.4.20.0-.2

exte

rnal

rat

ing

of te

am e

ffect

iven

ess

5.0

4.5

4.0

3.5

3.0

2.5

2.0

Figure 6.3 also presents the three factors predicting self-reported team effectiveness

on the 27-item Community Mental Health Team Effectiveness Questionnaire

(CMHTEQ), each independently of the other two: Overall score on the Team

processes Inventory, as shown in Figure 6.6; clarity of leadership, as shown in Figure

6.7; and the length of time the team had been in existence, as shown in Figure 6.8.

In other words, members of teams whose members described their processes

positively, members of teams whose members reported clarity as to the leadership

role, and members of teams that had been in existence for a relatively long time, all

tended to describe their team as more effective.

Figure 6.6: Impact of team processes on self-reported effectiveness (data from 113 CMHTs)

team processes

4.54.03.53.02.52.0

4.5

4.0

3.5

3.0

2.5

2.0

Figure 6.7: Impact of clarity of leadership on self-reported effectiveness (data from113 CMHTs)

clarity of team leadership

1.21.0.8.6.4.20.0-.2

4.5

4.0

3.5

3.0

2.5

2.0

Figure 6.8: Impact of length of time team inexistence on self-reported effectiveness (datafrom 113 CMHTs)

Length of time team in existence

5 or more years2 to 5 years2 years or less

4.0

3.9

3.8

3.7

3.6

3.5

3.4

3.3

3.2

3.1

3.0

3.5

3.4

3.3

Relationships between Team Composition and Processes, and Ratings ofInnovation

As shown in Figures 6.9 and 6.10, two variables predicted, independently of one

another, the quality of the innovations reported by team members as rated by

independent, expert judges: Reflexivity, beta = .51, t = 6.72, p < .001; and team size,

beta = .38, t = 5.00, p < .001. In other words, teams whose members rated their

teams as highly reflexive, as well as larger teams, described innovations that were

judged to be of higher quality.

Figure 6.9: Impact of reflexivity on expertratings of innovation quality (data from 113CMHTs)

reflexivity

6.05.55.04.54.03.53.02.5

exte

rnal

rat

ing

of te

am in

nova

tions

5

4

3

2

1

0

Figure 6.10: Impact of team size on expertratings of innovation quality (data from 113CMHTs)

team size

403020100

exte

rnal

rat

ing

of te

am in

nova

tion

5

4

3

2

1

0

Three factors, acting independently of one another, were associated with the

innovativeness reported by team members: self-reported reflexivity, beta = .45, t =

3.75, p < .001, as shown in Figure 6.11; overall score on the Team Processes

Inventory, beta = .32, t = 2.65, p = .009, shown in Figure 6.12; and freedom of

interaction (scheduled co-presence of the different disciplines at meetings), beta = -

.15, t = -2.20, p = .03. In other words, teams whose members reported a high level

of reflexivity, teams whose members reported team processes as positive, as well as

teams scheduling relatively little cross-disciplinary interaction in formal meetings, all

tended to be described by their members as relatively strong with respect to

innovation.

Figure 6.11: Impact of team reflexivity on self-reported innovativeness (data from 113CMHTs)

team reflexivity

6.05.55.04.54.03.53.02.5

inno

vativ

enes

s

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

Figure 6.12: Impact of team processes on self-reported innovativeness (data from 113 CMHTs)

team processes

4.54.03.53.02.52.0

inno

vativ

enes

s

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

Relationships between Team Composition and Processes, and Stress Levels inCMHTs

Three variables predicted, independently of one another, the level of stress reported

by team members on the General Health Questionnaire: Team processes (overall

score on the Team Processes Inventory), beta = -.50, t = -5.22, p < .001; informal

communication (social events, message boards, etc.), beta = .21, t = 2.20, p = .03;

and freedom of interaction (scheduled co-presence of the different disciplines at

meetings), beta = -.20, t = -2.20, p = .04. These findings indicate that the following

team characteristics are associated with relatively high stress (or low stress levels)

amongst staff: a positive team processes; plentiful opportunities for informal

interaction among members; and relatively little provision for scheduled, formal,

cross-disciplinary encounters at team meetings. Figure 6.13 shows the impact of

team processes on stress.

Figure 6.13: Impact of team processes on stress

team processes

4.54.03.53.02.52.0

1.6

1.4

1.2

1.0

.8

.6

GHQ = General Health Questionnaire, 12 item version

Relationships between Team Composition andProcesses, and Turnover in the team

As shown in Figures 6.14 and 6.15, two compositional variables predicted staff

turnover, independently of one another: mean age of team members, beta = - .25, t =

-2.21, p = .03; and the presence of social workers in the sample of respondents, beta

= -.23, t = -2.02, p = .05. These findings suggest that teams with older members

enjoyed more stable membership, as did teams including social workers.

Figure 6.14: Impact of mean age of teammembers on turnover (data from 113 CMHTs)

team processes

4.54.03.53.02.52.0

1.6

1.4

1.2

1.0

.8

.6

Figure 6.15: Staff turnover in CMHTs with andwithout social workers (data from 92 teams)

social workers in the team

yesno

20

15

10

5

0

8

15

Relationships between Team Composition and

Leadership

Two factors, acting independently of one another, were associated with the clarity of

leadership described by team members: the use of a single, integrated set of case

notes for each client, beta = .29, t = 2.85, p = .006, as shown in Figure 6.16; and the

presence of one or more psychiatrists amongst team members completing the

questionnaire, beta = -.27, t = -2.63, p = .01, as shown in Figure 6.17. It should be

noted that, before computing the mean clarity of leadership for each team, responses

from psychiatrists themselves were removed from the latter analysis. In other words,

teams using integrated case notes, and teams not including psychiatrists, were

clearer about the leadership role.

Figure 6.16: Integrated case notes and clarity ofleadership (data from 92 CMHTs)

Use of one integrated set of case notes

Yespartial accessNo

.8

.7

.6

.5

.4

.8

.7

.5

Figure 6.17: Psychiatric membership of the team and clarity of leadership (data from113 CMHTs, with responses of psychiatrists themselves removed from the analysis)

psychiatrists in the team

yesno

.8

.7

.6

.5

.4

.5

.7

Discussion

As for primary health care teams, the findings from this stage of the research reveal a

clear message for CMHT policy and practice, in relation to the factors predicting the

effectiveness and innovations of community mental health teams: positive team

processes, and clarity as to the leadership role within the team, make for a more

effective team, as judged by external stakeholders as well as the team members

themselves; requisite size makes for a higher quality of innovation; reflexive

processes aids innovation; longer-established teams are rated more innovative by

external judges and see themselves as more effective.

We have considered policy implications of the findings reported in this chapter in

relation to the Workforce Action Team issues identified in Chapter 1. In terms of

education and training, it is clear that teamworking skills are key to the effective

delivery of mental health care in the community. These skills are specific and

trainable. They are not acquired implicitly through professional socialisation into

such disciplines as nursing, medicine, clinical psychology, or social work. Nor are

they attained through unfocussed, unsustained “team-building” exercises of the kind

that are widely marketed into the NHS and other large organisations. Rather, they

comprise key types of knowledge, skill and ability required for effective teamworking

(Stephens & Campion, 1994). As depicted in Figure 6.18, these fall into 5 domains:

conflict resolution; collaborative problem-solving; communication; goal-setting and

performance management; planning and task co-ordination. Both initial professional

training and continuing professional development for CMHT members should

incorporate systematic training in these. To prepare and enable members of the

constituent professions to function effectively within CMHTs, the NHS should require

that the majority of such training should be delivered in multiprofessional rather than

uniprofessional learning environments. The requirement for such delivery may

create difficulties for HEI’s but the NHS should work with them to ensure that training

funded by the NHS meets the NHS’s pressing requirements for effective

multiprofessional teamwork.

Figure 6.18: Knowledge, Skill and Ability Requiredfor Effective Team-working (Stephens & Campion,1994)

Conflict resolution:• Recognise and encourage desirable but discourage undesirable team

conflict• Recognise type and source of conflict confronting the team and implement

appropriate resolution strategy• Employ integrative (win-win) negotiation strategy rather than traditional

distributive (win-lose) strategy

Collaborative problem-solving:• Identify situations requiring participative group problem solving and utilise

proper degree and type of participation• Recognise obstacles to collaborative group problem solving and implement

appropriate corrective actions

Communication:

• Understand networks and utilise decentralised networks to

enhance communication where possible• Communicate openly and supportively, sending messages which are (1)

behaviour - or event-oriented; (2) congruent; (3) validating; (4)conjunctive; and (5) owned

• Listen non-evaluatively and appropriately use active listening techniques• Maximize consonance between nonverbal and verbal messages, recognise

and interpret the nonverbal messages of others• Engage in ritual greetings and small talk, and recognition of their

importance

Goal-setting and performance management:• Help establish specific, challenging and accepted team goals• Monitor, evaluate, and provide feedback on both overall team performance

and individual team member performance•

Planning and task co-ordination:

• Co-ordinate and synchronise activities, information and task

interdependencies between team members• Help establish task and role expectations of individual team members and

ensure proper balancing of workload in the team

Stephens, M.J., & Campion, M.A. (1994). The knowledge, skill and abilityrquirements for teamwork: Implications for human resource management. Journal ofManagement, 20, 503-530.

In relation to recruitment and retention, we note the disturbing fact that only 12% of

our CMHTs included all 5 key disciplines: nursing, psychiatry, social work,

occupational therapy, and clinical psychology. This presents a challenge to effective

delivery of the full spectrum of mental health care, and confirms the importance of the

Workforce Action Team’s agenda. In that context, we draw attention to certain

features of the demographics of CMHT staff identified by this research. Sixty-seven

per cent of CMHT staff were women; the mean age of the staff was 40, with a

standard deviation of 8.4 years, and most workers aged between 30 and 50. This

profile highlights the importance of flexible working to accommodate family demands

and thereby retain staff. It also confirms the importance of retaining CMHT staff

beyond the age of 50. Within the somewhat restricted age range we observed,

teams with older members experienced less turnover. Turnover was greater among

smaller teams, suggesting that teams should be large enough to provide sufficient

support. Our findings implicate poor team processes in CMHT staff stress, which is

likely to be inimical to staff retention. Longer-established teams rating themselves as

more effective suggests that stability of the team itself may yield greater job

satisfaction through the experience of effectiveness.

In relation to leadership, we see this as a key and integral feature of team functioning

and hence, as shown by our data linking team processes to effectiveness, vital to the

delivery of effective mental health care. We found that clarity in relation to the

leadership role was reflected in external ratings of CMHT effectiveness, as well as in

team members’ own ratings of their team’s effectiveness. Clear and effective

leadership will be essential to delivery of the National Service Framework, and

development of the required leadership skills, which are learnable irrespective of

professional discipline, will require training resources. Training for CMHT leadership

must relate to the complex multi-agency environment, and combine clarity with

flexibility and innovativeness. Such training needs to be evidence-based, locally

available, and ongoing rather than occasional or intermittent.

In relation to primary care delivery of mental health care, we draw attention to

findings from Chapter 3 highlighting the benefits of larger PHCTs, and of a wide

spectrum of professions being members of the PHCT, alongside the value of clear

PHCT leadership. In the course of our PHCT research we obtained ample anecdotal

evidence that PHCT members are highly aware of the challenges presented to them

by their growing responsibilities in this area.

In relation to the Workforce Action Team’s interest in developing the role of

professionally non-affiliated staff in mental health care delivery, we have some

indicative findings on support workers within CMHTs. They comprised 7% of

respondents to our survey, and 75% of them were female. Their ratings of their

teams were very favourable, and interviews with them during the intensive analysis

reported in Chapter 7 were highly positive in tone. The intensive analysis also

revealed the considerable dependence of CMHTs on support workers for the time-

intensive, practical aspects of care of patients with severe and enduring mental

health problems. Accordingly, we strongly endorse the development of

professionally non-affiliated staff as a resource within CMHTs.

Chapter 7

Community Mental Health TeamsResults from Qualitative Research

Summary of findings

• Basic minimum standards of staffing and hence care are not yet universally

fulfilled by NHS mental health care

• The costs of mental health care vary across teams providing it, over and

above the apparent clinical requirements of the caseload as reflected in

diagnosis and severity

• CMHTs face conflicting demands from primary care and from the needs of

patients with severe and enduring mental health problems

• Venturing beyond our immediate data, we suggest that such key issues are

likely to have important effects on the morale, stress and effectiveness of

CMHT staff and on their capacity to initiate and maintain the innovative,

collaborative and flexible patterns of teamworking whose importance is

highlighted by our survey findings

Sub-sample team characteristics

In this section, we report an analysis of the sub-sample of CMHTs taking part in the

second phase of the project. In this qualitative work we looked in greater depth at

CMHT effectiveness by taking account of clinical and economic performance. Its

objectives were:

• To develop methods to assess the economic costs and clinical outcomes for

a representative patients seen by a CMHT

• To examine the relationships between task context (severity of caseload),

inputs (time, costs), team processes, and outputs (clinical improvement, goal

attainment and CMHT practitioner stress)

Process and effectiveness dimensions of the 10 teams

Figure 7.1 gives the aggregated team means for the primary process and

effectiveness variables.

Teams were selected for the second phase of the research on the basis of combined

standardised team processes scores, self report effectiveness score, and stress

levels (see Chapter 5). Figure 7.1 shows that the 10 teams varied on the

performance dimensions as ‘effective’ or ‘ineffective’: Teams B, G and J perceived

their teams as performing effectively, but this was only partly supported by external

judgements of team performance, in which innovativeness, but not effectiveness,

was rated highly for those teams.

Figure 7.1: Process and effectiveness dimensions of the 10 CMHTs

TeamA

TeamB

TeamC

TeamD

TeamE

TeamF

TeamG

TeamH

TeamI

TeamJ

OverallTeamprocessesMax = 5

2.95 3.85 3.20 2.96 3.06 3.46 4.24 3.30 2.92 3.94

Number ofmeetingsMax = 11

2 5 3 2 4 5 2 5 5 1

Multi-disciplinarycommunicationMax = 5

2.75 3.50 3.50 3.00 2.00 3.50 1.50 2.50 3.00 2.75

TeamMean stresslevel: GHQitem meanMax = 3

1.36 0.95 1.33 0.99 1.08 1.35 0.85 1.10 1.03 0.83

Clarity of teamleadershipMax = 1

0.67 1.00 0.77 0.09 0.77 0.89 0.87 0.15 0.22 0.83

Teamrelationshipwith GP’sMax = 5

4 4 4 5 3 3 3 4 3 3

Teamrelationshipwith TrustMax = 5

5 5 3 4 3 4 4 3 3 3

Teamrelationshipwith SSMax = 5

5 4 4 4 4 4 4 3 4 3

AssessmentWLimplemented

no no no yes no no yes no yes No

Integratedcase notes yes yes yes no no no no no no Yes

Referralspooled All some all all some all all all some All

Self reporteffectivenessMax = 5

3.10 3.70 3.28 2.82 2.66 3.41 4.00 2.72 2.85 3.65

External ratingeffectivenessMax = 5

- 3.13 3.28 2.85 3.12 4.19 3.19 3.59 2.67 -

External ratingof innovationsMax = 5

2.75 3.63 2.25 1.00 2.50 2.25 3.13 2.50 3.00 3.38

Caseload profile

But how does all this relate to the quality of care provided by this sub-sample of

CMHTs? In the intensive phase of the research, team practitioners were asked to:

• Complete a caseload analysis describing the diagnosis, severity, chronicity

and complexity for all clients. These factors were used to (a) describe the

team’s caseload profile; and to (b) select a representative sample of around

40 CMHT’s clients following procedures defined by the research team (see

Methods)

• Approach selected clients, and if appropriate their carers, and invite them to

participate in the research by completing a Service Satisfaction questionnaire

• Participate in interviews structured to elicit biographical, resource use, and

clinical information about the team’s 40 selected clients, retrospectively for

the past 6 months. Data on resource usage were collected relating to number

of contacts with practitioners in the team; number of inpatient days; use of

day hospital, day care or drop-in; medication; outpatient appointments

• Complete the HoNOS monthly for selected clients over the following 6 months

or up to discharge

• To provide information during a second site visit relating to resource use and

clinical outcomes for the team’s sample of representative clients.

Clinical and cost data were collected for 372 CMHT clients for the first 6 months of

the 12-month period, although an attrition rate of around 35 per cent meant that, for

the second 6 months, data for only 241 of those clients were available. The high

attrition rate was accounted for primarily by key workers or care co-ordinators leaving

the CMHT, and other key workers or care co-ordinators being unable to supply

clinical data during the second site visit.

Psychiatrists completed a caseload analysis in only two teams, and for Team J in

particular, this increased the number of clients on the team’s caseload. Figure 7.2

shows the client caseload team by team, in terms of severity and diagnostic group.

Oneway ANOVA tests showed that some teams differed in the severity of their

caseloads F(9,362) = 4.44, p < 001. Post hoc Bonferroni tests indicated that Team

B’s caseload (mean 2.18) was significantly milder than Team E’s (mean 2.70), p =

.003; than Team I’s (mean 2.64), p = .019; and Team J’s (mean 2.62), p = .021; and

marginally milder than Team H’s (mean 2.65), p = .054.

Figure 7.2: Client caseload profile by severity and diagnostic group6

Diagnostic category Mild Moderate Severe TOTAL

Team ASubstance misuse 3Depression 42Anxiety 50Psychosis 55PD 14

5 87 82 174

Team BSubstance misuse 3Depression 43Anxiety 51Psychosis 85PD 9

33 97 78 208

Team CSubstance misuse 3Depression 43Anxiety 44Psychosis 85PD 12

20 83 92 195

Team DSubstance misuse 8Depression 59Anxiety 8Psychosis 54PD 8

28 57 70 165

Team ESubstance misuse 6Depression 24Anxiety 9Psychosis 141PD 18

6 56 136 198

Team FSubstance misuse 3Depression 86Anxiety 45Psychosis 68PD 8

52 90 74 216

Team G

Substance misuse 7Depression 38Anxiety 30Psychosis 79PD 8

21 91 70 182

Team HSubstance misuse 8Depression 41Anxiety 24Psychosis 154PD 17

10 88 153 251

Team ISubstance misuse 0Depression 85Anxiety 42Psychosis 144PD 14

38 84 200 322

Team JSubstance misuse 29Depression 180Anxiety 43Psychosis 170PD 16

25 174 236 435

6 Some disorders (e.g. adjustment to disability) presented in low numbers, therefore were notincluded

During the site visits, all CMHTs emphasised that they experienced a tension

between policy requirements that their case load included enduring mentally ill

people, and the referral patterns of local GPs who continued to refer all adults with

mental health problems. The tension was increased by a third demand in some

localities that CPNs from within the team should conduct clinics in primary care

practices. Most teams had attempted to clarify with GPs the appropriate referrals,

but reported little guidance from their Trust management, and little change in types of

GP referrals.

Varying responses to these conflicting demands emerged from the data. An

important finding was that the three teams (B, C and F) carrying caseloads which had

clients who were relatively moderately ill, all provided clinics in primary care, and

inclusion of their primary care patients accounted for their lower overall caseload

severity.

Mental health status, health economic costs and clinical outcomes

Mental health status, first 6-month period

In this section we present more detailed information about the sample of

mental health status clients selected by teams and their clinical outcomes.

The period was 12 months, although as stated there was a 35 per cent

attrition rate for cases. Some of the selected clients who completed the

service satisfaction questionnaire criticised continuity of care, as a result of

practitioners leaving the team and being allocated to a new key worker.

Figure 7.3 gives a summary of the mental health status of a sample of clients

selected in terms of severity and CPA level. These data were collected 6

months into the review period. Comparing the information in Figure 7.3 with

Figure 7.2 above shows that the severity of selected clients' mental health

condition was broadly representative of teams’ entire caseload profiles.

Looking at Team J, for example, Figure 7.2 showed that 236 (54%) clients

were severely ill; Figure 7.3 below shows that 28 (62%) of Team J’s selected

clients were severely ill.

Figure 7.3: Summary of selected clients’ mental health status

Mild Moderate SevereCPA

level 1CPA

level 2CPA

level 3On

sectionNot onCPA

Team A 1 14 19 12 13 6 2 3Team B 5 23 12 22 7 7 2 3Team C 3 17 16 11 21 1 0 1Team D 3 12 26 6 14 6 3 13Team E 1 10 29 9 21 4 4 5Team F 3 14 19 12 16 5 3 0Team G 3 19 13 19 13 1 2 0Team H 0 9 17 5 9 3 4 9Team I 1 12 26 6 23 1 3 9Team J 0 17 28 - - - 2 -

During the site visits, it emerged that teams used different criteria for determining

CPA level. Some teams had produced guidelines for assigning CPA level, while

others had not. Most teams used the labels ‘level 1’ as least severe, but others used

‘level 1’ as most severe. One team in the sample used ‘level A’ as the most severe.

These had been recoded to represent severity as lowest, level 1, highest, level 3.

Thus how CMHTs used the CPA was non-standard, and indicated variation in the

management of CPA.

More importantly, in terms of the role of CPA in maintaining service standards, we

observed firstly that planned CPA reviews for selected clients were often overdue,

and secondly, that part of the difficulty was getting access to all the practitioners who

it was specified in the work plan must be present at the review. Some teams had a

particular difficulty with accessing psychiatry.

Across the teams, the percentage of CMHT clients who had a formal carer, in

the sense that their key workers judged that clients needed carer support to

live in the community, ranged from 22 to 55 per cent across the teams. We

found that many clients did not have a care plan with formal targets; however,

this varied across teams and was related to severity of caseload. Figure 7.4

gives a team breakdown of selected clients’ diagnosis and severity.

Figure 7.4: Selected clients’ diagnosis and severity

Diagnostic group Mild Moderate Severe TOTAL

Team A Substance misuseDepressionAnxietyPsychosis/PD 1/0

16

6/1

115

9/2 33Team B Substance misuse

DepressionAnxietyPsychosis/PD

23

68

8/13

8/1 40Team C Substance misuse

DepressionAnxietyPsychosis/PD

3 75

3/1

21

12/1 35Team D Substance misuse

DepressionAnxietyPsychosis/PD

12

37

2/0

272

12/2 40Team E Substance misuse

DepressionAnxietyPsychosis/PD

1 1

8/1

21

25/1 40Team F Substance misuse

DepressionAnxietyPsychosis/PD

2

0/1

53

5/0

152

10/1 35Team G Substance misuse

DepressionAnxietyPsychosis/PD

3

236

8/0

2

37/1 35

Team H Substance misuseDepressionAnxietyPsychosis/PD 8/0

32

10/0 23Team I Substance misuse

DepressionAnxietyPsychosis/PD

1 7

4/0

156

10/4 38Team J Substance misuse

DepressionAnxietyPsychosis/PD

94

4/0

131

13/1 45

Figure 7.4 illustrates that, although these 10 teams were similar in terms of

diagnostic categories for which care was provided, the numbers seen within the

serious and enduring categories varied. This was most apparent in team E, in which

83 per cent of selected clients were diagnosed with psychosis, in contrast to team G,

in which a diagnosis of anxiety was almost as prevalent as psychosis.

Health economic costs

Figure 7.5 gives a summary of costs 6 months into the period. Costs are derived as

mean cost per client, and rounded to the nearest pound. Economic costs for in-

patient days, outpatient appointments, day care sessions, and contact with

practitioners were calculated using Unit Costs of Health & Social Care compiled by

Ann Netten and Jane Dennett at the PSSRU University of Kent at Canterbury. We

commissioned advice on the analysis and interpretation of these data from the

Centre for Health Economics, University of York.

The following assumptions were made when calculating costs:

• In-patient days were calculated at £136 per day

• Outpatient appointments were calculated at £97 per appointment

• Day care costs were calculated at £32 per session. One session equates to

half-a-day. All contacts reported were assumed to be one session. Day care

included day hospitals, day centres, drop-ins and workshops

• Calculations for contacts with practitioners were based on a generic cost for

all members of the Community Mental Health Team. The unit cost used was

face to face contact calculated at an hourly rate. All contacts were assumed

to last one hour. Both a minimum cost of £26 per hour and a maximum of £50

per hour were calculated

• Medication costs were calculated using the British National Formulary. Costs

of generics were used in all calculations except where these were not

available

• Contact costs have been computed using the minimum generic costing (£26

per contact), so this is an underestimation of the cost of contacts, although

consistent across all teams

Figure 7.5: Summary of health economic costs: first 6-month period

Medication Contacts Day care OPappointments

Inpatientdays

TOTAL

Team A 247 2655 1156 120 1572 5750Team B 199 950 842 99 126 2216Team C 210 939 1043 119 162 2473Team D 100 2140 1304 213 1012 4769Team E 150 5001 1294 172 3515 10132Team F 169 1138 1003 136 382 2828Team G 193 2530 1197 77 1480 5477Team H 225 2727 1816 131 1496 6395Team I 180 3129 1092 246 1806 6453Team J 325 999 440 229 48 2041

One way ANOVA was used to determine significant differences between

teams. Significant differences were apparent in terms of outpatient (F = 3.94,

p < .001), inpatient (F = 3.41, p < .001) and practitioner contact costs (F =

3.01, p < .01). Using univariate analysis of covariance, controlling for

caseload severity, differences between teams remained statistically

significant, as shown in figures 7.6 to 7.8.

Figure 7.6: Inpatient costs, first 6 months

Source df F Sig.Corrected Model 10 3.785 .000

Intercept 1 1.014 .315SEV 1 6.688 .010

TEAM 9 2.854 .003Error 359

Figure 7.7: Outpatient costs, first 6 months

Source df F Sig.Corrected Model 10 4.192 .000

Intercept 1 .570 .451SEV 1 7.039 .009

TEAM 9 2.972 .002Error 231

Figure 7.8: Contact costs, first 6 months

Source df F Sig.Corrected Model 10 6.178 .000

Intercept 1 3.166 .076SEV 1 32.308 .000

TEAM 9 2.648 .006Error 353

Post hoc Bonferroni tests indicated that Team I had significantly higher outpatient

costs than Teams B, C and G. Post hoc tests also indicated that Team E had

significantly higher inpatient costs than teams B, C, F and J, and that Team E had

significantly higher practitioner costs than teams A, C and J.

Pearson’s R correlations were carried out on caseload severity with all health

economic costs for the first 6-month period. Medication costs were not associated

with other costs or caseload severity. Practitioner contact costs were associated with

outpatient (Pearson’s R .14, p < .01) and inpatient costs (Pearson’s R .25, p < .01).

Practitioner contact costs were also associated with caseload severity, Pearson’s R

.30, p < .01. Outpatient costs were associated with inpatient costs, Pearson’s R .17,

p < .01. Caseload severity was also associated with outpatient costs, Pearson’s R

.12, p < .05, and inpatient costs, Pearson’s R .18, p < .01.

Mental health status, second 6-month period

Whereas in some CMHTs we were able to collect cost and clinical data for

most of the sample of representative clients for the second 6-month period, in

others there was high attrition. Figure 7.9 shows the attrition rate within each

team.

Figure 7.9: Summary of selected clients lost at second 6-month period

First 6 months Second 6 months % lost

Team A 34 21 38Team B 40 27 33Team C 36 17 53Team D 41 12 71Team E 40 32 20Team F 36 29 19Team G 35 31 11Team H 26 7 73

Team I 39 26 33Team J 45 39 13TOTAL 372 241 35

Clients were divided into two independent groups, the first made up of clients for

whom data were collected only in the first 6 months, the second consisted of clients

for whom we collected data at both time points. Using independent sample t-tests,

we found that over the first 6 months clients retained in the sample were marginally

less severe (mean = 0.48) than those lost to the sample over the second 6 months

(mean = 6.80; t = 1.70, p = .09).

For those clients who were on CPA at the first time point, and remained in the

sample, 116 remained at the same level, 25 moved to a lower level, and 63 moved to

a higher level.

Health economic costs, second 6-month period

In Figure 7.10 below, costs are shown for the second 6-month period, based

only on the clients still in the sample, and calculated as described above.

Figure 7.10: Summary of health economic costs: second 6-month period

Medication Contacts Day care OPappointments

Inpatientdays

TOTAL

Team A 136 433 4302 185 648 5704Team B 179 216 551 86 584 1616Team C 188 177 452 102 0 919Team D 136 295 891 162 533 2017Team E 258 597 747 173 1466 3241Team F 269 267 164 171 1913 2784Team G 440 265 364 122 715 1906Team H 436 505 1563 83 2273 4860Team I 310 307 411 224 675 1927Team J 258 223 340 259 883 1963

One way ANOVA was used to determine differences between the teams in

terms of costs. Significant differences were found in practitioner contacts and

day care costs, but not for other costs. Post hoc Bonferroni tests indicated

that practitioner contact costs were accounted for by Team E being

significantly higher than all others except A, D, and H (F = 4.50, p < .001).

Day care costs were significantly higher for Team A than all other teams,

except teams C and H (F = 2.75, p < .01). Pearson’s R correlations were

carried out on client caseload severity with all health economic costs for the

second 6-month period. There were clear associations between severity level

and all costs except medication: with practitioner costs, R = .30, p < .001; with

outpatient costs, R = .23, p < .001; with inpatient costs, R = .20, p < .01; and

with day care costs, R = .13, p < .05.

Clinical outcomes, first and second periods combined

It was not until the end of the 12-month period that practitioners were asked to judge

whether clinical targets for each selected client had not been met at all, had been

partly met, had been fully met, or had been exceeded. Often, given the diagnosis

and chronicity of many of these clients, the targets were simply that they should be

stable or maintained in the community. Sometimes, though, there was a clear

objective to discharge the client, and tests were carried out to see how many of the

proposed discharges subsequently took place.

At the team level, in order to determine team differences, client severity level

and whether predicted targets were met were analysed using Oneway

ANOVA. The analysis indicated that Team E clients continued to be

significantly more severely ill (mean 7.19) than those of Teams B (mean 5.63)

and G (mean 5.77), F = 2.97, p < .01. A marginally significant difference was

found in the level of predicted targets met (F = 1.66, p = .10) and post hoc

Bonferroni tests indicated that Team J’s perceived performance in meeting

predicted targets was marginally better than Team C’s.

Team-level tests were carried out to determine whether the costs incurred by

different teams were related to the diagnostic profile of their clients. Univariate

analyses of covariance were carried out, controlling for the effects of diagnosis and

severity, as shown in Figure 7.11.

Figure 7.11: Costs and clinical targets met controlling for diagnosis and severity

Dependentvariable

Effects of diagnosis Effects of severity Team effects(controlling for diagnosis,

severity)

Fp

Fp F (11,351) p

Inpatient costs

< 1 ns 1.70 .19 3.06 < .01

Outpatient costsDf 11,227

3.00 .08 6.24 < .05 3.39 < .01

Medication costsDf 11,266

< 1 ns 1.56 ns 1.30 ns

Overall costsDf 11,352

< 1 ns 6.65 .01 3.68 < .001

Clinical targetsmet

3.94 < .05 < 1 ns 1.81 .07

Taking into account variation across teams in diagnosis and severity of

clients, teams were significantly different in terms of some health economic

costs incurred, although not for medication costs.

Inpatient costs (computed at £136 per day)

In relation to inpatient costs, Team J (mean cost per client £48) made little

use of inpatient facilities; teams B and C (means £125 and £167 respectively)

also incurred low inpatient costs. Inpatient costs for Team E, some of whose

clients were severely psychotic and at high risk, were higher than all others

(mean £3515).

Outpatient costs (computed at £97 per contact)

Outpatient costs were considered only for those clients who attended appointments.

As expected, both diagnosis and severity had an effect on outpatient costs, with

more serious diagnoses and higher levels of severity positively associated with

higher costs. At the team level, H and B (mean cost per client £97 and £161

respectively) incurred lower costs than did teams I (mean £589) and J (mean £499).

Medication costs

Medication costs were not influenced by diagnosis or severity, and teams did

not differ significantly in terms of their outlay on medication.

Overall costs

Overall cost per client was strongly influenced by severity, though not by

diagnosis. Teams differed significantly in relation to overall costs, with Team

E (mean £5001) higher than all other teams, a dimension partly accounted for

by the high use of support workers visiting clients at home, sometimes daily,

and high inpatient costs. Teams B, C and J (means £950, £964 and £999

respectively) incurred lower overall costs than the other teams in the sample.

We saw above that Teams B and C had relatively more moderate caseloads

than other teams, but this did not apply to Team J.

Clinical targets met

Teams differed marginally in terms of practitioners’ judgements of whether

clinical targets were met, although this was also influenced by client

diagnosis. Teams B and J (mean per client 2.67 and 2.74 respectively)

scored relatively high on this dimension, while Teams A, C and H scored low

(means 2.25, 2.07 and 2.25 respectively). However, a score of over 2.00

indicated that for the average client, clinical targets had been at least partly

met.

Psychiatry input

A serious problem for this sample of CMHTs was their lack of effective input from

psychiatrists. This was more of a problem for teams with a higher number of

severely ill clients, for example, Teams E and I. How this problem manifested varied

across the teams. One team reported that they had direct access to a psychiatrist for

over 2 years. Other teams had negative views about access to medical input. Their

clients had to visit outpatient departments for CPA reviews, rather than this being

conducted at the team base. Some teams reported that although a psychiatrist was

based within the team, and clients attended clinics at team premises for some CPA

reviews, these reviews were not conducted as frequently as specified in the client’s

care plan, because no medical input was available. It was difficult to find locum

psychiatrists so long term absence or secondment also resulted in a lack of

psychiatric cover for the team.

Team processes, team performance, team stress, clinical outcomes,user satisfaction and cost effectiveness

At the team level, in order to determine how the different dimensions of

effectiveness were related, correlation analyses were performed on relevant

compositional (age, tenure, severity of caseload, number of linked GPs,

MINI), process (team processes, clarity of leadership, ability to communicate

across disciplines, relationships with GPs, Trust and Social Services, pooling

of referrals, how quickly emergencies are seen), performance (overall self

report and external evaluation, external evaluation of innovations) and

outcome (targets met, overall costs, and user satisfaction variables). Some

associations were apparent, as shown in Figure 7.12. These data must be

viewed as exploratory and interpreted very cautiously, in view of the

probability of both Type I and Type II errors: calculating such a large number

of correlations invites Type I errors, whilst the small sample of teams incurs

substantial risk of Type II errors.

Figure 7.12: Summary of associations across composition, process and outcome

Association Pearson’s R probability

Caseload severity/annual costs .763 .010

Caseload severity/self report effectivenessTeams feel ineffective when they have severecaseloads

-.655 .040

Clinical targets met/team stress levelTeams whose members feel stressed also reportdisappointing clinical outcomes

-.628 .052

External evaluation of innovation/clarity of teamleadershipGood team leadership recognised by innovationraters

.560 .091

Team processes/self report effectiveness .890 .001

User satisfaction/how quickly emergency referrals areseen .789 .011

MINI/team uses integrated case notesIntegrated case notes tend to be used in moredeprived areas

.766 .016

Referrals pooled in the team/number of linked GP’sA central referral system in response to larger N ofGPs

.708 .049

CMHT relationship with Trust/CMHT relationship withSocial Services .642 .045

Tenure/external rating of effectiveness (n = 8) .699 .054

The survey results suggested that ‘good’ processes and outcomes are associated

with:

• small team size

• few part-time workers

• Health Service only commissioning

• single, clear line of leadership or co-ordination

• rapid response to emergency referrals

• effective communication processes

• external judgement about its effectiveness if the team itself rates its

functioning highly

When we looked in more depth at exemplar CMHTs, we found a wide variety in

practice. During the researchers' visits to the 10 CMHTs in four NHS regions, the

team numbers spent an average of around 20 hours with the research team, which

yielded rich anecdotal evidence in support of the ‘hard’ findings at both survey and

intensive stages of the study.

Size We found that in one very large team CHMT where three smaller teams had

been created, and separate meetings were held for nursing staff and social workers,

co-ordination and communication were problematic. The wider team met only once

every two months to debate and decide team policy and practice.

Single, clear leader or co-ordinator Seven of the sub-sample teams had a clear

leader or co-ordinator, but in three teams the lack of clarity about leadership was

problematic. One team had been without a leader for over two years, which was felt

to be an indicator of under-resourcing and lack of support from the local

commissioners. Practitioners agreed that this situation also made team meetings

difficult, not only in terms of process, but in terms of the struggle to implement and

communicate decisions.

Rapid response to emergency referrals One of the primary agenda items at

CMHT meetings was the implementation of duty systems to cover emergency and

urgent referrals. These clients often had to wait longer than practitioners felt was

ideal. Such new systems were also described by many team members across the

sample as one of the major innovations their teams had implemented in the previous

12 months. Of course, some of these teams had only been brought together within

the previous 12 months.

Effective communication processes Observation of team meetings indicated wide

variation in quality of team communication. In general, meetings had an agenda,

either formally written and circulated before the meeting or informally presented at

the start of the meeting. Most teams kept to the agenda and covered all business.

However, there were wide differences in process. In the most effective teams,

interaction was quick, responsive and supportive, and participation was equal; in

some teams though most people attending did not take part and merely ‘listened in’.

The issue then arises that those people who do not participate in discussion or

decisions do not feel they ‘own’ decisions and are slow to implement them.

Inclusion of social workers At the statutory level, social workers must be involved

in the care of the CMHT client group. Our survey findings indicated that social

workers tended to rate their team’s effectiveness lower than other disciplines. What

we discovered in carrying out the study was that the research process itself was not

so thoroughly underwritten by Social Services employees as by Health personnel. At

some site visits, we also observed the negative attitudes of health personnel towards

social workers, and vice versa. In open-ended statements in the survey, this cross-

disciplinary hostility was evidenced in many teams, partly because nurses were

expected to take on the duties previously seen as only related to social work, for

example, giving housing or benefits advice. However, in the most effective teams in

the sub-sample, social workers were well integrated to provide the delivery of care for

this client group.

Discussion

This intensive analysis of a sub sample CMHTs draws attention to some key issues

in delivering mental health care, as well as providing a demonstration methodology

for looking in detail at the effectiveness of services delivered in terms of participant

evaluations in relation to health care costs. However, in view of the small number of

teams we were able to study at this level of detail, our substantive findings cannot be

interpreted as more than tentative.

The intensively-analysed teams were selected on the basis of team member ratings

of team effectiveness, team processes, and personal stress. The 3 teams whose

members perceived them as effective were seen as more innovative, but not as more

effective, than the remaining teams in the sub sample.

We found that teams varied in the perceived severity of their caseloads, in the

proportion of their caseloads who were suffering from severe and enduring mental

health problems, and in the health care utilisation costs incurred in the treatment of

patients on their caseloads. Two of the 3 teams with relatively low overall costs per

case had relatively high proportions of only moderately severe cases. Health care

utilisation costs differed across teams even when controlling statistically for the

variation in caseload severity across the 10 teams. The different categories of cost

incurred (CMHT practitioner contacts, outpatient costs and inpatient costs) were

positively inter correlated across teams. This suggests that teams differ in terms of

their use of more or fewer services of all kinds, rather than differing in the priority or

availability of the 3 types of service in the care packages delivered by each team.

Only medication costs were unrelated to the other categories of cost.

Further analyses controlled statistically for both diagnosis and caseload severity and

showed that inpatient, outpatient, and overall costs all differed across the 10 teams.

In terms of meeting the clinical targets set by keyworkers themselves, however,

teams differed only marginally ( p = .07) when diagnosis and severity were

controlled. The associations between severity and costs were well-illustrated by the

team with the highest costs being the one that judged its patients to be most severely

ill.

In the course of collecting these data we made important supplementary

observations concerning the targeting, integrity, and likely effectiveness of CMHT

care. The most significant problem was the frequency of overdue Care Programme

Approach reviews, suggesting that CPA is often not implemented effectively. This

was attributed to the unavailability of psychiatric input. This echoes the finding from

the larger survey sample that only 12% of CMHTs included all core disciplines.

However, effective CPA requires availability of all relevant staff for review meetings,

a requirement that goes beyond the mere inclusion of relevant disciplines in the

membership of the team. A further problem was the threat to continuity of care

arising from the staff turnover that accounted for most of the 35% attrition when we

returned to collect data for the second 6-month period.

We were also forcibly reminded of the tension experienced by CMHTs between

policy requirements to focus on severe and enduring mental health problems and the

demands of GPs continuing to refer many adults with mental health problems. This

was reflected in the finding that the 3 teams with relatively moderately ill caseloads

all provided clinics in primary care, with inclusion of their primary care patients

accounting for their lower overall caseload severity. Fulfilment of CMHTs’ remit

under the National Service Framework will require primary care to shoulder its full

burden in relation to the less severe but considerably more prevalent disorders with

which it is tasked by the framework.

Finally, this intensive analysis of clinical data lent some further validation to the

measures used in the full sample of CMHTs: teams with relatively severe caseloads

considered themselves less effective; teams whose members felt under stress also

reported disappointing clinical outcomes; users of services provided by teams

reporting rapid response to emergency referrals were more satisfied with their team’s

service.

Although not definitive, on account of the small sample, this intensive analysis

highlights key issues in mental health service delivery.

Chapter 8

Secondary Health Care Team -Research methods and Sample Details

Introduction

Three studies were conducted with secondary health care teams. The first examined

the relationship between team processes and team member stress. The second

explored whether team membership per se, was associated with stress, and if so,

what factors in the team environment accounted for this association. The third was a

longitudinal study of the relationship between team functioning and subsequent

member turnover from the team.

The Sample in Study 1 in Secondary Care

The Association between Team Membership and Sress.

The purpose of this study was to determine whether team membership conferred

upon NHS employees better mental health than did membership of loose working

groups or a work situation which did not involve working in a team or loose staff

grouping. Four Trusts were selected from nineteen Trusts included in a larger study

of stress in the National Health Service (Borrill et. al., 1998) and 4,500 names were

selected from the Trusts’ staff lists. The next step involved selecting individuals from

these hospitals for possible involvement in the study. For small occupational groups,

where the number of possible respondents was a hundred or fewer, all those on the

staff lists supplied by the hospitals were included. For larger occupational groups,

individuals were randomly selected from staff lists, with a minimum proportion of

20%. Hence, greater proportions were sampled from smaller occupational groups

and from smaller hospitals. A total of 4,500 names was thus selected. Individually

addressed questionnaires were distributed either by hand to the person’s area of

work or through the internal post. 2,263 people returned completed questionnaires,

representing a response rate of 50%.

Questionnaires were sent to all those selected for the sample

The questionnaire was in four sections:

Section 1: Biographical information

Respondents were asked to give information about age, gender, marital status,

number of children, number of other dependants, job title, length of service, hours of

work.

Section 2: Work Characteristics

Scales were used to measure perceived job, work and hospital characteristics (e.g.,

work demands, autonomy, role conflict, influence in decision-making). Full details of

these measures can be found in Haynes et al., (in press) and are available from the

first author of this report. Measures of organisational climate were also included,

which examined 12 dimensions of climate.

The climate measure we employed is based on the Competing Values Model of

organisational effectiveness (Quinn & Rohrbaugh, 1981; Hill, 1998). This model

posits two fundamental organisational dimensions: internal versus external

orientation and emphasis on control versus an emphasis on flexibility. These two

orthogonal dimensions create four domains of organisational emphasis:

Rational Goal Approach - external focus with tight internal control

Open Systems Model - external focus and flexible relationships with the

environment

Internal Processes - internal focus with an emphasis on tight internal control

Human Relations - emphasis on well-being, growth and commitment of

employees.

These approaches reflect the rich mix of competing views and perspectives within an

organisation and Quinn (1988) argues that a balance of these competing

organisational values is required for organisational effectiveness.

The organisational climate questionnaire (Hill, 1998) was developed by selecting

scales from an existing team processes measure considered appropriate to service

organisations. Following extensive pilot work in four NHS Trusts, this Organisational

Climate Questionnaire (OCQ) was used to survey 5,275 health service employees

from 27 Trusts (Hill et. al., 1997). The results of this survey suggested seven core

dimensions: innovation, performance monitoring, autonomy, co-worker co-operation,

training, communication and resources (Hill, 1998).

Innovation

This dimension measures the extent to which the Trust is seen to be responsive to

change. More specifically the scale explored the extent to which senior staff were

interested in suggestions and the development of new ideas. This was a six-item

scale with five intervals and three stems ranging from strongly agree to strongly

disagree. An example item from this scale is “New ideas are readily accepted in the

Trust.” The coefficient alpha for the current sample was 0.91.

Performance Monitoring

This dimension addresses the perception of how adequately job performance is

monitored within the Trust as a whole, and how well staff are informed about their

work performance. The scale consisted of five items with five intervals and three

stems ranging from strongly agree to strongly disagree. An example item from this

scale is “Staff performance is measured on a regular basis.” The coefficient alpha for

the current sample was 0.86.

Autonomy

This dimension measures the extent to which employees feel that they have the

freedom to work in their own way and are given adequate scope and responsibility to

work without constant upward consultation. The scale has six items with five intervals

and three stems ranging from strongly agree to strongly disagree. An example item

from this scale is “Management tightly control the work of those below them.” The

coefficient alpha for the current sample was 0.87.

Co-worker co-operation

This factor measures the extent to which there is co-operation and conflict amongst

staff in the Trust. The scale comprised of six items with five intervals and three

stems ranging from strongly agree to strongly disagree. An example item from this

scale is “People can rely on one another in the Trust.” The coefficient alpha for the

current sample was 0.87.

Training

This dimension measures the employee’s perceptions of degree of emphasis within

the Trust on skill development and the availability of training resources. The scale

comprised of six items with five intervals and three stems ranging from strongly agree

to strongly disagree. An example item from this scale is “Staff are strongly

encouraged to develop their skills in the Trust.” The coefficient alpha for the current

sample was 0.86.

Communication

This dimension measures the employee’s perceptions of information sharing

throughout the Trust, particularly top-down/vertical communication between

management and workers. The scale comprised of five items with five intervals and

three stems ranging from strongly agree to strongly disagree. An example item from

this scale is “Communication between management and staff is excellent in the

Trust.” The coefficient alpha for the current sample was 0.85.

Resources

This dimension measures employees' perceptions of resource allocation and usage

within the Trust. The scale consisted of seven items with five intervals and three

stems ranging from strongly agree to strongly disagree. An example item from this

scale is “There is very little waste of financial resources in the Trust.” The coefficient

alpha for the current sample was 0.77.

Scale Structure and Reliability

Factor analyses and other multivariate techniques demonstrated the empirical

distinctiveness of the scales from each other (Hill, 1998). The scale reliabilities

reported for this study compared well with the original work, which quotes a range of

reliability coefficients from 0.69 to 0.89.

The notion of teamness was operationalised by using the definitions of teams

employed in the literature (e.g. Alderfer, 1977; Hackman, 1987; Guzzo & Shea,

1992; Guzzo, 1996, p. 8; West, 1996b). The following characteristics are commonly

used to define a team:

a) The group is perceived as a social entity by others and has an organisational

identity within a defined function.

b) This is a real group with a task to perform in an organisation from which shared

objectives are developed for the team.

c) There is a degree of interdependence between members of the group and

members interact together to achieve group objectives.

d) There is a degree of differentiation of roles and duties in the group.

e) There is collective responsibility for measurable outputs.

f) Groups are not so large that they constitute an organisation, which has vertical and

horizontal relationships and sub-groups. In practice this is usually a group of less

than 20 members (although there may be some exceptions to this number).

Distilling these characteristics suggests five components of teamness:

• Distinct roles for members of the team,

• Task interdependence - team members rely on each other to perform the task,

• Outcome interdependence - team members' achievement of team goals is

dependent on other members' knowledge, skill and task performance,

• Team identity - team members and other organisational members regard the

group as a team with a clear team level task to perform

• Clear team objectives - there are clear team level objectives.

Section 3: Stress

The main measure of stress was the 12-item version of the General Health

Questionnaire (GHQ-12; Goldberg, 1972; Goldberg & Williams, 1991). The GHQ-12

was designed as a self-administered screening test for detecting minor psychiatric

disorder in the general population. It covers feelings of strain, depression, inability to

cope, anxiety based on insomnia, lack of confidence and other psychological

problems.

Section 4: Team working

Respondents were asked to indicate, by ticking a ‘yes’ or a ‘no’ response option,

whether they worked in a team. To differentiate between those who did and did not

work in a clearly defined team according to our criteria of teamness, but who

indicated in answer to the categorical question that they did work in a team, we

summed responses to 4 questions:

• Does your team have relatively clear objectives?

• Do you frequently work with other team members in order to achieve

these team objectives?

• Are there different roles for team members within this team?

• Is your team recognised by others in the hospital as a clearly defined work

team to perform a specific function?

Those who did not answer, “yes” to all four questions were categorised as being in a

‘quasi team’. Out of the total sample, 283 responded clearly that they did not work in

a team. Of the 1,980 who answered “yes” to the question “Do you work as part of a

clearly defined team?” 692 answered “no” to one or more of these questions and

were therefore categorised as members of “quasi teams”. Thus 283 (12.5%) did not

work in a team, 1,288 (56.9%) worked in a team, and 692 (30.6%) worked in a “quasi

team”.

The Sample in Study 2

The Relationship between Team Processes and Team Member Stress

Using data from official records and the expertise of members of the National Health

Service Executive, ten Trusts were selected for inclusion in this part of the study

Identifying a sample of teams in each Trust was a lengthy process. Discussions

were held with senior managers, who identified teams in their organisations and

suggested contact persons from each team. The researchers then telephoned the

contact person, negotiated their collaboration in the study, and once agreement was

reached, secured the names and location of team members. Contact persons were

asked to distribute questionnaires to their team members.

There was considerable variation in the types of team in Trusts, and it was not

possible to identify a sufficiently large sample of a single type of team that was

common across all Trusts. Six team types predominated:

• nursing care

• management

• medical

• multidisciplinary

• support

• quality improvement teams

Members of 225 teams were invited to take part in the study. Members of 14 teams

declined this invitation. Over a period of 16 months, 193 teams in 10 NHS Trusts

continued to collaborate in the research. Questionnaire responses were received

from 1,237 team members. The numbers in each profession/occupation were: 752

nurses, 114 doctors, 98 administrative staff, 78 managers, 125 professions allied to

medicine (PAMs), 26 professional and technical staff, 26 ancillary staff and 18 of

unknown occupational group. Team sizes ranged from 2 to 44 (mean 11.4, SD =

6.93).

Figure 8.1: Characteristics of NHS Trusts

Type Budget*

(in £ million)

Number of

Staff*

Year of

Trust Status

Location

Teaching 100 3,000 1991 City

Teaching 125 5,000 1990 City

Teaching 120 5,000 1994 City

Teaching 90 5,500 1992 City

District 78 3.250 1994 City

District 56 2,500 1994 Rural

District 38 1,200 1991 Rural

Community 40 2,000 1992 City

Community 57 1,200 1992 Rural

Community 45 2,500 1993 Rural

* Data available from 1996

The ten NHS Trusts included four teaching hospitals, three community Trusts and

three full District Trusts. Numbers of staff ranged from 1,200 to 5,500 (as shown in

Figure 8.1).

Women formed 86 % of the sample. Mean age of team members was 39.58 years

(SD = 10.52, range 17 to 64 years). Mean team tenure was 4.3 years (SD = 4.65,

range one month to 38 years). 5% of the sample had worked in their team for less

than one year, 17% between 1 to 5 years, 25% between 6 and 11 years, 18%

between 12 and 16 years, and 36% had over 16 years service.

The mean caseness of teams was 23.3% (SD = 0.25), with a mean GHQ Likert

score of 0.95 (SD = 0.24). This level is comparable to a group of 71 primary health

care teams (caseness = 21.8%, Borrill & West, 1998).

At individual level this can be contrasted to 26.7% for a larger group of British

health care employees (n = 22,298, SD = 3.09, Mullarkey et al., 1999) and 18.4%

for the general working population (BHPS, Taylor, Brice, Buck, et al., 1995).

NHS employees often belonged to three or more teams (48%), with only 14%

belonging to one team. Teams either met infrequently (30% had not met in the last

month, and 39% ha met once), or frequently (21% of team had met four or more time

in the last month). Most people worked in permanent teams (90% of members).

Questionnaire completed by individual team members

Section 1: Biographical Information

Participants were asked for their job title, age, gender, and duration of tenure in the

NHS.

Section 2: Team Composition

Team members were asked to indicate the size of their teams (number of members),

how frequently they interacted together, whether they were members simultaneously

of other teams, team tenure, and the nature of the team‘s task. Teams were

classified as coming from Teaching, District General Hospitals or Community Trusts.

Section 3: Team Processes

This contained eight measures of team working. Four of these were drawn from the

Team Climate Inventory (Anderson & West, 1994,1998) that is based on a well-

developed theoretical model of team functioning (West, 1990). The four measures

assess levels of:

• team participation

• clarity of and commitment to team objectives

• emphasis on quality

• support for innovation.

Four other measures were included:

• Reflexivity – the extent to which team members reflect upon their team

objectives, strategies and processes and make changes accordingly (West,

1996; West, 2000)

• Teamness – The extent to which the team functions as a team versus a loose

grouping

• Roles - Team members' understanding of the distinctiveness of their own role and

the degree of differentiation of roles within the team. Team members are asked

to consider their understanding of their job in the team and the appropriate use of

skills and knowledge needed to carry out the work. Then they are asked to

consider these themes in relation to other members’ roles.

• Interdependence - Task interdependence is when group members interact and

depend on one another in order to accomplish work.

Section 4: Outcomes

This section of the questionnaire elicited members’ perceptions of team performance

and shared understanding of team goals.

Section 5: Objectives

A single item invited team members to describe their team objectives.

Study 3: Do team inputs and processes predict team member retention?

Research Design

This was a longitudinal research design with data collected six months after

participating teams had completed Study 2. 76 teams were selected from four NHS

Trusts (two community, one teaching and one District General Hospital).

Following Study 2 teams were sent feedback reports. This created opportunity for

dialogue with the teams. A single sheet of questions and an introductory letter was

sent to each team contact. Materials were clearly marked with the team name as

some contacts were members of more than one team. A stamped, addressed

envelope was included to return responses.

Response

Fifty-seven teams participated in the survey (31 nursing care, 13 management, 5

multi-disciplinary, 5 Quality Improvement, 2 medical and 1 administrative support

team). Community Trusts gave the most enthusiastic response (17 of 19 teams,

89%) and teaching Trusts the least (15 of 26 teams, 58%) (overall = 57%). Data

from the Quality Improvement teams were removed from analyses as these teams

are not permanent; and several teams had ceased to exist having completed their

tasks. The final sample for analysis comprised 52 teams (mean size 11.8, SD 6.03,

range 2 to 25).

Nineteen teams did not respond to this survey (mean size 9.0, SD 4.9, range 2 to

19). No information is available as to the turnover in these teams. A comparison of

means was undertaken across study variables obtained at Time 1 to identify any

differences between responders and non-responders. No significant differences

emerged.

Team size and type of Trust are associated with turnover. People are more likely to

leave a team if they work in a teaching Trust (mean rank = 36.0 Kruskal-Wallis one

way ANOVA; Chi-square 8.43, df 2, p < .05) and are members of a larger team

(Pearson correlation: r = .23 between log percentage leaver and team size p < .05,

one-tailed test).

Regression analyses showed team size and Trust type account for between 15% to

18% of the variance in turnover.

There is no evidence to suggest that stress at Time 1 is associated with team

viability at Time 2.

Perceptions of clear team objectives and high levels of participation are significantly

associated with low levels of turnover

Regression analysis (controlling for team size and Trust type) reveals that TCI

variables treated as a block (participation, support for innovation, team objectives

and task orientation) explain 10% of the variance in team retention.

Team Tasks

Secondary health care teams diagnose illnesses, plan and administer treatment for

various conditions, conduct health screening, and provide maternity care. These are

complex tasks that require co-ordination and management, both, in a professional

sense to ensure the best outcomes for patient care, and, in an organisational sense

to ensure that the work conforms to organisational objectives, budgets, and internal

and external standards. There were seven main categories of team included in the

sample and these are shown in Figure 8.2 below:

Figure 8.2: Types of Secondary Health Care Team

Classification Description

Medical Teams These are teams of doctors.

Nursing Care Teams This is a broader notion of team covering all nursingcare to patients / clients. These teams included groupsof staff in addition to nurses such as health careassistants, auxiliaries, ancillary staff, clerical staff, andprofessional staff.

Management Teams These were teams which undertook the task ofmanaging a department, group of wards, orspecialty/directorate. Therefore, a senior nursing teamthat manages a number of wards would be describedas a management team.

Multi-Disciplinary Team These were teams, often of professional staff, whichhad the task of delivering care or a service topatients/clients often in a boundary spanning roleacross departments, wards and specialities. Forexample, an Endoscopy team would contain medical,nursing, and professional staff who may work in amedical, surgical, or investigative context.

Support Teams: Administrative These were teams, often of administrative and clericalstaff, which provided support to the four team typesabove. This support may be secretarial, administrativeor record keeping in nature. For example, a medicalrecords team would be responsible for the storage,retrieval, and distribution of patients’ hospital records.

Support Teams: Ancillary These were teams of ancillary staff such as porters,domestics, and catering staff who provided supportservices for both patients and staff. For example, aportering team would provide support to the whole Trustto transfer patients and goods between locations,provide access to restricted areas and other duties suchas security and staff protection.

Quality Improvement Team These were temporary teams assigned discrete tasks inorder to improve quality of services provided in various

health care areas. For example, a bed hire team wouldbe monitoring in-patient admission activities andensuring appropriate mechanisms were created tofacilitate the provision of hospital beds in theappropriate locations depending upon demand.

The frequency of each type of team in the sample is shown in figure 8.3.

Figure 8.3: Frequency of Team Types in Study 2

86

29

10

51

116

0

20

40

60

80

100

NUR MAN MED MD SUPP QI

Type of team

frequency

In order to portray the variety of work that takes place within secondary health care

we offer a few examples from each task category and describe the team

composition, accountability, and work undertaken.

Nursing Care Teams

The Paediatric Nursing Team

This team is part of an urban community Trust. Team members provide for the

nursing care needs of acutely or chronically sick children at home, or, in other

community settings. There are 12 members of this team: a nurse manager, 2

paediatric community nurse sisters, 4 paediatric community nurses, 2 staff nurses, 2

district nurses, and a paediatric diabetes nurse specialist. The nurse manager is the

team leader. Organisationally this team is part of the Child Health Care Group and is

accountable to the Child Health Management Team. This team can be described as

a complex decision making team which performs multiple tasks using both basic and

specialised equipment. This team interfaces with many other specialised teams in

hospital and community organisations.

Key to Task Type

NUR = Nursing care 45%

MAN = Management teams 15%

MED = Medical teams 5%

MD = Multidisciplinary teams 25%

SUPP = Support teams 6%

QI = Quality improvement teams 3%

The Coronary Care Unit

This team is part of a busy urban teaching Trust. This team provides care for patients

with heart disease and problems associated with acute and chronic capacity. There

are 23 team members: 2 nursing sisters, 15 staff nurses, 2 enrolled nurses, one

domestic, and 3 doctors (a consultant, a senior house officer, and a house officer).

This is a complex decision making team that provides specialised care delivered

using highly specialised and technical equipment. Team members are involved with

patients and their families and take an essential role in rehabilitation. In addition, this

unit undertakes training and education of staff. The team leader is the medical

consultant who is responsible, at Trust Board level, to the Clinical Director of

Medicine.

Ward One

Ward One7 is a busy surgical ward that is part of an urban teaching hospital Trust.

The ward practices team nursing and divides patient care between four teams. Team

members provide care for patients in conjunction with other professional staff such as

occupational therapists and physiotherapists during patients’ post-operative

rehabilitation. Ward One has four members: two health care assistants, one primary

nurse (nursing sister), and one associate nurse (staff nurse). The primary nurse is

the team leader and will co-ordinate with the other three team leaders on the ward.

This team is responsible to the ward manager who is part of the Medical Directorate

Management Team.

Management Teams

The Child Health Management Team

This team is part of an urban community Trust. The team co-ordinates the Children’s

Service in the community. There are 10 members of this team: a general manager,

an assistant general manager, a primary care manager, a district dental officer,

manager of speech and language therapy, finance manager, care group planner,

consultant paediatrician, personnel manager, and a nurse manager. The general

manager is the team leader and is accountable for this team at Trust Board level.

This is a complex decision making team which can be considered the top

management team for the Child Health Care Group.

7 Team names are fictitious in order to maintain confidentiality.

The Clinical Management Team

This team is part of a busy district general hospital. The function of the team is to

manage the General Medicine Directorate and they are responsible to the Trust

Board. There are seven team members: one clinical head of service and a deputy

clinical head (both doctors), a speciality manager, two nurse managers, a bed

manager, and a secretary. The clinical head of service is the team leader. This team

can be described as a top management team.

Multi-disciplinary Teams

The Surgical Oncology Team

This team provides treatment for patients with breast cancer in a busy urban teaching

hospital Trust. There are 14 team members: two consultant surgeons, two ward

sisters, two medical secretaries, two senior house officers, four surgical house

officers, and two breast care nurses. This is a complex decision making team which

provides diagnostic services, treatment and follow-up care for breast cancer patients

using in-patient and out-patient resources. The team leader is the senior consultant

and he is responsible, at Trust Board level, to the clinical director of surgery.

The Medical Practice Team

The team is part of an urban community Trust providing a comprehensive health care

service to a practice population that involves working across the boundaries of local

and community care. There are nine members of this team: two district nursing

sisters, a community staff nurse, a district staff nurse, a community nursing auxiliary,

and four health visitors. A district nursing sister is the team leader responsible to the

Locality Management Team. This is a complex decision making team that works

across the boundaries of primary and secondary health care.

Medical Teams

The Transplant Team

This team is part of a busy teaching hospital Trust. The team provides bone marrow

transplants for adults and children and carers for patients before, during and after

their transplant. This specialist team is involved in the development of new

techniques, training, and research at an international level within the medical

community. There are five members of this team who are all doctors: one acts as

programme co-ordinator, and two other consultants take responsibility for adult and

paediatric patient care. The team leader is the specialty director. This is a complex

decision making team which operates across the boundaries of several medical

specialities. The clinicians are involved in highly complex work that requires the use

and understanding of complex medical and surgical techniques, drug therapies, and

highly technical equipment.

The General Medical Team

This team has a broad remit providing treatment for any medical or social problem

and is part of a large district Trust. This team has four members: a consultant

physician, a registrar, a senior house officer, and a pre-registration house officer.

The consultant physician is the team leader and there is a strict medical hierarchy

within this team. This is a complex decision making team which provides care for

patients within the hospital and the community for a wide variety of acute and chronic

conditions.

Administrative Support Teams

The Clinic Notes Team

This small team provides medical records for patients with outpatient clinic

appointments. This team is part of a busy rural whole district Trust. The team is

composed of three administrative staff. This team will undertake tasks that require

co-ordination across many departments within the Trust and will undertake many

problem-solving tasks. Information technology skills are well developed within this

team. This is a non-hierarchical team, which is responsible to the medical records

manager.

Ancillary Support Teams

St. Jane's Domestics

St. Jane’s is a small community hospital in a busy urban Trust. The hospital is due to

close in the next two years and services will be moved elsewhere. St. Jane’s

domestics are a team of 14 ancillary staff who provides domestic services throughout

the hospital over a twenty-four hour period. The team is lead by a supervisor who is

responsible to the domestic services manager. Each domestic assistant will have a

designated geographic area of work but will be required to work in other areas as the

need arises. The team performs domestic duties and assists ward staff in providing

food and beverages, keeping the ward clean, feeding patients, and helping visitors.

It is likely that the team members will feel more part of the ward team than the

domestic team. Although this type of work would appear to be of low complexity,

domestics need to be able to carry out their work in harmony with the health care

environment of the ward. This requires understanding of the health care process in

order to communicate appropriately with patients and visitors and to adequately

clean highly complex equipment. These tasks are of medium complexity.

Quality Improvement Teams

The Outpatients Quality Improvement Team

This is a temporary team, which is part of a busy whole district Trust. The aim of this

team is to improve the quality of patient care within the Outpatients’ Department,

which covers a broad range of care across medical and surgical specialities. There

are six team members: two outpatients' service managers, a quality assurance co-

ordinator, a senior midwife, an outpatient senior sister, and a medical records

manager. This is a complex decision making team which is involved in generating

many problem solving strategies in order to achieve standards set by the Trust and

the external Patients’ Charter standards. This is not the principal team for most of its

members and has many similarities to a quality circle.

In summary, a majority of these teams are complex decision making groups, which

undertake multiple health care tasks. With the exception of the quality improvement

teams, all are permanent teams with an on-going work remit. The composition of a

majority of these teams is made up of a variety of different occupational groups. In

addition, there are differences of status, pay, conditions of service and hours of work

across these teams.

Conclusion

The diversity of team types, tasks, composition and organisational contexts in

secondary care argues against the use of research designs we employed in primary

care and community mental health. There are no unitary measures of effectiveness

common across these diverse types of teams. Moreover, the nature of their tasks

varies across organisational settings as well as across team types. Consequently,

we focused on three questions:

• Does membership of teams buffer NHS employees in secondary care from the

negative effects of stress at work and, if so, why?

• To what extent and in what ways are team inputs and processes related to team

member mental health in secondary care?

• Do team processes predict team member retention in secondary care?

We provide the answers to these questions in Chapter 9.

Chapter 9

Secondary Care TeamsResults from Surveys

Summary of Findings

• Those working in clearly defined teams in secondary care have lower levels

of stress than those not working in teams or working in loose groupings (quasi

teams).

• Differences between team membership types in stress could be accounted for

by the higher levels of social support and role clarity experienced by those

working in clearly defined teams.

• Those working in teams also perceive greater co-operation amongst all staff

and clearer feedback from the organisation on staff performance than those

not working in clearly defined teams.

• This finding suggests that team membership somehow buffers individuals

from the vagaries of organisational climate. Poor training; resistance to

innovation, low levels of resources, co-operation, feedback on performance,

autonomy, communication and training, appear to affect stress levels

deleteriously much less among those working in clearly defined teams, than

among those not working in teams or working in looser groupings.

• It implies that teams can somehow compensate for the limitations and

frustrations of organisational factors in the work experience of their members,

and that this can significantly influence the level of stress experienced by

organisational members.

• Team processes are significantly associated with stress – better team

functioning is associated with lower team member stress.

• The more frequently team members interact and meet, the better does the

team function.

• The longer team members work together, the clearer their understanding

about each other’s roles.

• The more teams people were members of, the less clear they were about the

teams’ objectives. However, they reported higher levels of emphasis on

quality of care and understanding of others’ roles.

• Those working in larger teams reported lower levels of participation in team

decision making and less clear understanding of team objectives.

• Around 10% of teams in the sample ceased to exist in the six-month study

period and there is an average 6% turnover of team members in the

remaining teams.

• Clear team objectives and high levels of team participation positively predict

member retention.

• Data from our recently completed study at the Aston Centre for Health Service

Organisation Research8 show that the percentage of people working in teams

in acute trusts is associated with lower levels of patient mortality. The more

people who work in teams in Trusts, the lower the number of patient deaths

measured by the Sunday Times (Dr Foster) Mortality Index, deaths within 30

days of emergency surgery and deaths after admission for hip fracture.

Is team membership associated with lower stress?

The data from this study revealed that 283 (12.5%) respondents did not work in a

team, 692 (30.6%) were members of “quasi teams” and 1,288 (56.9%) worked in a

clearly defined team (i.e. they conformed to the criteria of teams specified in the

research design – see chapter 8).

Analysis of variance within and between the three groups (team, non-team, quasi-

team) in relation to stress scores on the GHQ-12 revealed significant differences

between those who worked in teams (mean = .95), those who did not work in teams

(mean = 1.09) and those who worked in quasi teams (mean = 1.03) (f = 15.68; df =

2,2250; p = > 0.001). Using the GHQ “caseness” method of scoring, 98 of those who

definitely did not work in a team were categorised as cases (equivalent to 34.9%).

275 of those who worked in a team were categorised as cases (21.8%) and 203 of

those who worked in a quasi team were categorised as cases (29.7%). Caseness

implies the individual is suffering from a sufficiently high level of stress that they

require and would benefit from some professional help.

The next step in the analysis addressed the question of what could explain these

differences in GHQ scores between the teamworking types. We examined, in turn,

demographic, work role and organisational climate factors.

Demographic factors

To determine the extent to which demographic factors accounted for GHQ

differences between those not working in teams, those working in teams and those

working in quasi teams, we conducted Chi-Squared tests of these groups by

demographic factors.

These included occupational group, gender, whether they had children, marital

status (single/married/living with a partner/separated/widowed/divorced), the

organisation, (i.e., membership of which of the four Trusts that participated in the

study), time employed in the National Health Service, time employed in current post,

and age. There was a significant Chi-Square value only in the case of occupational

group (Chi-Square of 47.73; df =12; p = <0.001). We therefore conducted an

analysis of variance by team type (non-team, team, quasi team) and occupational

group by GHQ scores. There were significant main effects for occupational groups

(f = 5.572; df = 6, 2232; p = <0.001) but this did not remove the effect of team type on

GHQ score (f = 18,645; df = 2, 2232; p = <0.001). There were no significant

interactions between occupational group and team type group on the GHQ.

Work role factors

8 Details are available from West or Borrill (co-authors of this report)

Next we conducted analyses of variance to determine whether work role factors

varied between team membership types (those not working in teams, those working

in teams and those working in quasi teams). There were significant differences

between these types in role clarity, supervisory leadership, social support, feedback,

autonomy and control, and influence over decision-making, with those working in

teams reporting higher levels that those in quasi teams, who in turn report higher

levels than those not working in teams. A directly opposite pattern was found in

relation to role conflict and role ambiguity. There were no differences between the

groups in perceived work demands and hours worked.

Which of those clear differences in work role factors between the groups might

therefore account for the variation between team membership types in GHQ scores?

To answer this question we conducted separate analyses of covariance to examine

the variation between team groups in GHQ scores controlling for each of the work

role factors in turn. Figure 9.1 shows the results, which reveal that none of these

work role factors alone accounts for the difference between the team membership

types in GHQ scores. However, the effect is most reduced by using role clarity and

social support as covariates. Indeed, when these two variables are entered as

covariates together, the difference between team membership types in GHQ scores

is no longer significant (F= 0.955; df = 2,221; P = 0.385). Thus it appears to be the

differences in social support and role clarity between those who work in teams (high

social support and role clarity) and those who do not work in teams or work only in

quasi teams, which account for variations in stress levels between these team

membership types.

Figure 9.1: Analysis of variance of GHQ scores by team membership type (team,quasi team, non-team) controlling for work role factors

Covariate F DF Significanceof F

MainEffect (F)

DF Significanceof F

Role Clarity 213.03 1,2231 <.0001 4.51 2,2231 0.011

Role Conflict 284.36 1,2233 <0.001 9.12 2,2233 <0.001

Feedback 273.24 1,2217 <0.001 4.99 2,2217 .007

Influence overDecisions

60.90 1,2232 <0.001 7.15 2.2232 .001

SupervisoryLeadership

124.55 1,2179 <0.001 5.37 2,2239 .03

Role Conflict 148.16 1,2147 <0.001 12.38 2,2147 <0.001

Autonomy andControl

5.30 1,2216 .02 14.67 2,2216 <0.001

We then examined the data to determine if there were any significant interactions

between work role factors, which were significantly associated with GHQ scores, and

team membership types. There was a significant interaction only between team type

and role conflict level in predicting GHQ scores (R2 change = .0025; F = 3.13; p =

.044). This showed that at low levels of role conflict there were no differences

between the team membership types in GHQ scores, but at high levels of role conflict

there were very large differences, principally between those who were members of

teams and the other two team membership types) those not working in teams or

those working in quasi teams). Perhaps health care teams enable their members to

manage role conflict through an overt process of shared role negotiation, whereas

those who do not work in teams do not have easy access to such a process.

Organisational climate

We also conducted analysis of variance to determine whether organisational climate

factors varied between team membership types. There were highly significant

differences between the team types in perceived autonomy, training provision, level

of organisational resources, organisational attitude to change, organisational

feedback on performance, communication and staff co-operation, with more positive

perceptions among those working in teams and the most negative perceptions

amongst those not working in teams. To determine which of these organisational

climate factors might account for the variation between team membership types in

GHQ scores, we conducted separate analyses of variance entering each climate

variable as a covariate. The results revealed that none of these variables could

entirely account for the difference between team membership types in GHQ scores.

However, when we entered co-operation and feedback on performance together, the

difference in GHQ scores between team membership types was removed (F = 2.01;

df = 2, 2208; p = 0.134). Thus, the difference in perceptions of co-operation between

staff working in the organisation, and the perceptions of the quality of the

organisation's performance feedback to staff, between those working in teams, those

not working in teams, and those only working in quasi teams, appears to account for

the differences between these team membership types in stress levels.

Finally, we checked the data to determine whether there were any significant

interactions between perceptions of organisational climate factors and team

membership types that predicted GHQ scores. There were significant interactions

between team membership type, innovation, feedback on performance, co-operation

between staff, communication, autonomy, and training and organisational resources

when predicting GHQ levels. They show that those working in clearly defined teams

seem less strained than those not working in teams, or working in ‘quasi teams’, by

perceptions of low levels of these organisational climate factors. It is as though, by

working in a team, team members achieve a shared level of self-sufficiency that

buffers team members from the inadequacies of their organisations. Those who are

not members of clearly defined teams seem more affected both positively and

negatively respectively, by the relative presence or absence of those organisational

factors.

Discussion

The results suggest that being part of a team in the high-strain setting of the NHS is

associated with lower levels of stress than if one is not a member of a team or

belongs to only a loosely defined and weakly interdependent team (what we have

called a ‘quasi team’). The results could not be accounted for by demographic

factors, or by individual work role and organisational climate factors. However, the

results clearly suggested that differences between team membership types in stress

could be accounted for by the higher levels of social support and role clarity

experienced by those working in clearly defined teams. This finding is consistent

with theoretical explanations of some of the beneficial effects of teamworking that

propose that teams contribute to a greater and shared sense of role clarity and social

support (Cohen & Bailey, 1997; West, Borrill & Unsworth, 1998; Mohrman, Cohen &

Mohrman, 1995). Roles are socially negotiated sets of mutual expectations and, by

working closely with those in one’s role set, role clarity results. Moreover, since

teamworking, by definition, involves interdependent working with close social contact

and communication, it is likely that team members will experience more support from

colleagues than those whose working relationships are less tightly linked.

Similarly, in relation to organisational climate perceptions, those working in teams

derive a sense of greater co-operation amongst all staff and clearer feedback from

the organisation on staff performance, as a consequence of their team membership

than those not working in clearly defined teams, and this accounts for the differences

between team membership types in stress levels. It is easy to appreciate how

membership of a team (whose members co-operate to achieve shared goals) might

lead to the illusory sense of high levels of co-operation among staff more generally in

the organisation. Another possibility is that staff do co-operate more with those who

are members of teams, perhaps because of their clearer roles and goals, or because

of the greater power conferred by their membership of a group, or as a result of their

more clearly defined social and functional identity - "This person is a member of the

Accident and Emergency Resuscitation Team and I know about their functional

significance and understand what information or resources they require"

Similar explanations can be offered for the effect of the relatively high level of

organisational feedback perceived by team members which accounts (in concert with

perceptions of staff co-operation) for the differences between team membership

types in stress levels These may be illusory perceptions with team members

mistakenly assuming that the higher level of feedback on their performance that they

experience (as a consequence of their team members' feedback to them), can be

attributed also to organisational feedback to staff on performance. It could also be

that as a result of the clear functional identity of the team in the organisation, the

team does get clearer feedback on performance.

Particularly intriguing is our finding of significant interactions between team

membership types and organisational climate perceptions as predictors of stress.

This finding suggests that team membership somehow buffers individuals from the

vagaries of organisational climate. Poor training, resistance to innovation, low levels

of resources, co-operation, feedback on performance, autonomy, communication and

training, appear to affect stress levels deleteriously much less among those working

in clearly defined teams, than among those not working in teams or working in poorly

defined teams. The consistency of these interactions and the fact that they are

largely absent when we examine interactions between work role factors and team

membership types, suggest their robustness. It implies that teams can somehow

compensate for the limitations and frustrations of organisational factors in the work

experience of their members, and that this can significantly influence the level of

stress experienced by organisational members

We can speculate about a number of possible explanations for our findings First, it

may be that those who have lower levels of stress, self select into teams People who

are relatively well-adjusted socially and have lower levels of anxiety may be attracted

to work closely with others since the challenges of teamworking are more

manageable for them than they are for those who experience high levels of anxiety

and uncertainty. Alternatively, it may be that those who are more relaxed and cope

with stress better are selected into teams by existing team members as they offer

less of a threat to the effective social functioning of the team. Finally, of course, it

may be that those who have relatively low levels of stress are less likely to leave

teams, than those with high levels of stress.

Thus those who experience high levels of stress may find teamworking too

demanding and challenging and their stress may also create social dysfunction that

leads to their attrition from the team. These attraction-selection-attrition explanations

for our findings are credible alternatives to the suggestion that it is the effect of

working in teams upon stress that we have discovered in this first study.

Study 2

Do the composition of and the way secondary health care team members work

together affect member stress?

• Team processes are significantly associated with stress – better team

functioning is associated with lower team member stress.

• The more frequently team members interact and meet the better does the

team function.

• The longer team members work together the clearer their understanding

about each other’s roles.

• The more teams people belonged to, the less clear they were about the

teams’ objectives. However, they reported higher levels of emphasis on

quality of care and understanding of others’ roles.

• Those working in larger teams reported lower levels of participation in team

decision making and less clear understanding of team objectives.

Questionnaires were distributed to team members as described in the previous

chapter. 193 teams from 10 NHS Trusts responded. The responses consisted of 1,

237 team members (752 nurses, 114 doctors, 98 administrative staff, 78 managers,

125 professions allied to medicine (PAMs), 26 professional and technical staff, 26

ancillary staff and 18 of unknown occupational group). Team sizes ranged from 2 to

44 (mean 11.4, SD = 6.93).

The overall response rate to this survey was 54%. There were variations in

response across organisational type (from teaching 42% to community 72%) and

across team task type (Support team 30% to Management team 65%). Women

formed 86 % of the sample. Mean age of team members was 39.58 years (SD =

10.52, range 17 to 64 years). Mean team tenure was 4.3 years (SD = 4.65, range

one month to 38 years). 5% of the sample had worked in their team for less than

one year, 17% between 1 to 5 years, 25% between 6 and 11 years, 18% between

12 and 16 years, and 36% had over 16 years’ service in their teams.

Nearly a quarter of those working in teams scored above the cut-off point on the

GHQ, indicating a high levels of stress. The mean caseness of teams was 23.3%

(SD = 0.25), with a mean GHQ Likert score of 0.95 (SD = 0.24). This level is

comparable to the primary health care and community mental health teams

(caseness = 21.8%) but somewhat lower than the 26.8% recorded amongst a

larger group of NHS employees (n = 22,298, SD = 3.09, Borrill et al., 1998) though

higher than the figure of 18.4% for the general working population (BHPS, Taylor,

et al., 1995).

Team Interaction

Teams either met infrequently (30% had not met in the previous month, and

39% had met once), or frequently (21% of teams had met four or more times in

the previous month). Team interaction frequency was significantly related to all

team process variables (with eight of the ten possible relationships being

significant). All relationships were in a positive direction suggesting that

higher frequency of team interaction facilitates team processes.

Team Task

A simple classification of six team task types was used (see previous chapter).

Quality improvement teams rated support for innovation higher than did medical and

ancillary support teams. Managerial teams described their team objectives in greater

detail than support and multidisciplinary teams. Administrative support teams report

greater understanding of each other’s roles compared to teams undertaking medical

and managerial tasks.

Multiple Team Membership

Most people worked in permanent teams (90% of members). Nearly half of this

sample belonged to three or more teams (48%). Multiple team membership was

negatively associated with clarity of and commitment to team objectives and

positively associated with emphasis on quality of care, interdependence, and role

understanding. This suggests a lack of clarity about team objectives may be

counterbalanced by a more vigorous emphasis on quality of care: working in other

teams may enable members to be more vigilant about and aware of quality issues.

Team Size

Team size is positively related to interdependence and negatively related to

participation and clarity of and commitment to team objectives. Many of the larger

teams are ward or department teams, which administer care or undertake

investigations over a 24-hour period. Dependency on other team members is critical

for the safe and efficient delivery of these services. However, membership of larger

teams is also associated with low levels of information sharing and influence over

decision making, and less clear understanding of the team’sobjectives. These

findings are consistent with previous studies (Blau, 1970; Shaw, 1981; Stahelski &

Tsukuda, 1990; Sundstrom et. al., 1990) that demonstrate that as group size

increases there are increased difficulties in communication, co-ordination and

interpersonal relationships.

Team Tenure

Team tenure was not related to stress but was related to mutual role understanding.

Relationships between Team Processes and Stress

Partial Pearson product moment correlations were undertaken with pair-wise

deletion, controlling for team size. Eight of the ten-team processes variables

examined were significantly and negatively associated with stress. This suggests that

good team processes are associated with lower stress. Five process variables

predicted stress: participation, support for innovation, emphasis on quality, team

objectives, and role understanding.

Team Processes and StressRegression analysis at team level revealed team processes accounted for 22.8% of

between team variance in stress. These results are summarised below. After

controlling for team size, three process variables negatively predicted stress (task

reflexivity, p < .01; emphasis on quality; p < .05; interdependence, p < .01).

Figure 9.2: Hierarchical Regression Analysis for Team Processes Predicting Stress(n=193)

Variable B SEB

β

Step 1Team Size .0043 .002

.5.13

R2 .016 df 1 p <.08Step 2Participation .098 .069 .20Support for Innovation -.0074 .060 -.014Team Objectives .065 .045 .14Emphasis on quality -.12 .051 -.31*Reviewing Processes -.14 .05 -.32**Social Relations -.016 .046 -.039Interdependence .11 .043 .24**Team Objectives .008 .02 .030Role -.11 .059 -.19Outcome .0033 .054 .0068

R2 .228 df 10 p <.001

Note * p < . 05, ** p < . 01, *** p < . 001.

Relationships between Stress and Team Processes at the Individual Level ofAnalysis

Process variables were entered in blocks to avoid suppresser effects and finally

entered together. Several process variables significantly predict stress at the

individual level: participation, emphasis on quality (p < .01), social reflexivity (p <

.001), role understanding, and knowledge of team outcomes (p < .05).

Stress and Team Processes Across Two Levels of Analysis

The analysis of this study is complex as a cross-level model (Rousseau, 1985) is

being explored: individual level of stress and team level processes. These levels

are “nested” (hierarchically ordered systems) and exist within the context of a

higher level unit (organisation, the NHS Trust). Strategies that examine

relationships at one level of analysis ignore the influence events at another level

may have on the data. In this case events may occur in the team that might

influence individual stress as well as individual events influencing team morale

(such as team members leaving).

Hierarchical Linear Modelling (HLM; Bryk & Raudenbush, 1992) is a strategy that

investigates data at more than one level of analysis. This analysis takes into

account both individual and team level variance. To explore the relationships in

this study the following research questions were examined. How much does

stress vary across teams? Do team process variables influence stress?

Data from 136 teams (n = 1,121) were analysed using HLM (teams with 4 or less

members were removed from analysis to improve reliability). Variables were

chosen that significantly correlated with stress: participation, support for

innovation, team objectives, emphasis on quality, reflexivity, social relations, role

understanding, knowledge of team outcomes and teamness (a composite of the

four criteria variables to identify team membership). Team tenure, age and gender

were used as control variables.

HLM demonstrates that 3.4% of the total variance in stress is explained by team

factors (this is significant, given that the GHQ measures general life stress rather

than work-related stress in particular). More than75% of this variance can be

explained by team processes (as measured by the TCI scales – clarity of team

objectives, emphasis on quality, participation and support for innovation).

Study 3:

Do Team Inputs and Processes Predict Team Member Retention?

• Around 10% of teams in the sample ceased to exist in the six-month study

period and there is an average of 16% turnover of team members in the

remaining teams.

• Larger teams have lower levels of retention.

• Clear team objectives and high levels of team participation predict

member retention.

Results

57 teams from 4 NHS trusts (75% of those originally participating) participated in the

survey (31 nursing care, 13 management, 5 multi-disciplinary, 5 quality improvement

teams, 2 medical teams and 1 administrative support team). Data from the quality

improvement teams were removed from the data set as these teams were not

permanent and several had ceased to exist having completed their tasks. The final

sample for analysis comprised 52 teams (mean size 11.8, SD 6.03, range 2 to 25).

Nineteen teams did not respond to the survey (mean size 9.0, SD 4.9, range 2 to 19.

No information is available as to the turnover in these teams. A comparison of means

was undertaken across study variables obtained at Time 1 to identify any differences

between responders and non-responders. No significant differences emerged.

A high rate of turnover was reported amongst the teams surveyed. 10% (5) of the

teams ceased to exist and only 12% (6) of teams reported no change in membership

(either joiners or leavers) over the six months period of the study. On average, there

were two leavers per team (mean = 1.90; SD = 1.94; range 0 to 8), and an average

16% turnover during the study period (mean = 16.1; SD = 15.8; range 0 to 71.4%).

26.3% of teams reported no leavers.

Team size and type of Trust are associated with turnover. People are more likely to

leave a team if they work in a teaching Trust (Chi-square 8.43, df 2, p < .05) and are

members of a larger team (Pearson correlation: r = .23 between log percentage

leaver and team size p < .05, one-tailed test). Regression analyses revealed that

team size and Trust type accounted for 15% and 18% of the variance in turnover

respectively. There was no evidence to suggest that stress at Time 1 was

associated with team turnover at Time 2.

Perceptions of clear team objectives and high levels of participation were significantly

associated with low levels of turnover, such that in teams whose members were clear

about and committed to the team objectives, and who reported high levels of team

participation at Time 1, there were lower levels of turnover between the two

measurement points. (See Figure 9.3).

Regression analysis (controlling for team size and Trust type) revealed that team

process variables treated as a block (participation, support for innovation, team

objectives and emphasis on quality) explained 10% of the variance in team turnover

or team member retention.

Conclusions

Overall these three studies suggest the value to NHS employees in secondary care

of working in teams, and particularly in teams that are characterised by clear

objectives, high levels of participation, emphasis on quality and support for

innovation. Taken together with the findings reported in earlier chapters, it

suggests teamworking is a means for promoting effectiveness in the NHS and the

well-being of employees. Perhaps most striking is the finding from a recently

completed study by the Aston research team (West, Borrill and colleagues)

revealing that the percentage of Trust staff working in teams in acute trusts is

associated with lower levels of patient mortality. The more people who work in

teams in Trusts, the lower the number of patient deaths measured by the Sunday

Times (Dr Foster) Mortality Index, deaths within 30 days of emergency surgery and

deaths after admission for hip fracture.

We now turn to examine the results of the intensive analysis of team meetings and

communication amongst primary health care and community mental health teams,

carried out by the Human Communications Research Centre at the Universities of

Glasgow and Edinburgh in collaboration with the Leeds and Aston research groups.

Chapter 10

Meetings and CommunicationResearch Methods

Introduction

Meetings are important for effective teamworking, providing teams with their main

opportunity for communicating. They are particularly important in health care teams

where the team members often work in different locations and therefore have little

opportunity for communication. For this part of the work, we have compiled two

different sources of information about team meetings. The first source details the

range of meetings held within the team and who is invited or expected to attend

them. This information was derived from the practice manager interviews in PHCTs

and the CMHT. This allows us to assess how much communication took place in a

team, especially across different disciplines, where links are usually the poorest.

However, even within teams which hold many meetings, with good links across

disciplines, communication can be good or poor, depending on how those meetings

are conducted. Therefore our second source of information is recordings of team

meetings, from which we draw both quantitative measures and observations about

meeting practice.

Communication and decision making in teams

Communication in meetings is important in teams for two different reasons. First,

effective teamworking requires everyone to be both well-informed and to be invested

in the team's overall goals and plans. Although there are other methods for keeping

team members informed, such as newsletters, bulletin boards, and informal

conversations, meetings are a common way of doing it. Meetings are the most

effective method of involving a group of people in activities which require discussion.

Team members who have been involved in the discussions of the team's goals or

plans, or in the decision-making process itself, are more likely to feel that they "own"

those goals and plans and to work actively to bring them about (Weldon & Weingart,

1993). Second, the plans which a team develops are likely to be better, the wider the

pool of views they take into consideration. Each individual in a team will have their

own unique perspective on how health care can best be provided in that team's

circumstances; in particular, team members from different disciplines encounter very

different situations in their daily working lives and therefore will have different ideas

about what should be done. Providing the best care means synthesising this

diversity of views into a co-ordinated plan which is understood and accepted by

everyone on the team. Therefore two important properties of team communication

are that everyone participates and especially that every discipline is involved in

discussion.

Barriers to effective communication

One of the classic problems for all teams, and not just ones from the health care

sector, is that when there are status differences between team members, higher

status members are more likely to attend the most important meetings. Even within a

single meeting, higher status members are likely to make the contributions which

drive the meeting, such as giving information, asking questions, and making

suggestions; lower status members are usually restricted to relatively short,

responsive contributions such as answering questions or expressing agreement with

something that has been said (Berger, Rosenholtz, & Zelditch Jr., 1980; Berger,

Fisek, Norman, & Zelditch Jr, 1977). This tends to make lower status members less

invested in the team's plans and to limit the range of ideas about potential changes to

increase effectiveness that the team actively discusses. In manufacturing industry,

for instance, it has been argued that strictly hierarchical management structures are

insufficient to deal with the rate of change in the modern business world because

they promote the flow of information downward but not upward, making it difficult to

adapt using information gained "on the ground." (Burns & Stalker, 1966)

Another classic problem, again universal, is that the larger a group discussion, the

more one person will come to dominate that discussion and the more people will sit

silently rather than contributing actively (Bales, Strodtbeck, Mills, & Roseborough,

1951). In fact, the optimal group size for free discussion is five people, and in any

group larger than around eight, no more than eight people say virtually anything

which is said. In status-differentiated groups, it tends to be the high status individuals

who speak and the low status ones who remain quiet (Berger et. al., 1980); where

one person has authority for decision-making, they tend to control the interaction

(Carletta, Garrod, & Fraser-Krauss, 1998). This can create difficulties for larger

teams unless they can find a way to discuss freely issues in small, cross-disciplinary

groups and then pass ideas forward from them. Finally, there are differences in the

properties of very small groups which make them more suitable for free discussion.

Even in relatively small groups, active discussion in a meeting tends to involve just a

few people at a time, but there is evidence that people who participate actively in a

discussion understand and react to it differently from those who simply overhear it

(Schober & Clark, 1989). In non-status-differentiated groups of five people,

discussions are highly interactive, with people's opinions influenced by whoever they

interacted with the most. Even in groups of just ten people, speakers make longer

utterances as if lecturing to the whole group, and whoever speaks the most has the

most influence (Fay, Garrod, & Carletta, 2000). This means that for good discussion

across disciplines, it is important to keep the groups small and make sure that status

is not an issue as far as that is possible, so that everyone has at least some chance

to interact with everyone else. Because people are more likely to have informal

conversations outside meetings with people they encounter and see as similar to

themselves — i.e., staff from the same discipline — this makes cross-disciplinary

discussion in meetings all the more important.

Methods

Our general interview methods have already been detailed in chapter 3; interviews

were held with the practice manager for primary health care teams and the team

leader in the community mental health care teams answered a questionnaire.

Information was collected about the set of meetings held within the team, who was

invited or expected to attend them, how long the meetings were, the purpose of the

meetings, and how frequently they were held. The sample sizes for primary health

care teams and community mental health teams are 67 and 92, respectively. The

rest of this section describes the methods used in order to obtain a corpus of

recorded meetings.

Team Selection

Within the primary health care sector, all teams undertaking the questionnaire and

interview section of the Health Care Team Effectiveness study were invited to

participate in recording of meetings for further study; selection was first-come first-

served, with no additional selection criteria. In particular, we did not select teams

based on meeting size or on the results of the effectiveness questionnaires. Twelve

teams volunteered for this part of the study. Meeting size ranged from three people

to twenty-five. The teams recorded were reasonably representative of the larger

sample of primary health care teams.

Meeting Selection

Recordings were made of multi-disciplinary decision-making meetings as it was

expected that these meetings would best reflect effective teamworking. Multi-

disciplinary meetings were defined as those attended by a range of disciplines.

Decision-making meetings were defined as those where in addition to exchanging

information decisions were made during them, again with active participation from the

different disciplines. Primary health care teams provided access to the team’s

business meetings, in which the day-to-day running of the practice was discussed. In

one team, all decisions affecting the practice were made solely by the partners, all of

whom were doctors, and therefore there were no multi-disciplinary decision-making

meetings to record. In this case the partners’ meeting was recorded. For community

mental health teams, the meetings again were multi-disciplinary business meetings in

which decisions were made about the running of the team. Wherever possible, two

meetings of the same type were recorded for each team. Dates for meeting

recording were at the discretion of the practice; therefore the two meetings recorded

were not always in sequence.

Before each meeting was recorded, the primary contact for the team, usually the

meeting chairman, was briefed that the meeting was to be kept as naturalistic as

possible. It was requested that recorded meetings should be held in their usual

locations, with their usual meeting protocols (agendas, minutes, chairing procedures,

etc.), and that attendance should be the same as if the meeting were not being

recorded. The researcher who tended the recording equipment made herself as

unobtrusive as possible.

Equipment used

Audio recording was conducted with two omni-directional PZM tabletop microphones

linked to different channels of a high quality audiotape recorder; the microphones

were set up so as to maximise channel differentiation but to be unobtrusive enough

that participants would not move them. A single static video camera on a tripod was

trained to record the gross movements of as many of the participants as possible;

this record was used only to aid speaker identification during transcription.

Transcription

Before each meeting was opened all participants introduced themselves and their

occupation and on the basis of this was allocated a speaker number. Therefore the

first person to introduce him/herself became speaker 1, the second speaker 2 and so

on. Each participant was referred to by the same speaker number for the transcripts

of both meetings regardless of when they spoke during the second meeting.

Meetings were transcribed from the audiotapes by an audio typist who had not

attended the meeting. Audio typists transcribed complete contributions in order

according to when they began, labelling each contribution by speaker number, but

did not code finer timing information. Speaker identification was facilitated both by

the video recording and by a seating plan drawn up during the meeting by the person

recording the meeting. In previous work using these methods on four to twelve

person meetings, transcribers were able to agree very reliably who made any one

contribution; using the kappa statistic, K=.93, k = 2, N = 230, with an average of 2%

and a maximum of 6% non-backchannel contributions left as unidentified (Carletta et

al., 1998). A contribution was defined as a period of speech from one individual in

which the only major pauses coincided with silence from the other speakers, so that

the pause was likely to be caused by the speaker thinking and not by the speaker

listening to someone else's contribution. Under this definition, speakers cannot follow

themselves in the speaking order. Overlapped speech was transcribed, with the

extent of the overlap roughly marked. Infrequently, parts of the meetings were

omitted because they were so badly overlapped that we could not track individual

contributions. After transcription, the transcripts were completely anonymised taking

out all staff, patient, place names, place and local authority names or possible team

or person identifiers.

An example transcription excerpt is given in Figure 10.1. Transcription proceeds one

contribution per row. Column one contains the speaker number. Column two

contains the words said, combined with some coding information, and column three

contains any notes which the transcriber wished to make (for instance, about people

entering or leaving the room). Column two codes include /num for the approximate

location of the start of another contribution during overlap, italics for anonymised

text, @ for words from the tape which could not be heard clearly, and some common

descriptions of non-linguistic behaviour such as general laughter within the group.

To make the example clear, everything but transcribed speech is indicated in red.

Figure 10.1: An example of the format used for meeting transcription

1 Shall I open /4 the meeting

4 Yep, let’s get on with it. phonerings

3 My apologies I am going to have to leave before the end. I havean appointment in Place 1.

1 Are you skiving off?

Group laughter

Analytical techniques

The primary interest is in how well teams communicate not just overall but also

across disciplines. The analysis therefore relies on a classification of meeting

participants by occupation. For ease of reference, categories are identified by colour

as well as number. For primary health care teams, we have used the following

categories.

1 GPs2 practice managers3 practice nursing staff, including nurse practitioners4 attached staff (mostly health visitors and district nurses)5 administrative staff (mostly secretaries and receptionists)6 Miscellaneous

For community mental health teams, the categories are instead:

1 Psychiatrists2 nursing staff3 occupational therapists4 psychologists, psychotherapists, and other therapists5 Managers6 staff from social services7 Miscellaneous

Both kinds of teams rarely had miscellaneous staff or miscellaneous meeting

attenders who were not included in the analysis. In most cases, these were visitors,

students, or staff associated with the physical location such as caretakers and

security staff; for community health care teams, there were also sometimes

representatives of user, carer, or voluntary groups and liaisons to local GPs.

For both the entire range of meetings described in interview and for the recorded

meetings, descriptions were produced of which staff categories interact with each

other. These descriptions are best explained in terms of diagrams. For instance,

consider the following diagrammatic representation of one of the recorded primary

health care team meetings:

Figure 10.2: Communication in a primary health care team meeting

1

2

3456

7

8

9

10

11

12

1 314 15

16

1 7

18

1 9

heavy: > 2 1

medi um: > 14

l i ght : > 7

70 minutes68% of team present

In the diagram, each person is represented by a coloured circle, where the colour

represents their occupational category. How often each person spoke immediately

before or after each other person is represented by the line between their two circles.

Heavy lines mean the people took adjacent turns relatively often; light or no lines

means that they took adjacent turns relatively rarely. When people take adjacent

turns in meetings of this type, they are usually (but not always) communicating

directly with each other and addressing the same topic. The actual line darknesses

are determined by the maximum number of times anyone followed anyone in the

meeting and using that to construct quartiles; no line is shown when the number of

adjacent turns the two people took is less than a fourth of this maximum, a light line

when it is less than half, and so on. For instance, in the diagram shown, nineteen

people attended the meeting, of which seven were doctors (indicated in red), and the

heaviest interaction was between participants 3 and 6, 3 and 8, and 8 and 1. Despite

the fact that no pairs were able to interact very many times (as indicated by the

numbers in the legend) this was a quite long meeting. Although the numbers in the

legend are affected by meeting length, smaller meetings are more likely to have high

numbers because there are fewer possible pairs to interact, and therefore the

potential for any given pair to interact is greater. High numbers in the legend of a

large meeting, unless it is unusually long, indicate that most of the possible pairs of

people do not interact with each other at all, and usually means that many of the

people attending the meeting say little or nothing.

In the diagram just observed, quite a few pairs of people interact, showing that, at

least among those who participate actively, the interaction is quite free. However,

although there are many people present at the meeting and they represent all of the

occupational categories, the interaction is almost exclusively among the GPs and

practice manager. It is also possible for interaction to occur primarily between one

person and others within the meeting (see Figure 10.3).

Figure 10.3: Communication in a primary health care team meeting

1

2

3456

7

8

9

10

11

12

1 314 15

16

1 7

18

1 9

heavy: > 7

medi um: > 4

l i ght : > 2

30 minutes40% of team present

This usually indicates strong chairing of the meeting, although the chair may only be

choosing who will speak next rather than actively controlling the topic of the

discussion or making contributions to the topic him or herself. Alternatively, it may

indicate that the purpose of the meeting was for the dominant person to give a report,

with others asking clarification questions as needed. Whether the diagrams indicate

interaction among many pairs or interaction through one central person, the

interaction which exists can be more or less cross-disciplinary, depending on the mix

of people who actively contribute to the meeting.

The same sort of diagram serves for the interview data. Consider the following

depiction of the set of meetings in a community mental health team:

Figure 10.4: Communication in a community mental health team

Here, instead of representing individuals, the coloured circles represent staff

categories. Instead of representing the number of times people took adjacent turns,

the lines show the number of minutes per month which representatives of the two

categories spend in the same meetings. Lines around a circle show all meetings

which someone from that staff discipline attended. For instance, this diagram shows

a team in which psychiatrists never attended meetings, not even ones only with other

psychiatrists. As in the diagrams for the recorded meetings, the darker the line, the

more communication occurred.

As well as providing a descriptive account of meeting practice in health care teams,

we also derive quantitative measures of communication from what we observed.

Recall that good communication among the pairs of individuals in a team and among

the pairs of disciplines is theoretically important for effective teamworking. To

measure this, we have devised a score for freedom of interaction (Carletta et. al.,

1998). For a recorded meeting, the freer the interaction, the more pairs of

participants take adjacent turns. This is reflected in the diagrams by how "starry"

they appear. Similarly, starriness in the diagram for a set of meetings reflects how

free the interaction is in general among the different staff disciplines. Freedom of

interaction is scored based on either the meeting transcripts or the interview data.

The scores vary between 0 and 1, with high scores reflecting high freedom of

interaction. Similarly, for the individual meetings, since it is important to know

whether high status individuals are over-represented in the communication, we score

12

3

4

5

6

v er y l i gh t : < 2 1 3

l i gh t : < 4 2 7

m edi u m : < 6 4 1

h eav y : < 8 5 5

equality of participation from 0 to 1. Meetings with equal participation have the same

darkness of lines coming from each of the participants if they are added together, but

do not necessarily link all the pairs and therefore do not necessarily have very free

interaction. Other measures for both individual meetings and the set of meetings for

a team as a whole consider the amount of communication which occurs, sometimes

divided by individual or discipline, and who attends meetings.

Chapter 11

Analysis of Communication in PHCT teams

Summary of Findings

§ Meetings in primary health care are often badly managed and dysfunctional

§ Attached staff (i.e. health visitors and district nurses) often miss team meetings,yet support for innovation is higher in teams where they attend.

§ Better meeting attendance in PCHTs is associated with care that is more patientcentred.

§ In primary health care, team meetings are often re-arranged, cancelled or start

late.

§ When teams do meet many PHCT team members remain silent throughout themeetings.

§ In over half the meetings we recorded no group decisions are taken.

Types of meetings

"A meeting" is defined as a set of people, who meet usually at some regular interval,

for a particular purpose. Adding together the time devoted to the different meetings

reported - all the meetings which involved any part of the team - shows that primary

health care team members spend relatively little of their time in meetings. On

average, there was a meeting involving some part of the team for 325 minutes per

month (range 22 - 1190, S.D. 240); that is team members spent about 3% of their

time in meetings. The primary health care teams in the sample had between 1 and 6

meetings. The frequency of these meetings ranged from weekly to yearly. Primary

health care meetings tended to fall into the following categories, divided by who

attended them:

Figure 11.1: Primary health care meetings

1

2

3

4

5

Whole team meetings, usually held monthly and attended byeither the whole team or at least by representatives of each ofthe disciplines (22% of sample).

1

2

3

4

5

Single discipline meetings for doctors or for administrativestaff, sometimes with practice management also attending (28%and 3% of sample, respectively). Doctors meetings weretypically weekly or monthly; administrative meetings wereusually monthly or every two months.

1

2

3

4

5

Practice clinical meetings attended only by those cliniciansbased in the practice, and sometimes by the practicemanagement (43% of sample). These meetings were usuallymonthly but some practices held them weekly.

1

2

3

4

5

Full practice meetings, usually held monthly and attended byall staff based at the practice: management, doctors, practicenurses, and administrative staff (7% of sample).

1

2

3

4

5

Clinical staff meetings, usually held monthly and attended byall clinicians included attached ones, and sometimes by thepractice management (16% of sample).

1

2

3

4

5

Nursing meetings attended by the practice nurses andattached staff such as health visitors and district nurses (6% ofsample).

Diagrams shown are representative of the types, but not all meetings in the category

conform completely to the diagram. Nursing meetings were the only ones which were

never attended by the practice management. For each meeting type, there was no

relationship between whether or not a team held a meeting of that type and the

team's size (unrelated t-tests, allowing for unequal sample sizes). The set of

meetings which a team held divided teams into the following categories, with the

following typical diagrams. For each category, we give the mean, minimum, and

maximum size of team with that meeting practice.

• Unitary: Teams with whole team meetings and nothing else (mean team size

14, min. 8, max. 26).

• Multiplex: Teams with both a clinical staff meeting which included attached staff

and either a full practice, practice clinical, or single discipline doctors meeting.

One-quarter of these teams also held a whole team meeting (mean team size 23,

min. 10, max. 51).

• Unitary-plus: Teams which hold whole team meetings plus either a separate

doctors meeting or a separate practice clinical meeting (mean team size 23, min.

10, max. 45).

• Practice-based: Teams whose most inclusive meetings were full practice clinical

ones. These teams sometimes had additional single discipline meetings. In this

category, attached staff such as health visitors never attended any meetings and

administrative staff never met with anyone outside of their single discipline

meeting (mean team size 21, min. 8, max. 64).

• Isolated: Teams which had nothing which could be categorised as a team

meeting. In these cases, the only cross-disciplinary meetings might mix practice

nurses and health visitors. These teams tended to report some single discipline

meetings (mean team size 24, min. 17, max. 37).

Of these types, multiplex, unitary-plus, and practice-based were the most common,

with relatively few teams having just a whole team meeting (the unitary category) or

no true cross-disciplinary meetings (isolated). (See Figure 11.2).

Figure 11.2: Proportions of multiplex, unitary-plus, unitary, practice-based, andisolated Teams

29%

28%7%

27%

9%

Multiplex

Unitary-plus

Unitary

Practice-based

Isolated

In practice-based and isolated teams, communications with attached staff could be

insufficient, causing inefficiencies and lack of direction. Unitary teams might require

a great deal of informal communication to supplement meetings; this is more likely to

be a successful strategy for fairly small teams, which is in fact where the meeting

practice tends to occur.

One might expect teams which have more inclusive meetings - those with whole

team, full practice, or clinical staff meetings - to have higher self-reported

team-working effectiveness scores, because these teams tend to have more chances

for cross-disciplinary communication. Although the occurrence of a whole team

meeting is unrelated to self-reported teamworking effectiveness (unrelated t-test,

t = -1.32, NS), the occurrence of full practice and clinical staff meetings is (for full

practice meetings, t = -2.44, df = 6S, p<.O2 with a mean teamworking score of 4.66

for teams without a meeting and 5.30 for teams with one; for clinical staff meetings,

Levene's F = 4.64, p <.04, with teams without a meeting having the more varied

scores; t with unequal variances = -2.11, df = 48.78, p<.05; mean without meeting

4.63, mean with meeting 5.01). Since the meetings we recorded are whole team

meetings, we will return to why this might be the case after we have considered what

form they take.

Processes within meetings

Meeting practice in PHCTs varies considerably. In this section we describe the

meetings recorded in terms of who attended, how long they were, meeting practice,

and interactional characteristics.

Interpreting date relating to the relationships between communication in the recorded

sample and other variables such as effectiveness and team processes, is

challenging. Primary health care teams are highly variable, not just in terms of input

factors, involving team context. For instance, one of the recorded teams, identified as

Team A, was made up half of new staff who had been brought in with the explicit aim

of changing the ethos of the practice. We observed that the atmosphere in this

practice was quite tense, and, they were one of only three teams in the wider sample

not to hold a Christmas party. This team had very low self-report effectiveness

scores compared to the rest of the recorded sub sample for all but the sub scale

reflecting professional delivery of care. They also declined to have a second meeting

recorded. For this reason, the team was omitted from analyses involving

effectiveness and team processes variables, but shown in graphs. In addition,

differences in meeting practice made it necessary to omit further teams for certain

parts of the analysis.

Because of the small sample size, it is not always possible to test whether

assumptions of normality and linearity are warranted for our statistical analysis. In

order to make the analysis more robust, we categorise teams into two sets, high and

low, for each of the properties of communication which we investigate, and employ

tests. Where these show a difference, we then go on to show the relationship

graphically and to characterise it using correlations.

In most cases, the teams held one meeting which they considered to be for the whole

team. These meetings were open, with all team members expected to attend.

Although the meeting remit was not always clear, the teams used these meetings as

their opportunity to discuss matters affecting the practice. There were three obvious

exceptions. One of the teams, identified as Team B, was strictly controlled by the

partners and never held multi-disciplinary meetings. Their team meetings were

attended by the partners and practice manager only, and even took place away from

the practice, in one of the partners' kitchens. This team had very low effectiveness

and team processes scores compared to the other teams in the sub sample. Like

Team A, they also declined to have a second meeting recorded. Another team,

identified as Team C, did not hold one whole team meeting, but had two highly multi-

disciplinary sub-teams with specific remits which were meant to improve the working

of the practice. One sub-team discussed how to make the best use of the nursing

staff within the practice, while the other discussed initiatives to improve preventative

care (for instance, an anti-smoking campaign). A third team reported that they held

team meetings and gave us permission to record them. However when we

attempted to arrange to record meetings, the team claimed that they were not holding

any meetings which would be appropriate. This team had average self-reported

effectiveness scores.

Figure 11.3: Length of recorded PHCT meetings

meeting length (minutes)

The recorded sample is larger than the set of analysed meetings due to recording difficulties.

PHCTs had regular time set aside for weekly or monthly team meeting. However they

were quite often rearranged or cancelled completely. Meetings often started late,

with people coming in late and leaving early in order to complete their other duties.

Many of those attending were silent throughout and appeared bored: many

commented informally to the researcher that the issues discussed in meetings were

irrelevant to them. Meeting agendas were quite vague. Early business in the

meetings, tabled on the agenda, tended to consist of items which the practice

manager felt it was important to discuss. However, most of the meeting time was

taken up by "any other business" raised by other people present. In many cases,

items which the practice manager were put off in order to accommodate unscheduled

discussion. We observed that for many teams, items raised at the beginning of the

first meeting we recorded had not been dealt with by the end of the second meeting.

Only one team kept rigidly to the agenda, with no unscheduled discussion. Meeting

chairs were usually practice managers, GPs, or practice nurses. (See Figure 11.4).

However, the degree of control which chairs exerted over the meetings varied

considerably, with some chairs, particularly those who were nursery and

administrative staff, merely announced the next step on the agenda as prior

discussion came to a close.

Figure 11.4: Who chaired the recorded PHCT meetings, by occupational class

In the meetings we recorded the decisions made which affected the entire team.

Often the issues discussed were logistical; financial or business issues were

discussed in different meetings. The official agenda in individual meetings tended to

focus on one or two large issues, such as auditing team performance or clinic

management. However, the majority of meeting time was spent discussing less

weighty, more social issues such as what to do on practice nights out, whether to

have a fish tank in the waiting room, and where to go for a Christmas party. These

discussions rarely remained focused therefore they tended to take up more time than

the critical issues.

Who cha ir ed t he r ecor ded PHCT meet ings,by occupa t iona l cla ss.

GP

PM

PN

ATTACHED

ADMINMISC

Although the meetings we recorded had been identified by teams as decision-

making, we found that decisions were not made in the meetings. Where decisions

were made, they were often about how to proceed with the issues discussed; for

instance, in these meetings, the team might decide to call another, often smaller,

meeting for more discussion. Major decisions affecting the team members, such as

changing a clinic date or recruiting more staff, were taken in a different forum and

reported back to the 'decision making' meetings. Thus these meetings were largely

for exchanging information and a forum so that there would be a place where team

members could express their opinions. Some of the individual participants

complained informally to the researcher carrying out the recordings that the meetings

were boring and that the issues which they addressed were completely irrelevant to

them.

Figure 11.5: Primary health care meetings vary considerably in size

Number of People attending PHCT meetings

..

.

To score general attendance, one can use the average proportion of team members

who attended recorded meetings. All of the recorded teams considered their

meetings to be open to all team members except for Team B, which openly restricted

20 75 100 125 200 225 300

Meeting

size

Number of People

attendance to GPs and the practice manager, and Team C, which was organised into

sub-teams. Omitting these two teams, there is a relationship between team size and

the proportion of members attending recorded meetings (see Figure 11.60 (dividing

teams into two sets, small and large, (t = -3.64, df = 7, p = .01 two-tailed; small teams

have the higher proportions).

Figure 11.6: The relationship between team size and meeting attendance for whole team meetings

Team C

Other teams

The relationship between team sizeand

meeting attendance for whole teammeetings.

Team B is omitted from the graph because team size isunavailable. Fitshown without TeamC.

team size

5040302010

.9

.8

.7

.6

.5

.4

.3

.2

.1

One likely reason for this is the workload of health care team members. The bigger

the team, the harder it is to schedule meetings at times that are suitable for

everyone, and the more likely is that members will be unable to attend. This needs to

be taken into account when interpreting results based on this general attendance

score. Team A was omitted from the analysis because it had disproportionately low

effectiveness and Teams B and C because they did not hold whole team meetings.

When the average attendance was divided into two sets, low and high, a relationship

was found with one of the self- reported effectiveness sub scales, patient-

centredness of care (see Figure 11.7) (t = 2.42, df=6, p= 0.5 two tailed). Among the

teams that have a higher proportion of team members attending meetings, the

effectiveness score is higher. If a linear correlation between general attendance and

this effectiveness sub scale is assumed, the same result emerged (r = .73, df = 8, p =

.04). This is the case despite the fact that there is no relationship between team size

and patient-centredness of care, either in the sample of teams we recorded or in the

wider sample.

Figure 11.7: The relationship between general attendance and patient-centredness of care

Team A

Other Teams

The relationship between general attendance

and patient-centredness of care.

Teams B and C have been omitted because their meetings are not open to

the whole team. Fit shown without Team A.

average proportion of team members attending meetings

.9.8.7.6.5.4.3.2.1

6.5

6.0

5.5

5.0

4.5

4.0

These results show that higher levels of attendance at meetings is associated with

effectiveness with respect to quality of patient care. The explanation for this, even

though there are no relationships with the other effective variable, can be found if we

consider how the team use meetings. Even during the meetings, agenda items were

often delayed while team members passed on information about individual patients to

other members who were also involved in their care. Team members also used the

time just before and after meetings to have such discussions, although there the

opportunities were less certain because people often came late or left early.

Although these discussions were only useful to a few of the people present and

therefore might be seen as wasting team time, this was the only opportunity many

had to exchange information.

Team members from the differing occupational groups were not equally likely to go to

the meetings recorded; whereas GPs and practice managers nearly always attended.

(See Figure 11.8). The graph includes team members whether they work full or part-

time with the team. Although some categories are more likely to be part-time than

others, and therefore have difficulty attending meetings, all team members still need

some opportunity to communicate with each other.

Figure 11.8: Who attended at least one meeting (in solid), by occupational class,versus total team membership (complete bars)

In addition, not all meetings recorded had at least one representative from each

of the constituent disciplines; whereas all meetings had at least one GP and one

practice manager present, nearly half the recorded meetings did not involve any

attached staff.

Who attended at least one meeting (in solid),by occupational class, versus total team

membership (complete bars).

0

10

20

30

40

50

60

70

80

90

GP PM PN ATTACHED ADMIN MISC

occupational class

numberinsample

Figure 11.9: For each occupational class, percentage of meetings recorded whichhad someone from that class attending

Team B deliberately excluded all members except GPs and the practice manager.

Two teams where we were able to record two meetings did not have any attached

staff present at either one. Some practice managers remarked informally that for

some individuals, failure to attend was quite regular and tended to cause resentment

among the other team members.

The meetings which we recorded were in all cases the largest and most

multi-disciplinary meetings which the teams held. As a result, we can use who

attended the recorded meetings as a measure of integration in between the different

disciplines. Although team meetings were only one of many ways in which a team

communicated internally, it is likely that team members and disciplines that had poor

attendance at meetings would be less well-informed and less involved in decision-

making. This was most likely to apply to attached staff such as health visitors and

district nurses. Not only were they least likely to attend team meetings, as the

interview data shows, in most team these were the only meetings they were

expected to attend. Much of the work of attached staff was carried out away from the

practice premises, therefore they were least able to communicate with other team

F o r e a c h o c c u p a t io n a l c la s s , p e r c e n t a g e

o f m e e t in g s r e c o r d e d wh ic h h a d s o m e o n e

f r o m t h a t c la s s a t t e n d in g .

0 %

2 0 %

4 0 %

6 0 %

8 0 %

1 0 0 %

1 2 0 %

GP P M P N A T T A DM IN M IS C

o c c u p a t i o n a l c l a s s

members in other ways. From the content of the meetings it was clear that because

attached staff do most of the home visits for the practice more than other team

members, they were more in touch with both the circumstances of individual patients

and patient needs in general. The recordings of team meetings also suggested that

detached staff were the most aware of inefficiencies within the practice arising from

GP's performing tasks that other team members could do (for instance/GPs making

home visits on the same day as an attached team member). Thus for this analysis,

we use two measures to reflect the discipline mix of meeting attendance. The first, a

score for multi-disciplinarity of meetings, is the average number of occupational

categories present at a team's recorded meetings, out of our list of five. The second

is simply whether or not at least one attached staff member was present at one of the

meetings we recorded. These scores are highly related, since no teams ever had an

attached staff member present unless all of the other disciplines were represented as

well; that is, for teams that ever had an attached staff member present, the multi-

disciplinarity rating was over 4.

There is no relationship between self-reported effectiveness and either of these

measures. However, the data suggest a relationship between self-reported support

for innovation and multi-disciplinarity of team meetings, if we omit Team A on the

grounds of its disproportionately poor team processes. If Team A is omitted, support

for innovation is higher when attached staff are present for at least one of the team's

recorded meetings (t = -3.76, df = 8, p = .006 two-tailed).

Dividing the multi-disciplinarity score into two sets, low and high, more multi-

disciplinary teams have higher support for innovation (t = -2.8, df = 8, p = .02 two-

tailed; r = .8347, df = 10, p = .003 two-tailed).

Figure 11.10: Fit shown with (solid) and without (dotted) Team A

This result suggests that team members viewed these meetings as their chance to

raise new ideas within the team. For this sample, as for the wider study, only around

a quarter of the team members providing self-reports for team processes were

attached staff (in this sample, mean 22.56%, min 8.33%~ max 38.46%). Therefore it

is unlikely that the differences were a result of attached staff themselves reporting

that they feel the team supports innovation; a more likely explanation is that their

presence affects the entire team.

We used the 'freedom of interaction' and 'equality of participation' scores previously

to study interactions during meetings in small to medium-sized manufacturing firms.

In that study we confirmed that the scores differentiate groups which operate as

teams, with equal responsibility among the members, from groups in which one

person has overall authority. Equality of participation and freedom of interaction

were higher for the teams, showing that they engage in freer discussion. Primary

health care teams behave like equal responsibility teams and not like the managed

groups. Restricting consideration to teams in the same size range (fewer than

thirteen members), PHCT scores are higher and less varied than the industry scores

as a complete set (for equality of participation, F = 6.725, p = .014 two-tailed;

t = 3.76, df = 3l.53, p = .001 two-tailed; for freedom of interaction, F = 5.028, p = .032

two-tailed; t = 2.50, df = 33.05, p = .017 two-tailed), but indistinguishable from the

T e a m A

O th er te am s

i t sho w n w ith (so lid) andwithout (dotted) TeamAA.

A t least one attached mem ber w as presen t fo r at least one

m eeting if number o f categories presen t exceeds 4.

number of occupational categories present

5 .55.04 .54.03.53.02.52 .01 .5

4.0

3.8

3.6

3.4

3.2

3.0

2.8 *

equal responsibility subset (for equality of participation, F = .72l, NS; t = l.57, df = 22,

NS; for freedom of interaction, F = 3.582, NS; t = .60, df = 22, NS). Although the

teams have nominal chairs, for the most part the meetings are not strictly led. This is

surprising because it is generally difficult to have free discussions in such large

groups. In addition, status differences tend to make interaction less free, and GPs

are both traditionally high status and the employers of many of the team members.

Under the circumstances, if free discussion is what is required, these teams are

doing better at allowing them to occur than one would ordinarily expect.

Conclusion

In primary health care teams, good general attendance at team meetings was linked

to self-reported patient-centredness of care. An explanation for this seems to be that

meetings gave individuals the opportunity to have conversations and exchange

information about patients. However, whole team meetings do not appear to make

the team believe they are more effective in other ways. In particular, teams that have

whole team meetings do not believe they are any better at teamworking than teams

that do not, even though full practice meetings and clinical staff meetings do improve

a team's impression of their teamworking skills. These differences may arise from a

sense of the purpose of a meeting. Team members may have felt that it was

important to meet, but have been unsure about who should go to the meeting and

what should be discussed. Being aware that a meeting is necessary requires that

team members know that people need to communicate, but knowing how to

communicate requires more preparation. Where there is uncertainty about the

purpose of the meeting, the practice was to suggest that everyone (or at least

representatives from each staff group) attended just in case something important was

discussed. It was only possible to know which staff could be excluded when the

remit of the meeting was clear. Although whole team meetings could be useful if

they had a clear purpose that included everyone who attended, the meetings that we

observed often did not have this character. Instead, because of the lack of direction,

many staff members saw the meetings as irrelevant and a waste of their time.

Attending meetings that are seen as irrelevant may have a demoralising effect on

staff with further ramifications for the team's work. Therefore it is important for teams

to consider their meeting practice and to make sure it is designed to best fit their

circumstances.

Chapter 12

Analysis of Communication in CMHT's

_____________________________________________________________________

Summary of Findings

• CMHT members spent three times more time in meetings than PHCT members

• The more cross-disciplinary meetings held in a CMHT the lower the stress levels

in the team

• CMHT meetings were generally well organised and multi-disciplinary

• In 90% of CMHT meetings effective group decisions were taken

• The prototypical CMHT meeting combined operational and clinical decision

making and contained about 9-13 members

• CPNs and Social workers were the best represented in the meetings but there

was also regular attendance of occupational therapists, psychologists and

psychiatrists

Types of Meetings

The number of different meetings was larger and more varied in community mental

health teams than in primary health care teams (range 1 to 11, mean 4.36, SD 1.77).

(See Figure 12.1).

Figure 12.1: Number of meetings

Number of meetings in total

111087654321

30

20

10

0

Teams identified over 20 types of meeting, which we considered to fall within fourcategories:

• Clinical, including audit/quality; day care meetings; ward rounds; representation

at PHCT meetings; CPA reviews; allocation; referrals

• Operational, including MDT meetings; business; locality, sector or patch

meetings; team leader meetings; management; communication

• Strategy, including planning meetings, away days, and team building

• Professional development, including education or training meetings;

professional group meetings; supervision; support

Some of the meetings described, for example, PHCT, ward rounds, locality, sector,patch, or team leader meetings, were not strictly team meetings, although the CMHTwas represented and received clinical, operational or strategic input from such

meetings. There is some confusion between clinical and operational meetings in thedata set because teams tended to mix these two functions, for instance, by callingone meeting which first performed case allocation and then handled business issues.81% of teams reported having clinical meetings and 97% reported having operationalmeetings; these two types of meetings are probably ubiquitous, with the teamsreporting no meetings of a type performing that function as part of another meeting.Strategy and professional development meetings were reported by 16% and 48% ofthe teams, respectively. Probably because they held so many different meetings,counting up the number of minutes per month that at least part of the team is in ameeting gives an average of around 1000 minutes per month (range 140-2940, S.D.608). This means that on average, there was a meeting happening a tenth of thetime involving at least part of the team.

Because of the variety of meeting types, it is less useful to characterise team practice

as a whole in terms of the set of meetings which a team holds than to consider which

disciplines engage in meetings with each other. In these teams, communication was

usually very strong across the constituent disciplines. Teams generally fell into one

of four categories of practice. (See Figure 12.2).

Figure 12.2: Four Categories of Practice

12

3

4

5

6

Teams exhibiting complete connectivity might have somedirect links missing --- for instance, in the example shown,managers never met directly with occupational therapists --- butall disciplines were involved in some cross-disciplinary meetings.In just under half of these teams, all disciplines encountered allothers in meetings.

12

3

4

5

6

Team with one isolate exhibited complete connectivity for five ofthe six disciplines, but one discipline was never involved incross-disciplinary meetings. In two-thirds of these cases, theisolated discipline was management; the remaining cases weredistributed evenly among psychiatry, occupational therapy, andpsychology.

12

3

4

5

6

Teams with a psychiatry + nursing + social services axisshowed good connectivity for these three disciplines. Just overhalf of these cases only ever had these three disciplinescommunicate together in meetings. In the remaining cases,these three disciplines were included, but so was one other, withall the others equally likely to be the additional inclusion.

12

3

4

5

6

Teams with a nursing + social services axis showed goodconnectivity between these two disciplines, and also usuallyincluded cross-disciplinary meetings with one or two otherdisciplines, but never with psychiatry. In over half of thesecases, occupational therapists were involved in meetings, butthere were also examples with management and psychologyinvolvement.

All of these categories were reasonably common in the sample, but the categories

showing better overall connections were more prevalent:

Figure 12.3: Cross-disciplinary communication in CMHT meetings

Cross-disciplinary communication inCMHT Meetings

33%

32%

23%

12%

Complete connectivity

One isolate

Psychiatry+nursing+socialservices axis

Nursing+social services axis

Processes in Meetings

What we can see from this analysis is that in terms of communication in meetings,

nursing and social services staff tend to form the core of the team, with psychiatry in

close contact and management most likely to be isolated. Although what happens in

meetings is not necessarily indicative of communications in the team as a whole,

meetings provide opportunities to discuss work and develop good relationships not

just during the meetings themselves, but also beforehand and afterwards. Therefore

we would expect this pattern to hold for the teams overall, even outside their

meetings.

Which meeting communication pattern a team has is not completely arbitrary.

Omitting three teams with more than 35 members and six teams which were

strangely constituted (usually nursing-only teams rather than cross-disciplinary

teams), freedom of interaction is related to commissioning (one-way ANOVA F (2,79)

= 3.41, p<.05). In our descriptive analysis, the categories divided according to

whether psychiatry was connected to the core team and whether management was

ever involved in cross-disciplinary meetings. Whether or not psychiatrists ever met

with staff from social services is related to how the team was commissioned (X2 =

6.78, df = 2, p<.05).

Teams which are commissioned by health service and social services separately are

more likely to have psychiatrists meet with social services than teams which are

commissioned jointly or by the health service only. Whether managers were

involved in meetings at all was also related to commissioning (X2 = 6.45, df = 2, p<.05

two-tailed); again, they were more likely to be involved in teams commissioned

separately. The general pattern is that teams which are commissioned separately

have stronger cross-disciplinary links than the other types.

The complete details of how our quantitative measures of communication in meetings

relate to other variables within our theoretical model of teamworking are given in

Chapters 1, 2 and 5. Summarising from this analysis, the number of minutes per

month a team's meetings takes is related to the length of time the team has been set

up, the percentage of staff full-time in the one team, and the age of the team

members (younger team members spend more time in meetings). Although we can

not be certain why these relationships occur, they do have reasonable explanations.

Part-time staff are simply harder to schedule into meetings than full-time ones.

Teams which have younger and presumably less-experienced members may provide

a somewhat less complicated service for their local clientele or may be less status-

differentiated than other teams, and therefore meet together rather than dividing by

function into smaller meetings. Mature teams have had more opportunity to structure

themselves to fit their circumstances; they presumably divided their meetings by

function to involve smaller sets of people, or, as sometimes happens, they may

accumulate new meetings for new functions without remembering to end ones which

are no longer useful.

Social workers are one of the key disciplines involved in meetings. If there are social

workers in the CMHT itself rather than accessible from outside the team, then the

overall meeting time for the team was lower. This may be because under these

circumstances more of the team's communication can occur informally. The higher

the percentage of men on the team, the more overall meeting time the team had;

percentage of men is almost certainly standing in for some hidden variable, but the

real cause is not clear. The more meeting time a team had, the more quickly the

team deals with emergency referrals and the more quickly emergencies are seen;

this is probably a result of being able to communicate information related to the

emergency itself. Finally, the freer the interaction among disciplines exhibited by the

set of meetings the team holds, the lower the average level of stress for team

members and the lower a team's self-reported innovativeness. This relationship is

probably complex, since communication with colleagues should, by and large, reduce

work stress and allow team members to express ideas which eventually turn into

team innovations, but innovative teams undergo more changes than non-innovative

teams, and change increases stress.

Results derived from CMHT recorded meetings

Meeting practice in CMHTs followed a more consistent pattern than with the PHCTs.

Overall we found the communication in these meetings to be extremely effective. The

meetings were used to make important group decisions and the content of the

meetings was appropriate to their stated purpose. In this section we describe the

recorded meetings in terms of their purpose, meeting practice, multidisciplinary

representation, and the general interactional characteristics. We then use the

analysis to identify good practice in such meetings and highlight what we believe to

be important contributory factors toward good practice.

As in the case of the recorded PHCT meetings we did not think it appropriate to try

and draw strong conclusions about the relationships between communication in the

recorded sample and other variables such as effectiveness or team processes. The

sample is not sufficiently large or diverse to do this. Instead we describe the results in

more qualitative terms and use them to identify prototypical meeting practices in a

CMHT. On the basis of the purpose, content and general interactional characteristics

of the meetings we then define a good practice prototype for CMHT meetings of the

kind recorded.

The purpose of the recorded CMHT meetings

Teams were asked to select for recording routine meetings with strong

multidisciplinary membership. Generally, they chose meetings that fell into the

operational category described earlier. So the meetings were typically weekly team

business meetings, but frequently they also had a clinical component. We recorded

18 meetings from 9 teams and where possible ensured that they were two

consecutive meetings of the same type from each team in the sample. Unfortunately,

for one of the meetings there was a technical problem in recording so it had to be

dropped from the sample. Hence, the analysis was based on a sample of 17

meetings in 9 CMHTs.

Although the meetings fell into the operational category, as business meetings, they

did vary in terms of purpose and this affected their style. In one case the team

selected for recording two special meetings designed to respond to and influence

health department policy on CMHTs. Membership was much larger than for other

meetings in the sample and was quite different. For example, it included health care

managers and a ward manager. The content and communicative style of these

meetings was also quite different from others in the sample. We refer to these as

High Level Policy (HLP) meetings to differentiate them from the others.

A second kind of meeting associated with two of the teams involved dissemination of

policies imposed from above. The purpose of these meetings was to ensure a

detailed understanding of Trust or DoH policy initiatives. In many respects

communication in these particular meetings was less effective than in others. For one

of the teams much of the meeting time was spent reading documents that would

have been better dealt with outside the meetings and decisions about how to deal

with the policy initiatives were regularly put off to subsequent meetings. However,

this team had adopted a rotating chair practice for their meetings and this limited the

ability of the team leader to control the team’s decision making. It could well have

been this factor which led to the apparent ineffectiveness of the decision making. We

take up this issue in the section on meeting practice. We refer to this kind of meeting

as a Policy Dissemination (PD) meeting.

The most frequent kind of meeting in the sample was the weekly team meeting in

which both clinical issues, such as deferrals, and team policy were formulated.

Typically, teams split the meeting into a section on clinical reporting and case

allocation and a subsequent business section. In most cases the business section of

the meeting dealt with team operational issues rather than high-level policy issues.

We shall refer to these as Mixed Purpose (MP) meetings.

Running meetings in CMHTs

Members of CMHTs spend much more time at meetings than do members of

PHCTs. This greater experience is reflected in a generally high standard of meeting

management. In the majority of recorded cases, team meetings were held regularly,

they were well organised and chaired, and had clear agendas. However, there was

some variation in the organisation and style of chairing that did affect the quality of

the decision making at the meetings.

Three important factors in the running of meetings are their regularity, their size and

their duration. In all cases except the rather anomalous HLP type, the team meetings

took place on a weekly basis. The HLP meetings only occurred every two months.

Twenty-one people attended the HLP meetings, but this was also not representative

of the sample. For the meetings as a whole the average size was 11 and it varied

between 5 and 21. (See Figure 12.4). The average duration of the meetings was 71

minutes and it ranged from 38 to 140 minutes. (See Figure 12.5). In most cases the

meetings were scheduled for no more than an hour, but there were a few occasions

when this was extended to two hours for both clinical and business meetings.

Figure 12.4: Size of meetings in terms of number of people present

0

3

6

9

12

15

18

21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

People present

Num

ber

of p

eopl

e

Meeting

Figure 12.5: Duration of meetings in minutes

For multidisciplinary decision making teams choosing the right size of meetings is

always going to be difficult. On the one hand, groups of more than about 8 or 9 tend

to fall into a non-interactive mode of communication in which each speaker

broadcasts information to the rest of the group. This hinders the process of finding a

true consensus in the group. On the other hand, it is important to have a sufficient

number of people present to ensure effective representation of the different

disciplinary interests. In this situation meetings should be restricted to no more than

about 12 or 13 members and there needs to be an experienced chair. The chair can

then ensure that all relevant voices are heard and promote interactive discussion

when it is needed. This was the practice in a large proportion of the MP type

meetings that we recorded.

For example, in one of the MP meetings the team was confronted with a crucial

decision about how to respond as a team to a change in the sectors they were to

cover. This change, which had been introduced without consultation, meant that

there would be a reduced overlap between the areas covered by team members. The

question was whether they should respond by splitting into two separate groups for

their meetings or remain as a single group. The meeting contained 13 members and

up until this point was characterised by a non-interactive broadcast style of

discussion. So the team leader and chair of the meeting first promoted an extensive

discussion of this issue in which he did not intervene. Then, when everyone had had

0

20

40

60

80

100

120

140

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Leng

th o

f Mee

ting

Meeting

their say, he went round the whole group and allowed each member to indicate their

feelings on the issue before confirming the decision. So although the group contained

13 members it was possible to establish a clear consensual decision about this

important matter. Interestingly, this meeting also managed to get through as much

business as most of the others. Yet, this was all achieved in 40minutes; it was one of

the shortest meetings in our sample.

By contrast, one of the much larger HLP meetings (21 members) faced a situation in

which a group decision should have been made, but it was left up in the air. A young

CNP raised a problem of risk management and safety in the light of a recent

traumatic experience. She had been on a routine visit and on arrival confronted a

suspected suicide. Because she had no portable ‘phone and was in a poorly

provisioned area of the City, she had to return to base before being able to call for

support. The question confronting the meeting was whether to push for provision of

portable phones to all staff engaged in domiciliary visits. This item received

considerable discussion, but because the group was so large the discussion

amounted to a series of long broadcasts where different members expressed their

opinions on everything to do with risk management. After nearly 10 minutes

discussion the problem had not been resolved and was not deferred for subsequent

decision. This does not reflect on the will of the people at the meeting or the ability of

the chair, but rather on the extreme difficulty of making effective group decisions in

meetings of 21 members. In a smaller MP meeting, when confronted with a less

dramatic example of the same problem, the team managed to come up with a co-

ordinated policy, including the provision of portable phones, for just such cases as

this one.

The point illustrated here is that meeting size is crucial to effective group decision

making and that even with relatively small groups of around 12 or 13 members the

process relies on skilful chairing. In relation to this point, two of the teams in our

sample adopted a practice of rotating chairs for meetings. Both had relatively small

meetings (between 6 and 9 members at each). However, it was apparent that the

quality of the group decision making was affected by the practice. For instance, in

one case there was real confusion about who was to monitor and control the

decision; whether it should be the team manager or the chair. In another case at

least 5 minutes was wasted establishing who was to chair the meeting and who to

take the minutes. Although it may seem helpful to give members experience of

chairing meetings, such cases illustrate that ineffective chairing will certainly reduce

the effectiveness of the team’s decision making processes.

For about half of the teams in this sample meeting size was restricted to less than 9

members, which is quite an appropriate size for effective interactive discussion and

group decision.

The content of CMHT meetings and decision making

The content of the recorded meetings fit in with the goals of the meetings. There

were three main areas of discussion: policy, team operations and clinical allocations.

The proportion of meetings in the sample that covered each of these topics is shown

in the Figure 12.6.

The three broad categories of meetings HLP, PD and MPD discussed the three kinds

of issue according to their goals. The HLP meeting predominantly discussed a “green

paper” on mental health care provision in their city. This was quite appropriate to the

meeting because the team had been chosen to elicit feedback from relevant

community groups on the content of the paper. The PD type meetings also discussed

policy, but more in the context of detailed policy documents that had been sent to

their team leader. In MP type meetings there was often also reference to policy, but

only in so far as it was pertinent to particular issues arising from either clinical cases

or team operation.

The second main topic of discussion was what we have called team operation: by

this we mean practices or policies to be adopted by the team that affect the way the

team works. Not surprisingly this topic arose in most of the meetings we recorded. It

was also the source of most team decision making that occurred in the meetings.

Finally, the MP teams also discussed clinical matters. In the meetings we recorded

most clinical (i.e., patient oriented) discussion concerned allocation of cases.

However, there was also discussion of particular problems associated with difficult

cases. For example, in one such case a sectioned patient had been on leave in her

hometown in India. She had written to indicate that she was being held by the family

against her will and was requesting repatriation as a British subject. The team had to

work out an appropriate response to this situation.

Figure 12.6: Content and decision making in CMHT team meetings in terms of %meetings discussing these topics and making group decisions

The figure also shows the proportion of meetings which resulted in 1 or more group

decisions. As can be seen such decisions were made in 88% of the recorded

meetings. These decisions varied from straightforward matters of how to co-ordinate

reports between the nurses, social workers and occupational therapists to more

complicated matters such as formulating an effective security policy for staff on

domiciliary visits. Interestingly these two issues arose in a number of the meetings

and the teams tended to come up with slightly different solutions. The diversity of

decision in operational matters reflected the different circumstances of the teams and

seemed perfectly appropriate.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Policy Teamoperation

Clinical GroupDecision

% o

f mee

tings

Type of decision

Attendance at meetings

The CMHT meetings were attended by a broad range of different categories of staff

representing different professional groups. The figure shows percentages of

meetings attended by each of the major staff categories (see Figure 12.7).

Figure 12.7: Percentage of meetings with a representative from each staff category

All the meetings had representation of both the nursing staff and the social services.

In fact, community psychiatric nurses and social workers were nearly always in the

majority at the meetings. Across all meetings 37% of attendees were CPNs and 24%

were social workers. In more than half of the meetings there were also occupational

therapists, but they only represented 9% of the membership across all meetings. The

two other staff categories of psychiatrists and psychologists were less well

represented. Psychiatrists represented 6% of the attendees and psychologists only

4%.

In a situation where different disciplines are associated with marked differences in

status (e.g., between consultant psychiatrists and CPNs) multidisciplinary

representation at meetings can become a barrier to group decision making. On the

one hand, it is important to have representation from as many groups as possible;

but on the other hand, it is also important to minimise any major disparities in the

0%10%20%30%40%50%60%70%80%90%

100%

psyc

hiat

rists

nurs

ing

staf

f

occu

patio

nal t

hera

pist

s

psyc

holo

gist

s, p

sych

othe

rap.

..m

anag

ers

staf

f fro

m s

ocia

l ser

vice

sm

isce

llane

ous

staf

f

% of meetings

status of the group members. This is especially true for larger groups in which high

status dominant speakers exert a disproportionate influence on the discussion. This

means that notional group decisions tend to become individual decisions that do not

reflect the broader interests of the group.

So it is interesting that there were psychiatrists at only 1 in 10 of the meetings.

Furthermore, we had the strong impression that the meetings in which they were

present were not so effective at making group decisions. They tended to be longer

(average 91 minutes compared to 72 minutes for the others) and generally the

communication was less interactive.

One strategy adopted by CMHTs was to have nursing staff from the hospitals

represent the consultant’s cases at the meeting. This seemed to be a very effective

way of ensuring broader representation without introducing marked disparities of

status within the meeting itself.

Communication dynamics in CMHT

Unfortunately, we encountered some problems with the quality of the video recording

for CMHT meetings. This made it difficult to establish quantitative measures of the

degree of interaction of the kind made with the PHCT meetings. Also the mixed

nature of many meetings with a clinical reporting section followed by a business

section would have made it difficult to interpret overall measures of interactivity and

participation.

From other sources of information, such as the relative length of contributions it is

possible to draw general conclusions about the interactivity of the discussion in the

meetings in the sample. The major contributory factor to interactivity was simply the

size of the meeting. In general, the larger the discussion group the lower the

interactivity in the discussion and this was reflected in the CMHT meetings.

The prototypical CMHT meeting: recommendations on good practice

On the basis of this sample, which represents around 20 hours of discussion in

CMHTs, it is possible to define a prototypical meeting. This can be used both to

describe the overall nature of CMHT team meetings as they occur in our sample and

as a framework for giving recommendations about good practice.

Within our sample of meetings a clear pattern of practice emerged that corresponded

to the results from the interview data. The most common type of meeting was an

operational meeting, which included both a clinical component and a separate

business component. This was the case for more than half of the recorded meetings.

In terms of size this prototypical type of meeting ranged from 5 to 13 members. In

terms of multi-disciplinary representation it fit into the one isolate style of meeting. In

other words, the meeting always contained representatives from nursing and social

work but had either psychiatrists or occupation therapists or psychologists not

represented. We feel that this is probably quite satisfactory in terms of group decision

making because it allows for a sufficiently broad representation without

compromising the size of the group or producing disparities in status.

In relation to this prototype, meetings with around 9-13 members were generally the

most satisfactory for both routine reporting and case management as well as more

general group decision making. However, for the larger meetings to be effective there

had to be skilled chairing of the meetings. We would strongly recommend that

meetings be chaired by team leaders where possible and that they be given some

training in effective meeting management.

In contrast, the three meetings with membership of 15 or more were noticeably less

effective in terms of decision making. It seemed that the very small team meetings

with 5 or 6 members were also less effective. Despite often being very interactive in

terms of the discussion they seemed to have more trouble making group decisions

and on one occasion ended in conflict between two of the groups being represented.

This was the only occasion in all the recorded meetings where such a conflict arose.

Conclusions

The prototypical CMHT meeting contains between 9 and 13 members, it combines

clinical discussion with team business and is used to make team decisions.

Communication in community mental health care teams was much more effective

than in the primary health care teams. On average team members spent three times

as much time at meetings as members of PCHTs. Also, there was strong cross-

disciplinary interaction over the range of meetings that they held. Interestingly, the

freedom of cross-disciplinary interaction was associated with reductions in team

stress levels (as derived from the interview data). Furthermore, the meetings we

recorded showed evidence of high quality chairing and were generally well

organised. Unlike many of the PHCT meetings group decisions were made in 90% of

the meetings and many teams came up with effective ways of ensuring

multidisciplinary representation without introducing too much disparity in the status of

the members.

General Conclusions

In Chapters 11 and 12 we have looked at communication in both primary andcommunity mental health care teams. The analysis was based both on the interviewdata from a large sample of teams in the two health sectors and the detailed analysisof recorded team meetings from a sub-sample of teams. The findings highlight majordifferences in the quality of communication and group decision making betweenPHCTs and CMHTs. In PHCTs there are few occasions where the whole team getstogether in a recognized forum to discuss their activities as a team. When they do soin team meetings the communication and decision making is not particularly effective.In less than half the meetings we recorded not group decisions were taken. Teammembers who are attached to a practice regular do not attend the team meetings.Yet, there was evidence that teams were they do attend have stronger support forinnovation. To the extent that the team meetings are effective, they enable pairs ofmembers to sort out bilateral problems, such as coordination of patient visits.

In contrast, meetings play an important role in the day-to-day operation of CMHTs.On the basis of the interview data it seems that members spend about 3 times asmuch of their time at meetings as do members of PHCTs. Most of the meetings thatthey attend are multidisciplinary and there is generally good connectivity across thedifferent disciplines in teams at these meetings. Overall, the recorded meetings werewell managed and spent most of their time discussing topics on their agendas. Inabout 90% of meetings appropriate group decisions were made and there wasevidence in some meetings of skilled management of this process by the chair.

In relation to the primary health care teams we would recommend that someattention is given to training in the management of multidisciplinary meetings. We feelthat PHCTs should recognize the importance of group decision-making processes insupporting an effective team. For CMHTs the communication and meeting practicesare in general more professionally organized and on occasion match the higheststandards observed in industrial team meetings in which the members have hadtraining in communication and decision making. However, we would also like tounderline how important it is for these meetings to be chaired effectively. This wasone of the most striking observations from the CMHT meetings.

Chapter 13

Conclusions and Recommendations

We began this research programme by asking whether team working in healthcare

was associated with better quality patient care. We also asked what factors were

associated with effective teamwork in healthcare organisations. Using a model that

related team inputs and team processes to healthcare outcomes we were able to

conduct research which has directly addressed these questions. The results paint a

clear picture and provide firm answers to the questions we addressed.

§ Inputs such as team composition and organisational factors have a strong

influence upon innovation and effectiveness in healthcare teams.

§ Team processes are directly related to the innovation and effectiveness of health

care teams across sectors.

§ The quality of teamwork is directly and positively related to quality of patient care

and innovation in healthcare.

• There is a significant and negative relationship

between the percentage of staff working in teams

in acute hospitals and the mortality rate in those

hospitals, taking account of local health care

needs and hospital size1. Where more

employees work in teams the death rate among

patients is significantly lower (calculated on the

basis of the Sunday Times Mortality Index Dr.

Foster; deaths within 30 days of emergency

surgery and deaths after admission for hip

fracture)9.

§ Effective teamwork in primary healthcare teams is associated with lower stress

among team members.

§ In those teams characterised by clear leadership, high levels of integration, good

communication and effective team processes, team members have good mental

health and low stress levels. In secondary health care settings the retention rates

of staff are higher in those teams characterised by good team processes.

§ In this setting also, the research demonstrated that team membership itself

confers role clarity and social support on team members, helping them to achieve

better mental health or lower stress than their counterparts not working in clearly

defined teams.

§ Good team processes means clear, shared objectives amongst team members;

high levels of participation including frequency of interaction, quality of

information sharing and shared influence over decision making; emphasis on

high quality patient care within teams and a preparedness to encourage

constructive controversy but to discourage interpersonal conflict; and practical

support for ideas for new and improved ways for providing healthcare.

§ Effective and innovative teams are characterised by a pattern of reflexivity. Team

members collectively and individually take time out to review the objectives,

strategies and processes of the team; they prepare plans for making changes

accordingly; and they implement those plans in action.

§ Such reflexivity leads to both better quality healthcare and higher levels of

innovation.

9 This finding is based on separately funded research recently completed by the research

team at the Aston Centre for Health Services Organisation Research (further details availablefrom West or Borrill).

§ Leadership also emerges as critical in healthcare teams but is often absent. A

single clear leader in highly complex large teams is associated with higher levels

of effectiveness and innovations

§ As teams develop and become more sophisticated in their ability to work

effectively, distributed leadership or shared leadership amongst different

functions is associated with higher levels of effectiveness, innovation and better

quality teamwork

§ Clear shared leadership is associated also with better team processes - clear

objectives, participation, emphasis on quality, support for innovation and

reflexivity.

§ But conflict over leadership is disastrous for teams. Where conflict over

leadership exists, teams are ineffective, not innovative and team processes tend

to be very poor.

The implications of our research are clear. First it is important that teamwork is

encouraged in healthcare organisations and second that leaders are trained and

encouraged to work effectively in teams. Leadership should be group-centred rather

than traditional.

The traditional approach to leadership of healthcare teams is that the leader

is responsible for the group and has control over final decisions. He or she

guards their position power and perceives the group as individuals to be

managed by the leader. The leader shapes the task for the team and

ignores the socio-emotional processes within the teams. He or she

discourages expression of needs or feelings during team meetings.

Our research suggests that healthcare teams needs group centred leaders

who see responsibility as shared by both the leader and the team; where

control over final decisions is vested in the team; where leader position

power is de-emphasised; where the leader perceives the team as a

collective entity and shares responsibility for shaping the tasks of the team.

Moreover, the team leader should emphasise and share with the group the

responsibility for maintaining the group as a social entity. He or she should

closely observe socio-emotional processes in the group and encourages

discussion in meetings of team members' needs and feelings.

§ The research reveals that multi-professional working is associated with high

levels of innovation in healthcare. Where a variety of professional groups are

involved in healthcare teams it is much more likely that innovation will be a

consequence. This is particularly so when the team processes are healthy.

§ Larger teams are seen as more innovative and effective, partly because they

have the resources, organisational structures and processes in place which

enable radical changes in the delivery of healthcare to be accomplished.

Policy makers should be cautious about how they respond to these findings.

Teams cease to exist above around twelve to 14 members. They become

small organisations. Primary health care "teams" consisting of around 40

members are in effect medium sized enterprises. As these teams transform

at around 15 to 20 members into organisations it is likely that they will divide

into sub-teams. This is a normal structural process in organisations.

Leaders and managers then have to make decisions about the appropriate

sub-teams to be formed. These should be formed taking account of the

team's context and tasks. The purpose of these sub-teams should be to

manage uncertainties in relation to particular patient groups, particular

stakeholder groups and particular illness categories.

A wise differentiation within healthcare teams is necessary as they develop

into organisations. This differentiation or specialisation of activities is

necessary for the team to cope effectively with its environment. But at the

same time it creates new pressures upon the organisation. The sub-groups

must learn to integrate effectively with one another to communicate, liase

and co-operate. The danger is that sub-teams, particularly if they are

composed of like professional groups, end up competing rather than

collaborating and co-operating for the greater good of patients. Therefore,

managers and leaders must ensure that these groups and sub-teams liase

and collaborate together. It is harder to work as a single team as the

organisation grows in size and the required differentiation and integration

processes necessary require sophisticated leadership. Leaders of health

care teams must therefore be trained in appropriate leadership knowledge,

skills and attitudes.

These observations are reinforced by our finding that higher levels of integration in

healthcare teams are associated with innovation and effectiveness. Communication,

quality of meetings and information sharing activities to increase integration are all

associated with innovation and effectiveness in healthcare teams. However, our

research also reveals that quality of communication and meetings varies

considerably. In general meetings are very poorly managed in primary health care

and professionals in this domain have much to learn from their colleagues in

community mental health teams. Indeed, another important conclusion which can be

drawn from the research is the importance of training for those working in primary

health care to better plan and manage their meetings.

Recommendations

There are a number of key elements to effective teamwork (Guzzo and Shea, 1992).

1. Conditions for effective teamwork

• First, Individuals should feel that they are important to the success of the team.

When individuals feel that their work is not essential in a team, they are less likely

to work effectively with others or to make strong efforts towards achieving team

effectiveness. Roles should be developed in ways which make them

indispensable and essential.

• Individuals' roles in the team should be meaningful and intrinsically rewarding.

Individuals tend to be more committed and creative if the tasks they are

performing are engaging and challenging.

• Teams should also have intrinsically interesting tasks to perform. Just as people

work hard if the tasks they are asked to perform are intrinsically engaging and

challenging, when teams have important and interesting tasks to perform, they

are committed, motivated and co-operative (Hackman, 1990).

• Individual contributions should be identifiable and subject to evaluation. People

have to feel not only that their work is indispensable, but also that their

performance is visible to other team members.

• Above all there should be clear, shared team goals with built-in performance

feedback. Research evidence shows that where teams are set clear targets at

which to aim and they receive feedback on their performance, their performance

is generally improved.

2. Selecting team members

Regardless of their task specialism, there are certain attributes that all team

members need to demonstrate if the team is to achieve its goal. Selection should

focus not just on professional skills but also on knowledge skills and attitudes (KSAs)

for teamworking. (See Figure 13).

Figure 13: Knowledge, Skills and Abilities for Teamworking

A Conflict resolution KSAse.g.

Fostering useful conflict, whileeliminating dysfunctional conflict.Using integrative (win−win) strategiesrather than distributive (win−lose)strategies.

B Collaborative problem solving KSAse.g.

Having the right level of participationfor any given problem.Avoiding obstacles to team problemsolving (e.g. domination by some teammembers).

C Communication KSAs e.g. Employing communication patternsthat maximise an open flow.Using an open and supportive style ofcommunication.Using active listening techniques.Paying attention to non-verbalmessages.

D Goal-setting and performance Management KSAs

. e.g. Setting specific, challenging andattainable team goals.Monitoring, evaluating and providingfeedback on performance.

E Planning and task co-ordination KSAse.g.

Co-ordinating and synchronising tasks,activities and information.Establishing fair and balanced rolesand workloads among team members.

Source: Stevens & Campion, 1999 (Reprinted with permission).

3. Team composition

Today’s health care teams are being formed to respond to increased complexity and

demands in the environment; and they bring together people from diverse

professional backgrounds. Such diverse teams therefore embody different attitudes

and working practices as a result of differences in age, gender, educational

background, nationality, organisational culture, etc.

There is growing evidence, supported by the results from our research, that teams

that are diverse in terms of skill and educational specialisation produce high levels of

innovation and clear strategic thinking.

To date however it seems that other forms of diversity produce less positive effects

on team performance. Teams whose members have diverse cultural backgrounds

initially tend to perform more poorly than culturally homogenous teams, although this

effect diminishes over time. Turnover rates are higher in teams which are diverse in

terms of demographic features such as age, educational level, status and non-

industry work experience.

Diverse teams are not only advantageous if team performance is to be maximised: in

current health care settings, it is simply not practical to select teams that are highly

homogeneous. The challenge is to achieve the positive effects of diversity whilst

building stable teams that will grow and develop together. It is essential therefore to

provide induction and training for individual team members which will minimise the

impact of differences that can be disruptive.

Susan Jackson (1996) has drawn a distinction between team-member differences

that are task related (for example, educational level, work specialisation,

organisational function) and those that are relations-oriented (e.g. gender, age,

nationality, political views). It would seem that difficulties in teams are more often, in

the short term at least, related to relations oriented-differences. In these areas,

individuals will tend to make shallow or stereotyped decisions about others. Both

awareness training and opportunities for social or informal contact between team

members therefore play an important part in breaking down stereotypical reactions

and developing more appropriate judgements.

4. Team Leadership

• Leadership is creating alignment around shared objectives and strategies to

attain them.

• Leadership is increasing enthusiasm and excitement about the work and

maintaining a sense of optimism and confidence.

• Leadership is helping people appreciate each other and helping them to learn

how to confront and resolve differences constructively.

• Leadership is helping people to co-ordinate activities, continuously improve,

develop their capabilities, encourage flexibility, encourage an objective analysis

of processes, and foster collective learning about better ways to work together.

• Leadership is representing the interests of the group or organisation, protecting

its reputation, helping to establish trust with external stakeholders and helping to

resolve conflicts between internal and external partners.

• Leadership is creating a unique group or organisational identity.

Organisations that introduce team based working stress the importance of selecting

the right people to lead teams at the very beginning. Leaders who find it difficult to

move from a directive/controlling supervisory role to one of participative leadership

can cause lasting problems. Those organisations which have successfully overcome

this difficulty actively encourage all members of staff to apply for team leader

positions. This begins the process of breaking down stereotypical thinking about who

can lead teams, the criteria for application, etc. In addition, these organisations

provided considerable training and support for new team leaders in the initial months

of their appointment.

In the early stages of a team’s development, training should be made available for

team leaders to enable them to design and implement appropriate team processes

and to develop the skills needed for effective team leading.

Team leaders need to be skilled in responding appropriately to meet the needs of

their teams, i.e. to be more or less directive in supporting a team. Their aim should

always be to move as quickly as possible away from being directive and towards

allowing the team to be autonomous. When supported by a team leader who

provides an autonomous environment, a team can achieve more highly by becoming

self-directing in its development and its work.

5. Organisational Support for Teams

Hackman and his colleagues at Harvard University have concluded that there are six

principal areas within which teams need organisational support: targets, resources,

information, education, feedback and technical/ process assistance in functioning.

Examining the extent to which organisations provide team support in these areas can

help in discovering the underlying causes of team difficulties.

Targets

Teams need support from an organisation in determining targets or objectives.

Surprisingly few health care teams are given clear targets by their organisations often

because organisational targets and aims have not been clarified sufficiently. It is

striking, when team members are asked to outline their objectives and team targets,

how few have clear notions of what is required of them. There is an implication that

teams should derive their targets and objectives by scrutinising the organisational

objectives or mission statements. However, these are often such vague good

intentions or positive but abstract sentiments that it is almost impossible for a team to

derive clear targets and objectives. Where, through a process of negotiation, teams

are able to determine their targets in consultation and collaboration with those

hierarchically above them, there is usually a better level of performance.

Resources

The organisation is required to provide adequate resources to enable the team to

achieve its targets or objectives.

Resources include: having the right number and skill mix of people; adequate

financial resources to enable effective functioning; secretarial or administrative

support; adequate accommodation; adequate technical assistance and support (such

as computers, blood pressure testing equipment, or appropriate equipment for testing

infants' hearing, etc).

Information

Teams need information from the organisation which will enable them to achieve

their targets and objectives. Changes in strategy or policy which are not

communicated to teams can hamper their effective functioning. Ensuring that

relevant information reaches a team to enable it to perform effectively is an essential

component of an organisation's management. For example, GPs need to provide

health visitors with ready access to age/ sex registers, medical records and other

information about the practice population, in order for the health visitors to function

effectively within the teams.

Education

Part of an organisation's responsibilities for effective team functioning is to provide

the appropriate levels and content of education for staff within teams. The purpose of

such training and education is to enable team members to contribute most effectively

to team functioning and to develop as individuals. This includes on-the- job training,

coaching via supervisor, training courses, residential training courses or distance

learning courses. There should be adequate access to training which is relevant to

the team's work and of a sufficient quality and quantity to enable them to perform to

maximum effectiveness. And, as indicated above, team members should be trained

in the knowledge, skills and abilities, for team working.

Feedback

Teams require timely and appropriate organisational feedback on their performance if

they are to function effectively. Timely feedback means that it occurs as soon as

possible after the team has performed its task, or occurs sufficiently regularly to

enable the team to correct inappropriate practices or procedures. Appropriate

feedback means that it is accurate and gives a clear picture of team performance.

For some teams it is difficult to gain accurate feedback. For example, primary health

care teams have almost no feedback at all. For a team responsible for providing

training in one division of, say, a major oil company, organisational feedback might

take the form of senior managers' satisfaction with improved performance. This could

include measuring the results from technical training courses in customer service in

retail outlets (i.e. filling stations). Such information could come from surveys of

customer satisfaction with retail operators' services. Clearly there are large

differences in the extent to which organisations can and do provide feedback to

teams, but the aim should be for the organisation to improve continuously in the

extent to which it provides useful, accurate and timely feedback to teams.

Technical and process assistance

Organisations have to provide the specialised knowledge and support which will

enable teams to perform their work effectively. A primary health care team engaged

in developing its practice objectives, by identifying the health needs of the practice

population, might need the health authority to deploy a community medical officer to

advise the team on patterns in local health and ill-health. For a training team in an oil

company, technical assistance might take the form of specialist computing experts

and marketing strategists, advising the company on how to communicate most

effectively to managers throughout Europe, in order to market their training courses

to managers in different functions.

Process assistance refers to the organisational help available when team process

problems are encountered. Are consultants and facilitators available to help the team

identify, diagnose and overcome problems of team functioning from time to time?

But the implication of this work is that NHS organisations should not simply create

teams. They should recreate themselves as team-based. We now turn to address

this important issue.

6. Developing team-based organisations

Teams working within team-based organisations have more discretion and scope

than those working within traditionally managed organisations. In practice, team-

based organisations reflect a management philosophy that incorporates certain

fundamental principles.

In team-based organisations, most employees are clear about and committed to the

objectives of the organisation as a whole. Senior management take time to

communicate information to all employees about organisational objectives and also

encourage team members and teams to influence the development of organisational

objectives.

In team-based organisations employees are more fully involved. They are

encouraged to contribute ideas, opinions and information to decision-making

processes, and their teams have influence over decisions that are made. The

organisation as a whole promotes acceptance of and commitment to processes of

debate about how to perform work most effectively.

Managers within team-based organisations are committed to encouraging

constructive debate within the organisation. They listen carefully to the views of team

members and take time to explore diverse views and differences of opinion. They

also encourage the expression of minority points of view and value opportunities for

careful discussion about the best ways of delivering products and services.

In team-based organisations there must necessarily be a climate supportive of

creativity and innovation. Teams are hothouses for creative ideas, and the

organisation must encourage the expression and implementation of ideas for new

and improved health care processes and ways of working. If it fails to do this, both

the impetus for and the value of team-based working are lost.

To ensure the achievement of these aims, team-based organisations must reflect the

belief that organisational goals will largely be achieved not by individuals working

separately but by groups of people who share responsibility for outcomes and who

work in efficient and effective teams.

In traditional organisations, there tend to be individual command structures with

various status levels representing particular points in the hierarchy. There are

supervisors, managers, senior managers, assistant chief executives and so on. In

team-based organisations, the structures are collective. Teams orbit around the top

management team or other senior teams, both influencing and being influenced

rather than being directed or directive. The gravitational force of different teams

affects the performance of the teams around them. This is a flexible, fluid structure in

contrast to the mechanical, hierarchical structure of traditional organisations.

In traditional organisations, the manager monitors the performance of employees. In

team-based organisations, the team monitors the performance of members within the

team and the team as a whole is appraised by those it provides services and

products for. Thus the Human Resource Management team may be appraised by all

of the teams within the organisation for which it provides services.

In traditional organisations, power is invested in the hierarchy. The further up the

hierarchy you go, the more power you find located there. In team-based

organisations, the emphasis is on integration between teams and on reducing the

number of levels in the organisation so that there is less vertical difference between

different teams and groups. Whereas in traditional organisations the emphasis is on

maintaining power and control through the use of a clear hierarchy of command

(which may be important for example in an organisation dealing with crises), in team-

based organisations the emphasis is on achieving shared purpose across teams and

achieving shared understanding and integration across teams.

In traditional organisations the emphasis is on stability and keeping things the same.

Rules and regulations, formalisation and bureaucracy encourage uniformity and

control. In team-based organisations the emphasis is on encouraging innovation,

change and flexibility in order that the organisation can adapt appropriately to its

changing environment and be innovative health care services.

Traditional organisations tend to adopt ‘one best way’ and to seek for universal

models of effective organisational functioning. The team-based organisation

emphasises its uniqueness, adopts ways of working that are appropriate to the

organisation in its current circumstances, environment and economic context, and

adapts as the environment changes.

In traditional organisations, managers manage and control; whereas in team-based

organisations, the teams are self managing and take responsibility for setting their

(perhaps in consultation with senior managers) and monitoring the effectiveness of

their strategies and processes. Changes in the process of achieving the team-based

organisation are therefore deep, wide and pervasive.

In the face of the inevitable complexities within organisational environments, within

teams themselves and between the people who constitute those teams, there are no

simple prescriptions for implementing effective team based working. In order to be

effective, team members must therefore learn to reflect upon, and intelligently adapt

to, their constantly changing circumstances as the team develops.

There are however certain areas where problems can be predicted and where

effective initial design greatly improves the chances of success. Many of the

common problems in the introduction of team-based working (TBW) result from

impatience: effective TBW takes time to implement and requires multiple changes

that create almost inevitable difficulties. Long-term benefits can only be achieved

through persistent and consistent action in each of these three key areas:

• organisational context

• team structure

• team processes

Organisational context

The top management team’s level of commitment towards TBW is a key factor in an

organisation's introduction of TBW. But the attitude towards TBW amongst

employees generally is another powerful issue under this heading. Reward systems

that focus on competition between individual employees for bonuses undermine the

introduction of TBW. Similarly, information systems that are characterised by secrecy

rather than openness will impede the implementation of TBW. The training and

education priorities of the organisation must also be geared towards developing the

knowledge, skills and abilities required for TBW, including leadership skills and

teamworking skills. The organisation will also need access to coaching expertise to

support teams both during their development and when they experience difficulties in

the course of their work (such as conflicts between team members), either provided

by someone within the organisation or an outside consultant.

Team structure

Teams structure refers to the composition of the teams, i.e. who will be the team

members. This is not simply a matter of the skills required to perform the task, but

also raises questions of variety in functional background and balance in demographic

characteristics such as background culture, gender, age and even personality.

A key aspect of team structure is the nature of the task that the team is required to

do. The goals should be clear, the task should be motivating and team members

should have clear feedback on how effective their performance has been. It also

refers to effective team leadership, as we have emphasised above, and the need to

appoint team leaders who know how to lead teams and are not hierarchical,

traditional supervisors.

Team processes

When TBW is introduced, most organisations focus initially on team processes and

send prospective team members on team-building workshops. Though the motive

behind this is valid, i.e. to build cohesion, the first step in building effective teams is to

ensure that team members:

• make sufficient effort (that they are motivated to perform the task).

• have adequate knowledge and skill within the team both to perform the task and

to work in a team.

• have developed appropriate ways of performing their task, i.e. task performance

strategies.

As organisations implement TBW, there are major pitfalls in each of these three key

areas that must be avoided, any of which could considerably delay or impede the

process.

• the creation of teams throughout the organisation, regardless of the need or the

nature of the tasks.

• setting up teams but continuing only to appraise, reward and manage individuals.

• creating teams but neglecting to train people to function effectively within and

across teams.

• introducing TBW while leaving teams without expert assistance when problems

such as major conflicts arise.

• creating well functioning teams but ignoring the vital need to ensure these teams

communicate with each other, integrate their work and otherwise liaise

effectively.

• failing to negotiate with the teams clear and challenging team-level objectives.

• giving the teams challenging objectives but not the training, skills and resources

to meet those objectives.

Conclusion

The following quotation illustrates just how fundamental team working is to our

species and we include this to remind the reader of the importance of groups and

teams to human societies throughout their development. There is little new about

teamwork.

"He makes tools (and does so within more than one technical tradition),

builds shelters, takes over natural refuges by exploiting fire, and sallies out

of them to hunt and gather his food. He does this in groups with a

discipline that can sustain complicated operations; he therefore has some

ability to exchange ideas by speech. The basic biological units of his

hunting groups probably prefigure the nuclear family of man, being founded

on the institutions of the home base and a sexual differentiation of activity.

There may even be some complexity of social organization in so far as fire-

bearers and gatherers or old creatures whose memories made them the

data banks of their 'societies' could be supported by the labour of others.

There has to be some social organization to permit the sharing of co-

operatively obtained food, too. There has to be some social organization to

permit the sharing of co-operatively obtained food, too. There is nothing to

be usefully added to an account such as this by pretending to say where

exactly can be found a prehistoricial point or dividing line at which such

things had come to be, but subsequent human history is unimaginable

without them."

[Extract from J.M. Roberts (1995), The History of the World, page 18].

The activity of a group of people working co-operatively to achieve shared goals via

differentiation of roles and using elaborate systems of communication is basic to our

species. The current enthusiasm for team working in and in health care reflects a

deeper, perhaps unconscious, recognition that this way of working offers the promise

of greater progress than can be achieved through individual endeavour or through

mechanistic approaches to work. That is what this report has demonstrated in

relation to health care teams to and quality and innovation in patient care.

Appendix I

Measures Used

Primary Health Care Team Questionnaire

Section 1 Team working. This contained seven measures of team working. Five of

these form the Team Climate Inventory (Anderson and West, 19xx): participation, a

15 item scale covering information sharing (α = ); innovation, an eight item scale

covering support from new ideas (α = ); team objectives, covering clarity and

relevance of objectives (α = ); task style, covering the monitoring and appraising of

work in the team (α =). Three other measurers were included: reflexivity, covering

the reviewing process in the team (α = ) and XXX was measured using XX (xx

19xx) (α = ); and team innovation measured using (West?) (α = ). Respondents

were also asked to list the major changes introduced by the team in the previous 12

months.

Section 2 included measures of team effectiveness adapted from Poulton and West

(199x). This includes three dimensions: team working (α = ); , patient orientation (α

= ); and organisational efficiency (α = ).

Section 3 included a measure of psychological stress, the GHQ-12 (Goldberg, 1991)

(α = .88 ).

Section 4 included questions eliciting biographical and team information (e.g. age,

gender, ethnic origin, job title, employer, team composition, team leader).

Initial construction of the effectiveness measure

Effectiveness criteria were generated using an iterative process within the

constituency model approach (Connally et al., 1980). After consultation with the local

Health Authority and Community Mental Health Trust, the Department of Health, and

local community mental health teams, representatives of the range of stakeholders in

the provision of mental health care were invited to a one-day workshop. A total of 13

interest groups were approached: users, carers, advocacy agencies, mental health

charities, consultant psychiatrists, community mental health nurses, occupational

therapists, psychologists, social workers, managers, policy makers, researchers in

mental health, and general practitioners. All groups were represented by the 50

people attending the workshop.

The aim of the workshop was to provide the basis for an agreed set of definitions of

effectiveness in CMHTs. Stakeholders sharing a perspective were grouped together,

so that consensus could be achieved more easily within each working group.

Participants were asked to generate a set of criteria which they agreed would

measure CMHT effectiveness, with the proviso that any criterion must be supported

by concrete examples of how good practice could be audited. Groups then

reconsidered their criteria in order to prioritise aspects of practice.

Output from the workshop was analysed and categorised by the research team.

Duplication and ambiguity were removed. The 76 remaining criteria, grouped into the

three broad categories of user and carer issues, team development and viability, and

organisational issues, were re-circulated to all workshop participants. They were

invited to comment on wording and clarity, to suggest modifications, indicate

significant omissions, and approve priority ratings. Items rated as less important by a

majority of respondents were removed, and any items rated down in the final

consultation were respositioned. Remaining ambiguities and duplication were

removed, together with items already covered in other sections of the proposed

survey questionnaire.

The reduced set of CMHT effectiveness criteria, together with supporting measures

of good practice, were piloted amongst local community mental health nurses, a

CMHT, psychologists and other mental health professionals. Practitioners agreed

that they captured the complexity of the work and the diversity of environments in

which CMHTs operate. The final set of 27 criteria was incorporated into the main

survey questionnaire, along with the Team Climate Inventory (Anderson & West,

1994) and the 12-item General Health Questionnaire (Goldberg, 1970.)

Each statement was clarified by additional concrete examples of elements of practice

which individuals could use to aid their rating. A 5-point Likert-type scale was used

to rate how effective the team was on each criterion. Thus, to illustrate, the content

of the first criterion was

Accessibility of the service to users and carers has been identified as a measure of CMHT effectiveness.

[For example: identification and contactability of a key worker; clear referral procedures; time taken to respond to users and or carers; a clear point of access.]

Not at all To a great extent Overall, to what extent does your CMHT make services accissible to users and carers? 1 2 3 4 5

Appendix I I

Knowing the Way:

Effectiveness in Primary Health Care

A description of national workshops aimed at defining

effectiveness criteria for primary health care

Introduction

The World Health Organisation define primary health as:

“..essential care based on practical, scientifically sound and acceptable methods and tehcnology made

universally accessible to individuals and families in the community through their full participation and

at a cost that the community and the country can afford to maintain at every stage of their development

in the spirit of self reliance and self determination. It forms an integral part of the countries health

care system, of which it is a central function and main focus, and of overall social and economic

development of the country. It is the first level of contact of individuals, the family and community with

the national health system, bringing health care as close as possible to where people live and work”.

World Health Organization, 1978.

In order to enable primary health care professionals and the populations they serve to develop the

health and stress of those populations, it is important that clear objectives for primary health care are

established. Moreover, what constitutes effectiveness in primary health care also has to be determined

by each primary health care team or organization. Indicators of effectiveness, once established,

provide professionals with clear guidelines over how best to allocate resources in order to achieve

effective primary health care. This consultative document represents the endeavours of 63

professionals working in the area of primary care who spent four days in workshops designed to help

achieve this overall aim. 10

The vision of this exercise is to promote health care for the population nationally, by providing primary

health care professionals and others concerned with the health and stress of the population with clear

indicators of effectiveness. These can be used as a basis for discussion and development in primary

health care teams across the country, taking into account their local circumstances, philosophies of

primary care and the needs and views of their local populations.

Effectiveness in Primary Health Care

In an important analysis of the evaluation of health services’ effectiveness, St Leger, Schneiden &

Walsworth-Bell (1992) observe that “Surprisingly, routine data [sources relating to primary care] are

relatively sparse, especially when one considers that the majority of contacts that the general public

have with the health service are with general practice.” (p.41). Indeed, it is striking how little research

has focused on developing theoretically grounded or practically useful indices of effectiveness of

primary health care services. In a rare exception, Pearson & Spencer (1995) employed a two-stage

Delphi questionnaire to determine agreed indicators of effective teamwork in primary care. Using

responses from 137 people involved in primary care teamwork - primarily from FHSAs - they rated the

importance of twenty indicators. Four emerged as particularly significant:

10 Within primary health care, terminology is a source of concern and conflict. For example, someprefer the term ‘primary care’. There are sensitivities around the terms ‘general practice’ and ‘primaryhealth care’ being used interchangeably; and a distinction between medical and nursing care is alsosensitive. Those who attend for treatment or advice are called ‘patients’ by some and ‘clients’ byothers. We have preferred the terms ‘primary health care’ and ‘clients’ in this document.

• Agreed aims, goals and objectives

• Effective communication

• Patients receiving the best possible care

• Individual roles defined and understood

What is effectiveness?

How can the effectiveness of primary health care teams therefore be judged? At the simplest level,

effectiveness might be viewed as the influence the primary health care team has in improving health

and ameliorating ill health within the practice population. Such a criterion of effectiveness begs a

number of questions, however. Examples include:

• Is the population in an area of social deprivation?

• Does the team have unusually good resources, in terms of number of staff or technical equipment?

• Will the effects be long-term?

In the literature on organizational effectiveness, a distinction is made between efficiency (doing things

right) and effectiveness (doing the right things) (Sundstrom, DeMeuse & Futrell, 1990). Efficiency

may be defined as the output for a given input and how a team compares with other similar teams in

this regard. Effectiveness can also be seen as the team’s capacity to perform, adapt, maintain itself and

grow (where growth may refer to size, innovation or skill development) (Goodman, 1986).

Clearly, teams can be seen as more or less effective depending upon the criteria adopted.

Consequently, the assessment of team effectiveness has come to be seen as much a political as an

empirical process. Below we consider (briefly three examples of approaches to managing this problem.

The constituency approach

Recognition that effectiveness is a political concept has led to the use of the ‘constituency approach’

(Connally et al, 1980; Bedeian, 1986) which seeks to incorporate all significant views in the judgement

of team effectiveness. Each of the major constituents is identified (e.g. clients, carers, staff health

authorities, professional organizations) and the effectiveness criteria they would use are adopted as

indicators. Effectiveness is then measured using multiple indicators rather than an aggregate, since, in

many cases, effectiveness in one area will necessarily imply ineffectiveness in another (consumer

satisfaction may not always coincide with quality of care if consumers require prescriptions for drugs,

the use of which is not in their best interest).

From this conceptual, background Poulton & West (1994) developed a set of 23 effectiveness criteria

for primary health care teams using a focus group methodology involving multiple stakeholders in

primary care. In a study involving more than 500 practice nurses (Poulton, 1995) the criteria were then

factor analysed revealing four underlying factors: good teamwork, task excellence, organizational

effectiveness and patient-centred care. More recent analyses with larger samples suggest only three:

quality of health care, organizational efficiency, and teamworking.

Poulton and West (1997) employed these outcome measures in a six month longitudinal study of the

impact of teamworking on effectiveness. The research demonstrated that clarity of and commitment to

objectives was the most important and statistically significant predictor of effectiveness in all four

areas.

Multidisciplinary auditAnother important approach to measuring effectiveness is multidisciplinary audit. A major centre for

research and advice is the Eli Lilley National Clinical Audit Centre (Hearnshaw, Baker & Robinson,

1994; Baker, et al 1995). In a study conducted by this Centre of three practices, those supported in the

development of multidisciplinary audit showed a significant improvement in specific areas of

functioning. The procedure involves teams identifying particular problems (in a diagnostic area or an

area of team functioning) and a priority is identified. The team then sets standards in the specific area,

observes current practice and achievements, and compares these with the standards. Discrepancies

prompt changes in practice and the results are then regularly reviewed. Such an approach clearly

enables teams to assess and improve performance in specific areas, although the approach is somewhat

atomistic and relies on the effective identification of priorities. An excellent overview of this approach

is provided by Crombie, Davies, Abraham & Florey (1993).

ProMES

In the broader organizational literatures on team effectiveness, a widely adopted approach is the

Productivity Measurement and Enhancement System (ProMES) based on research by Naylor, Pritchard

& Ilgen (1980). Effectiveness criteria are established in group discussions with team members and

supervisors. The variables are then “psychologically scaled” to a common effectiveness scale. Based

on group consensus about expected levels of effectiveness, which are given a zero value, maximum

effectiveness levels (set at +100), and minimum levels (-100) are set. Each variable is also weighted in

terms of its perceived contribution to the overall effectiveness of the team or organization. The system

is then used to set objectives, develop indicators monitor and improve performance and give feedback

to the team (Pritchard, 1990). This is promising for primary health care, because of the sophistication

of the approach, its theoretical robustness and practical utility in complex contexts. It was therefore

used as a basis for developing effectiveness measures by the Health care Team Effectiveness project.

Overall, however, it is clear that conceptual and empirical development in evaluating the effectiveness

of primary health care is urgently required, if cost, clinical and community value are to be convincingly

demonstrated. In order to take this forward a series of four workshops on primary health care team

effectiveness was sponsored by the Institute of Work Psychology. The methods are described below.

Prior to detailing these we consider the need for a manageable set of objectives or effectiveness

indicators; and the need for a theoretical framework within which to locate any set of dimension.

(i) The need for relative simplicity

It is apparent from any analysis of research in the domain of effectiveness that the development of

indicators and effectiveness dimensions for primary health care is a complex task (West, 1996). A

major problem is that many measures of effectiveness and many indicators can be developed. Trying

to use this large number of measures within an organization as an effective means of targeting

resources simply becomes overwhelming for the practitioners concerned. The sheer multiplicity of

potential indicators is cognitively too complex for people within the organization to cope with the task

effectively. Indeed, some research suggests we can only cope with 7 plus or minus 2 categories,

whereas other organizational researchers specifies to more than 8 to 12 key dimensions. Consequently,

we sought a parsimonious theoretical model which would provide guidance for researchers and

practitioners.

(ii) The Competing Values Model

The most useful model appears to be the Competing Values Model. This model incorporates two

fundamental dimensions;

• flexibility of the organization versus control within the organization;

• external orientation versus internal orientation (see figure 2).

M e e t i n g t a r g e t s

E m p h a s i s o n p e r f o r m a n c e

‘C u s t o m e r ’ s a t i s f a c t i o n

C o m p e t i n g V a l u e s M o d e l

H u m a n R e l a t i o n s M o d e l O p e n S y s t e m s M o d e l

I n t e r n a l P r o c e s s M o d e l R a t i o n a l G o a l M o d e l

T r a i n i n g

T e a m d e v e l o p m e n t

L e a r n i n g o r g a n i s a t i o n

T e a m m e m b e r m e n t a l

h e a l t h

C o m m i tm e n t &

s a t i s f a c t i o n

S c a n n i n g o f t h e

e n v i r o n m e n t

R e s o u r c e a c q u i s i t i o n & c o -

o r d i n a t i o n

I n n o v a t i o n

C o l l a b o r a t i o n w i t h o t h e r

o r g a n i s a t i o n s

E f f i c i e n t c o n t r o l o f

r e s o u r c e s

C l i n i c a l a u d i t

R e v i e w i n g & e v a l u a t i n g

e f f e c t i v e n e s s

B u d g e t m a n a g e m e n t

C o n t r o l

I n t e r n a l

O r i e n t a t i o n

E x t e r n a l

O r i e n t a t i o n

F l e x i b i l i t y

F i g u r e A . 1

Combining these dimensions identifies 4 domains of effectiveness:

human relations,

open systems,

rational goal,

internal process

• Human relations model

The primary emphasis is on norms and values associated with belonging, trust, respect, skill

development, growth and stress. Motivational factors are attachment, cohesiveness and

organization membership. Areas of effectiveness include:

• development of skills

• team development

• learning organization skills

• team member mental health

• commitment and satisfaction

• Open systems model

The primary emphasis is on change and innovation and appropriate scanning of the environment.

Norms and values are associated with knowledge of the environment, resource acquisition,

innovation and adaptation. Motivating factors are:

• development of services

• resource acquisition and co-ordination

• innovation

• collaboration with other organizations

• Rational goal model

The primary emphasis in this model is on the pursuit and attainment of well-defined objectives. Norms

and values are associated with good performance, goal-fulfilment and achievement. Motivators

are: successful achievement in pre-determined areas. Effectiveness dimensions include:

• meeting ‘production’ targets

• ensuring high quality

• high client or customer satisfaction

• Internal process model

The emphasis here is on stability, internal organization and adherence to rules and protocols,

where norms and values are associated with efficiency, co-ordination and uniformity. Motivating

factors are needs for order, rules, regulations and efficiency. Effectiveness dimensions include:

• efficient control of resources

• reviewing and evaluating effectiveness

• good budget management

The model of competing values stresses how the allocation of resources to any one area, for example

the open systems model, is likely to lead to a restriction of resources in the opposite domain (internal

process). Similarly a focus on external control (rational goals) may well lead to a neglect of the area of

human relations (internal flexibility). Effectiveness in organizations, is likely to be maximised when

emphasis is placed equally in each of the four domains of effectiveness. Using this framework as a

guide we embarked on a series of 4 workshops to clarify our thinking about effectiveness in primary

care.

The Workshops

MethodThe methodology used to develop effectiveness measures for primary health care was developed taking

account of the following: the importance of including the complexity and diversity in primary health

care and of taking into consideration the wide range of views and perspectives held by the professional

groups who contribute to meeting patient needs; the need to develop a method which would most cost-

effectively use the time of primary health care professionals; and the importance of developing

measures which were generalisable across primary health care.

The measures were therefore developed in two stages:

i) Indentifying objectives for primary health care and developing indicators for these objectives, was

carried out in four workshops with domain relevant experts from primary health care, based on an

approach developed by Connolly et al, 1990. Such an approach enabled the views of a range of

professionals to be taken into account. Working intensively with an expert group enabled considerable

knowledge and insight to be gained in a short space of time. In addition, these professionals, who had

a background in primary care but were mainly working in an advisory, policy or research role,

provided a broader, more generalisable perspective on effectiveness in primary health care.

ii) The measures developed were used by primary health care teams teams and feedback provided on

effectiveness.

Workshops

Objectives

There were three principle overall objectives for the workshops:

♦ to identify the important issues relating to developing effectiveness measures for primary health

care,

♦ to develop a set of effectiveness measures acceptable to all perspectives in primary health care,

♦ for participation in the workshops to be a valuable experience for the participants.

Participants

An initial stakeholder analysis identified 13 stakeholders in primary health care. Advise was sought

from contacts in primary health care about key experts who could represent the views of each

stakeholder group, and about whether the initial list of stakeholder was sufficiently comprehensive.

The experts suggested by the contacts were sent information about the research programme, invited to

attend the four workshops, and asked to suggest additional or alternative key experts who could also

make a contribution. In addition, representatives from primary health care teams were invited, that is,

professionals who were currently engaged in clinical practice. The majority of those contacted were

keen to attend the workshops, and were able to commit themselves to attending two or three. X were

able to attend all four. Each workshop was planned so that the participants covered the main

stakeholder views.

Workshop process

A focus group methodology was used.

The delegates were divided into three working groups. These were designed so that (a) a range of

stakeholder views were represented, and (b) one or two or group members had attended most or all of

the workshops and so could share with new members the learning and experience from previous

workshops. Each group worked with a facilitator, trained in ProMES, and a notetaker who recorded

the group discussion and the decisions made.

Workshop 1

Objective: to develop objectives for primary health care.

After an initial introduction to the Health Care Team Effectiveness project and a presentation on

ProMES, the delegates were presented with a set of objectives for primary health care developed by the

researchers. Each group worked on (a) refining the objectives, (b) critically evaluating the objectives

in relation to the criteria for objectives (see appendix xx).

The outputs from each group were presented at the end of the workshop in a plenary session. After the

workshop the outputs were discussed with members of four primary health care teams (who endorsed

their relevance and value), combined into a single list and then circulated to delegates.

Workshops 2 and 3

Objective: to develop indicators for the objectives for primary health care.

Both workshops started with a presentation on team working in primary health care and issues relating

to the development of effectiveness measures.

Delegates were presented with the final version of the objectives for primary health care. Each group

worked on (a) developing indicators for an objective, (b) critically evaluating the indicators in relation

to the criteria for indicators (see appendix xx).

In Workshop 2 the working groups selected the objective to discuss. The objectives - Quality of Care

and Client Satisfaction were selected. In Workshop 3 groups were assigned an objective so that each

was discussed at least once. The objectives - Effective Management of Resources, Development and

Satisfaction of Primary Health Care Team Members and Quality of Care, were discussed. In the third

workshop the output from each working group was given to another group in the afternoon session for

discussion and refinement.

The output from each group were presented at a plenary session at the end of both workshops.

After Workshop 2 and 3 the outputs were amended and the circulated to delegates. In addition, after

the third workshop the outputs from all three workshops was written-up in the document ‘Knowing the

Way: Effectiveness in Primary Health Care’ and circulated to delegates.

Workshop 4

In the fourth Workshop, those attending critically analysed the objectives and indicators developed, and

considered how they could be applied in practice by PHC teams and others. A review of the data

derived from 100 teams, examining their definitions of effectiveness was also presented. The focus

groups commented on the next steps in taking forward the work completed to date.

1. Objectives were identified in the first of the four workshops and after an additional three

workshops with somewhat differing attendees who worked with them, they remained unchanged.

(The point is they have been tested and found acceptable by lots of different people.)

2. In addition, they were shown to a number of primary health care teams. These teams found the

objectives useful and accurate.

3. The themes in these objectives are similar in principle to the themes of objectives that other types

of professional organizations and in other settings have developed. Thus, there is some consensual

validation.

Bearing in mind the problems of deluging primary health care teams/organizations with long and

unmanageable lists of objectives and indicators of effectiveness, we endeavoured to produce a core list

of key objectives

Below we describe each and suggest possible indicators.

Effectiveness measures are a guide not stone tabletsThe key objectives identified and the indicators developed for each are a synthasis of the outputs from

the four workshops. Some of the contributions made by the working groups have been modified by the

researchers, and the Competing Values Model used to provide a conceptual structure. The majority of

objectives and indicators described below, however, were suggested and critically reviewed by the

Primary Health Care professionals who attended the workshops. The objectives and indicators are not

definitive, nor are they comprehensive and applicable to every primary health care team. They are

meant to be used as an aid for teams which wish to measure and enhance their effectiveness. In

relation to each objective, the focus groups developed a set of indicators by which progress towards

objectives could be measured. Again, these indicators are meant as examples. If they fit and make

sense to individual primary health care teams, fine; but they may well not fit. Primary health care

teams have different missions and the measurement must be tailored to that mission. Moreover, the

focus groups produced many more possible objectives and measures of them are described below.

These can be used as a resource for teams wishing to explore areas beyond those we have designated as

likely to be core.

Objectives and Indicators for Primary Health

The core key objectives developed in the workshops are showh in Figure A2, mapped

in to the competing value model.

* Im p r o v i n g h e a l t h

* H i g h q u a l i t y o f h e a l t h c a r e

C l i e n t s a t i s f a c t i o n

C o m p e tin g V a l u e s M o d e li n P r im a r y H e a lth

C a r eH u m a n R e la t io n s M o d e l O p e n S y s t e m s M o d e l

I n t e r n a l P r o c e s s M o d e l R a t io n a l G o a l M o d e l

G o o d T e a m w o r k i n g

C o n t i n u i n g

p r o f e s s i o n a l

d e v e l o p m e n t

H ig h t e a m m e m b e r

c o m m itm e n t &

s a t i s f a c t i o n

A c c u r a t e i d e n t i f i c a t i o n o f

h e a lth n e e d s

R e s p o n s i v e n e s s t o c l i e n t s

a n d c o m m u n i t y

E f f e c t i v e c o l l a b o r a t i o n w i t h

o t h e r o r g a n i s a t i o n s

E f f i c i e n t u s e o f

r e s o u r c e s

H e a lth c a r e r e v i e w i n g &

im p r o v i n g e f f e c t i v e n e s s

C o n t ro l

I n t e r n a l

O r i e n t a t i o n

E x t e r n a l

O r i e n t a t i o n

F le x ib i l i ty

F ig u r e A . 1

* In d i c a t o r s m a y d e p e n d o n t h e h e a lth c a r e p h i l o s o p h y o f t h e p r i m a r y h e a l t hc a r e t e a m s e .g . h o l i s t i c , p r e v e n t i v e , b i o m e d i c a l

The indicators developed in the workshops for each of the objectives are listed below.

♦ Improving health

♦ High quality of health care

♦ Improving client satisfaction

♦ Efficient use of resources

♦ Reviewing and improving health care effectiveness

♦ Good teamworking

♦ Continuing professional development

♦ High team member commitment, stress and satisfaction

♦ Accurate identification of health needs

♦ Responsiveness to clients and community

♦ Effective collaboration with other relevant organizations

Rational Goal

Objectives Example Indicators

Improving health ⇒ Improvement in the health of the practice poulation including

reductions in e.g. coronary heart disease, smoking, mental health

problems.

⇒ Percentage of clients improving at the expected rate after treatment.

⇒ Effectiveness of preventive practice in reducing specific treatment

requirements.

High quality of health care ⇒ Effective knowledge of and management of chronic diseases (e.g.

diabetes, epilepsy, asthma) measured by conformance with evidence-

based good practice.

⇒ Effective health education and preventive health care programmes.

⇒ The PHCT holds regular meeting to review a sample of cases. This

review would include the appropriateness of who saw the client,

procedures, and outcomes. Percentage of cases managed entirely

appropriately, based on all staff’s views.

Improving client satisfaction ⇒ Measures of client complaints and adequacy of procedures for

complaints.

⇒ Questionnaire or telephone surveys using standardised measures*.

⇒ Measures of waiting times, satisfaction with consultations,

appropriateness of appointments.

* See Appendix III for examples

Internal Process

Objectives Example Indicators

Efficient use of resources ⇒ Percent client contact time as a percentage of total time (there is an

optimal level between extremes). Assesment of DNA’s.

⇒ Number and effectiveness of initiatives developed to help team

members use time better.

⇒ Review and evaluate budget allocation and improvements in resource

utilisation.

Reviewing and improving health

care effectiveness

⇒ Review and use of evidence-based treatment protocols (all staff).

⇒ Planned clinical audit (all staff).

⇒ Intra-team referral practices regularly reviewed and adapted (all staff).

Human Relations

Objectives Example Indicators

Good teamworking ⇒ Clear, shared objectives (partly related to health needs analysis) set

annually by the team.

⇒ Regular (at least monthly) meetings to review team objectives,

strategies, processes and procedures to coordinate sub-groups and

whole team.

⇒ Positive team/organizational climate assessed annually.

Continuing professional

development

⇒ Clear and specific written annual training and development plans

agreed for each staff member (percentage of staff covered; percentage

of development plan items completed.)

⇒ Research and development budget and plans agreed by team annually.

⇒ Access for all team members to training/development resources.

High team member commitment,

stress and satisfaction

⇒ Annual review of staff commitment, stress and satisfaction using

standardized measures*.

⇒ Mechanisms to deal with and review staff dissatisfaction, conflicts and

complaints.

⇒ Low absenteeism and staff turnover.

*See Appendix III for examples

Open Systems

Objectives Example Indicators

Accurate identification of

population health needs

⇒ Collection of practice level data (demographics, disease patters, socio-

economic patterns, activity levels); and local, regional and national

data.

⇒ Involvement of clients, community groups and other relevant

organizations in health needs analysis.

⇒ Data used to inform daily planning, and longer term strategy and

direction; setting annual objectives; identifying gaps in provision and

skill mix.

Responsiveness to clients and

community

⇒ Involvement of clients and community in team/organization decisions

concerning team objectives, strategies and processes.

⇒ Frequency, quality and usefulness of contacts between team members

and representatives of community stakeholder groups.

⇒ Extent of planning within team/organization to seek feedback from

clients and community stakeholders/opinion leaders/groups.

Effective collaboration with other

relevant organizations

⇒ Appropriate admissions to hospital (and referral rate)

⇒ Number of effective and appropriate contacts with agencies (e.g.

palliative care, social services, education).

⇒ High ratings of team/organizations on salient dimensions*.

*See Appendix III for example

Application of the effectiveness measures

General Principles

There are three levels of application of these effectiveness measures, ranging from simple through to

comprehensive.

• Simple. The simplest way of using the effectiveness measures is to use them as a basis for group

discussions in the primary health care team; for members of the team to consider the areas of

effectiveness described and how they can make use of the measures in facilitating of the

effectiveness of the primary health care team. They may also consider what other measures they

may wish to add, given their local circumstances and which of the measures are not applicable. In

other words, the simple approach is to use the effectiveness measures as a basis for ongoing

discussions about monitoring and improving the effectiveness of the primary health care team.

• Moderate. The primary health care team can use the effectiveness dimensions and indicators to

develop measures of effectiveness within their primary health care organization in relation to each

effectiveness measure (as appropriate). The team can develop measures and make action plans in

terms of how they can improve their performance in this area. Again they may wish to consider

which of the measures are applicable in their organization and which are not, and what measures

which are relevant to them are missing from document.

• Comprehensive. This application involves the use of full productivity measurement and

enhancement system, which has the following steps: (a) a design team, which includes

representatives from all groups of staff in the primary health care team, is formed; (b) in a series of

meetings the design team works with a facilitator identifying objectives and indicators for these,

where necessary consulting with other members of the team; (c) the design team develops

contingencies for each indicator, that is, determines the relative contribution that improvements on

a indicator will make to overall effectiveness; (d) the team uses the indicators and receives

feedback on performance.

Quality of Health Care

Quality of Health Care

• The PHCT would have a monthly (or more frequent) staff meeting where a sample of cases was

reviewed. This review would include the appropriateness of who saw the client, what procedures

were done, and whether that client was handled appropriately in all aspect. The measure would be

the percentage of cases which were considered as being handled appropriately. This would also be

the basis for discussion of what improvements need to be made for those specific clients and for

clients in general.

• Some PHCTs will feel a health needs analysis is valuable but do not know how to do one or how

to use it. For such a PHCT, the task of developing such an analysis could be broken down into

definable steps. E.g. get information on how to do such an analysis, decide on a plan for doing the

analysis for that particular PHCT, gather the information, put the information together into a form

that the PHCT can use to make decisions. Each of these steps would be given a time for

completion. The indicator would be the percentage of the analysis completed compared to the

anticipated time for completion.

Survey on client perceptions of health improvement after treatment. For example, each client is

given a questionnaire or a sample of clients are called by phone and asked about improvements.

Measure is the percentage of clients improving. For the various specific targets given by agencies

outside the PHCT such as immunisation rates, develop a scoring system whereby each level of

meeting the objective gets a certain number of points. E.g. if the target immunisation rate was

80%, actually doing 80% would give 100 points, 60% immunised would be 20 points, 70% 80

points, 90% 130 points, etc. The number of points would be based in the importance of that

target. The index would be the percentage of actual points earned compared to the maximum

possible points received if all targets were met. (Note, this assumes there are lots of such targets.

If this is not true, a composite measure such as this is probably not necessary.

• The percentage of required reports completed on time.

• The number of required reports returned by agencies requesting corrections or additional

information. (This would be an index of the quality of the reports.)

Client Satisfaction

Establish a formal procedure where clients can make complaints including a process for following up

on these complaints. Measure is the number of such complaints which were not concluded to the

client’s satisfaction within one week.

Effective Management of Resources

• Number of new initiatives developed that are designed to help team members use their time better.

These initiatives should also be reviewed on a regular basis to ensure they are still effective.

• Percent client related time as a percentage of total time. This measure gets at how much time is

devoted to clients. It does not measure how well that time is being spent. Other indicators are

needed to address this issue. (RDP: Note that this indicator is one where there is probably an

optimal level between the extremes. To little time with clients may suggest too much

administration time. Too much time with clients may suggest too little administration time.)

• Percentage of staff turnover over time. High staff turnover leads to inefficient resource utilisation

because it takes time to teach procedures to new staff and work is lost as a departing staff member

leaves. This measure would also be an indicator for the satisfaction of team members.

• Percentage of appointments which are unfilled or where the client did not come.

Development and Satisfaction of Primary Health Group/Team Members

• Training and development. A list of training and development experiences for each person on the

team would be developed each year. For example, attendance at a certain type of conference,

training on a piece of office equipment, learning a new procedure, etc. This list would be the

development plan for that person for that year. There would be two measures for training and

development. The first would be the percentage of team members who had the written plan. The

second measure would be the percentage of the development plan items actually completed.

• Who are reviewed, given feedback, and have a formal, jointly developed action plan for making

improvements.

• Satisfaction. Measure overall satisfaction on a monthly or bi-weekly basis with a very brief

questionnaire that would take no more than 2 minutes to compete. Measure would be the

percentage of staff indicating Satisfied or Very Satisfied with their jobs.

• Staff turnover is also a satisfaction measure. Note this measure under Effective Management of

Resources.

Figure A.2 List of participants

NAME JOB TITLE PLACE OF EMPLOYMENT

John Horder President CAIPE

Debbie Mellor Section Head of Workforce

Non-Medical Planning

NHS Executive

Thelma Sackman Nursing Officer NHS Executive

Kate Andrews Clinical Research Fellow Dept. General Practice

Rosemary Field To Be Advised To Be Advised

Nicki Meade Research Associate The National Primary Care

Research & Development

Centre

Steven Campbell Research Associate The National Primary Care

Research & Development

Centre

Brenda Leese * Research Fellow The National Primary Care

Research & Development

Centre

Bonnie Sibbald Research Associate The National Primary Care

Research & Development

Centre

Ann Richards Research Fellow Psychological Therapies

Research Centre

Malcolm McCoubrie Senior Lecturer in Community

Based Medical Education

Standards - Medical Director

Wandsworth Community

Health

Sheelagh Richards * Occupational Therapy Officer London

Jane Cannon* Practice Nurse Larwood Surgery

Sue Jenkins-Clarke Research Fellow University of York

Peter Bundred Senior Lecturer in Primary

Care

University of Liverpool

Judy Mead * Physiotherapist Chartered Society of

Physiotherapists

Richard Brown * To Be Advised To Be Advised

Alan Chapman Management Education &

Development Manager

Primary Care

East Norfolk Health Authority

Lance Gardner Professional Officer The Queens Nursing Institute

Terry Brugha * Senior Lecturer & Honnary

Consultant Psychiatrist

University of Leicester

Rosamund Bryar Professor of Community

Healthcare Nursing Practice

University of Hull

Stephen Rogers Senior Lecturer in Primary

Care

University College London

Joan Lole Director of Nursing & Primary

Care

Mancunian Community Health

Paul Thomas Senior Lecturer Dept. Of

General Practice

Imperial College School of

Medicine @ St.Mary’s

Ruth Hudson Education Officer Community Practitioners &

Health Visitors Association

Jacky Hayden Dean of Postgraduate Medicine University of Manchester

Christiana Johnson Health Promotion Officer Princess Royal Community Health

Centre

Beverley Haynes Senior Health Promotion Specialist Princess Royal Community Health

Centre

Peggy Newton Lecturer in Psychology Dept. Of General Practice

Jeanette Naish Senior Lecturer in Primary Care Dept. General Practice & Primary

Care

Stuart Mee * Practice Manager The Crookes Practice

Kay Robinson Primary Healthcare Facilitator South Humber Health Authority

Brian McAvoy Professor of Primary Health Care Dept. Of Primary Care

Susan Lonsdale Senior Principal Research Officer Dept. Of Health

Sandra Dodgson Senior Development Manager N H S Development Unit

Frances Fogg Primary Healthcare Facilitator North Notts Health Authority

Wendy Whyte Regional Community Nursing

Team Leader

British Forces Overseas

Mike Sharpe Regional General Manager of

Medical Services

British Forces Overseas

Ron Pollock Assistant Director Support &

Development/Finance

Wakefield Health Authority

Mike Vaughan Total Purchasing Project Manager Wakefield Health Authority

Sasha Wishard Research Facilitator Tayside Centre for General Practice

Marion Duffy Education Facilitator Tayside Centre for General Practice

Chris Simmonds * Practice Manager Medical Centre Doncaster

Jane Solomon * Locality Management Nottingham Health Authority

Catherine Booth * General Practitioner G.P. Unit

Ann Netton Assistant Director of PSSRU University of Kent

Gwen Wilson * Development Manager Community

Nursing

Sheffield Community Health

Appendix III

Effectiveness Measures

Developed for Primary Health Care Teams

Core Objectives for Primary Health Care teams

Promote, maintainand improve health

§ Provide high quality health care§ Accurate identification of individual and population health

care needs§ Review and improve the effectiveness of health care

provision§ Manage illness, injury and disease taking account of

agreed standards and evidence based practiceEnable personaland communityresponsibility forindividual health

§ Enable patients/clients to make informed decisions abouttheir own health.

§ Proactively encourage positive health behaviour§ Implementation of health education and preventative

care programmesEfficient use ofresources

§ Human resources – skills, knowledge, expertise, time§ Physical resources – budgets, equipment, premises

Continuouspersonal andprofessionaldevelopment

§ Individual annual training plans which take account of theplans of the PHCT

§ Equal access to training/development resources

High team membercommitment, stressand satisfaction

§ Team working

§ Mechanisms for reviewing and acting upon staffdissatisfactions, conflicts and complaints

Responsiveness toclients andcommunity

§ Gather information and feedback from clients/communitystakeholders/opinion leaders

Collaboration andpartnership withother relevantorganisations

Objective: Promote, maintain and improve health

Techniques for reviewing whether services meet client needs

A. What are the main aims of this service?

B. What does the team (in collaboration or in addition to other agencies)currently do to meet a particular health/health promotion need?

List all the provisions currently available in the team (and from other

agencies, if relevant).

C. How do you know whether these provisions meet these health/healthpromotion needs?

List evidence that can be used to determine this.

D. Which aspects of this evidence suggest that you are meeting thishealth/health promotion need?

E. Which aspects of this evidence suggest that you are not meeting thishealth/health promotion need?

F. What provision would the team ideally like to have in place to meetthis health/health promotion need?

Next steps:

Use the evidence discussed in C, D and E to develop measures to enable the teamto evaluate more systematically whether they are meeting clients needs.

Objective: Promote, maintain and improve health

Measure 1 - Review of quality in case management

Measure =Percentage of cases judged to be managed appropriately on the mostrelevant quality dimensions.

Steps to clarifying the measure:

∗ Determine the types of cases to be reviewed (specific condition, e.g.

diabetes/asthma, or a specific age group or type of patient population).

∗ Decide which of the quality dimensions are most relevant to the cases being

reviewed.

∗ Decide what is an acceptable quality level on each dimension.

∗ decide what is an acceptable % of cases to be judged as having been managed

appropriately.

Using the measure:

∗ Rate each of the selected cases on the quality dimensions and give a total score.

Note the dimensions where quality is above and below the acceptable level.

∗ Calculate % of cases which fall above and below the acceptable level of cases

being managed appropriately.

∗ The review will result in two types of information

⇒ dimension of quality for individual cases which fall below the acceptable

standard.

⇒ % of cases overall which are managed appropriately.

N.B. For this measure need to develop an instrument for rating cases on each

dimension which suggests evidence that can be used to make judgements,

emphasise the importance of standardising ratings across cases and gives guidance

in how to complete the instrument.

Dimensions of Quality

Effectiveness: Is the treatment given the best available in a technical sense,

according to those best equipped to judge?

What is their evidence? What is the overall

result of the treatment?

Acceptability: How humanely and considerately is this treatment/service

delivered? What does the patient think of it? What

would/does an observant third party think of it (“How would I

feel if it were my nearest and dearest?”) What is the setting

like? Are privacy and confidentiality safeguarded?

Efficiency: Is the output maximised for a given input or (conversely) is the

input minimised for a given level of output? How does

the unit cost compare with the unit cost elsewhere for

the same treatment/service?

Access: Can people get this treatment/service when they need it? Are

there any identifiable barriers to service - for example,

distance, inability to pay, waiting lists, and waiting times -

or straightforward breakdowns in supply?

Equity: Is this patient or group of patients being fairly treated relative to

others? Are there any identifiable failings in equity - for

example, are some people being dealt with less

favourably or less appropriately in their own eyes than

others?

Relevance: Is the overall pattern and balance of services the best that

could be achieved, taking account of the needs and wants of

the population as a whole?

Dimensions of Quality

Effectiveness

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Is the treatment/service beinggiven technically the bestpossible?

� � � � �

Does the treatment/servicebeing given conform to agreedprotocols/standards?

� � � � �

Is the current outcome fromthe treatment/service as wouldhave been expected, given thepatient’s condition at the start?

� � � � �

Acceptability

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Is the patient’s privacysafeguarded?

� � � � �

Is the patient’s confidentialitysafeguarded?

� � � � �

Is the patient treated withconsideration and respect?

� � � � �

Efficiency

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Are the inputs to thetreatment/service (e.g. stafftime, medication) minimisedfor a given level of output?

� � � � �

Is the unit cost the same as forthis treatment/servicedelivered elsewhere?

� � � � �

Access

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Can patients access thetreatment/service when theyneed it?

� � � � �

Do any of the following pose abarrier to accessing thetreatment/service?:

� � � � �

Location� � � � �

Distance� � � � �

Time of availability� � � � �

Inability to pay� � � � �

Waiting lists� � � � �

Waiting times� � � � �

Lack of appointment times

� � � � �

Lack of resources to supply

treatment/service

� � � � �

Equity

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Is this group of patients beingfairly treated relative to others?

� � � � �

Are the resources available forthis treatment/servicecomparable to those availablefor others?

� � � � �

Relevance

To what extent......

To a verylittle

extent

To someextent

To a verygreatextent

Are the resources used for thistreatment/service appropriatein the context of the needs andwants of the practicepopulation as a whole?

� � � � �

Objective: Promote, maintain and improve health.

Measure 2 - Young People’s health - Sexual Health

Measure = Percentage unwanted teenage pregnancies in a 6 month period Percentage of teenagers prescribed the morning after pill in a 6 month

period Percentage of teenagers requesting pregnancy tests in a 6 month period

Steps to clarifying the measure:

∗ Over a 3 month period monitor the number of unwanted teenage pregnanciesand terminations, number of morning after pills prescribed, number of teenagersrequesting pregnancy tests. This will establish a base line.

∗ Compare the numbers (or % of total number of teenager girls on the practice list)with the teenage pregnancies, use of morning after pill, teenagers requestingpregnancy tests in other PHCTs, and/or with regional figures. This enables theteam to assess the extent to which they are meeting the sexual health needs ofyoung people.

∗ Decide what is an acceptable level of unwanted teenage pregnancies, morningafter pill, requests for pregnancy tests.

Using the measure:

∗ Over a 6 month period log each: unwanted teenage pregnancy; request formorning after pill; and request for a pregnancy test.

∗ Note whether it is a small number of teenage girls who make the requests, orspread across a wide range of girls.

∗ Note whether there are any patterns (i.e. times of the week/month).

∗ After 6 months (or sooner if there are sufficient incidents of pregnancies/requestfor morning after pill/requests for pregnancy tests to form a judgement), collatethe information collected.

Next steps:

∗ Compare the % for teenage pregnancies, requests for morning after pills andrequests for pregnancy tests with (a) what were considered to be acceptablelevels, and (b) with figures for other practices.

∗ On the basis of this determine whether the current provision to meet the sexualhealth needs of young women is (a) being met (how do the figures for the PHCTcompare with the acceptable level? Are they better, worse, the same?), and (b)how the extent to which these needs are being met compares with otherPHCTs/regional averages.

∗ If the measures suggest that the sexual health needs of teenagers are not beingmet, introduce interventions to address this. The information about whether it isthe same small number of young women requesting morning after pills/pregnancytests will help to determine the type of interventions required.

∗ Once interventions have been introduced, re-use the measure to determinewhether the provision of services has improved.

Interventions:

∗ What type of follow-up is there when a teenager requests the morning after pill/apregnancy test?

∗ Gather more information about why young people take risks.∗ Implications for HIV/AIDS.

Objective: Promote, maintain and improve health.

Measure 3 - Young People’s Health - Alcohol and Drug Misuse

Measure = Number of teenagers attending A & E after drug overdose in a 3month period.Number of teenagers attending A & E after excessive alcohol

consumptionin a 3 month period.

Steps to clarifying the measure:

∗ Over a month monitor the number of A & E slips which record that a teenager hasattended A & E for drug or alcohol abuse.

∗ Compare these numbers with national/regional figures, and the number of otherPHCTs. This will help establish the extent the team is meeting these healthpromotion needs of young people compared to other PHCTs.

∗ Decide what is an acceptable number of A & E attendances for drug and alcoholabuse among teenagers.

Using the measure:

∗ Over a 3 month period log each A & E attendance by a teenager for (a) drugabuse, (b) alcohol abuse.

∗ Note whether it is a small number of teenagers who attend A & E for drug andalcohol abuse, or if it is spread across a wide range of teenagers.

∗ Note whether there are any patterns (times of the week/month).

∗ After 3 months collate the information requested. Determine (a) number ofincidents of drug and alcohol abuse at A & E, (b) number of teenagers whoattend A & E once, number who attend regularly.

Next steps:

∗ Compare the number of A & E attendances for drug and alcohol abuse with (a)what the team judged to be an acceptable number and (b) with figures from otherpractices and regional/national figures.

∗ On the basis of this determine whether the current health promotion to raiseawareness about the changes of drug and alcohol abuse are (a) being met (howdo the recorded numbers compare with the agreed acceptable level? Are theybetter, worse, the same?) and (b) how the extent to which the health promotionneed is being met compares with other PHCTs/regional figures.

∗ If the measures suggest that health promotion is not being effective, introduceinterventions to assess this. The information about whether is it the same ordifferent young people who misuse alcohol and drugs to determine the type ofinformation required.

Measure 4 - Patient access to consulations with a GP

Measure = The number of days that patients wait to see a GP of their choice

The PHC design team believed that an important part of providingquality care was to ensure continuity of patient care. The aim wasto ensure that patients had access to the GP of their choice (the GPwho ha most often provided health care in the past) by reducing thelength of time they had to wait to see this GP.

Steps to clarifying the measure

• Monitor the length of time patients have to wait to see each GP over a 1 monthperiod.

• If there are variations in the waiting time from week to week note the factorswhich might account for this (eg GP absences, increase in patient demand,services provided by other team members).

• Decide what is the target waiting time for each GP. Set this target taking accountof the factors which increase and reduce the waiting time. This target might bethe number of days a patient has to wait to see a GP of their choice, or it mightbe more realistic to set a target which specifies the maximum and minimumrange, to allow for fluctuation which are outside the team’s control.

Using the measure

• Over a 3 month period monitor the length of time patients have to wait to seeeach GP in the team.

• If there are variations across weeks, months, or between GPs, note the factorswhich might account for these.

Next steps

• Depending on the target set, calculate the average length of time that patientshave to wait to see a GP or calculate the maximum and minimum lengths of timethey have to wait. Compare this with the target set.

• If the target has been met, use the information gathered which explainedfluctuations in the length of time patients had to wait to assess whether it mightbe possible to reduce waiting times further (ie, if increases in patient demandsincreased waiting times, is it possible to anticipate and plan for these increases?).

• If the target was not met, use the information gathered which explainsfluctuations to assess what changes need to be made so that the target can bemet. Also consider whether the target is realistic.

Measure 5 – Patient access to a quality consulation with GPsMeasure 1= Percentage of patients whose appointment with a GP is minutes

duration in a 3 month period.

Much of the discussion in the PHT design team concerned how to most effectivelyuse the staff resources within the team to meet patient needs. The aim was toachieve this by having mechanisms in place which ensured that patients saw thehealth professional in the team most appropriate for their needs and as aconsequence, reduce the number of patients who needed/wanted to see a GP. Thiswould enable GPs to have longer (10 minute) appointments with those patientswhose health needs require a GP consultation.

GPs being able to have a longer appointments with patients was judged by the teamto be a measure of quality of care because it is seen as an indication that patientsneeds are being met by the appropriate health professional in the team.

It can also a measure effective use of resources.

In addition, ithe measure is an indication that the mechanisms put in place to ensurethat patients see the most appropriate health professional are effective (on thecondition that the longer GP appointments do not increase the workloads of the otherhealth professionals in the team).

Steps to clarifying the measure

• Over a month monitor the number of patients who have 10 minute appointmentwith a GP (this is a booked appointment, not a shorter appointment which overruns).

• Calculate the average number or % of patients in a week who can be offered a10 minute appointment.

• Decide what is the target number or % of patients who can be offered a 10minute appointments. When setting the target it may also be useful to considerthe types of patients on the practice list who might benefit from longerappointment so as to establish the level of possible demand. It will also be usefulto consider other factors which might affect the demand. For example, will therebe seasonal variations?

Using the measure

• Over a 3 month period monitor the number or % of patients each week who areable to have a 10 minute appointment.

• Note the types of patients seen and, if there are weekly variations, the factorswhich might account for these variations

• Also note whether there are any unforeseen consequences (eg, increased workload for other team members, administrative difficulties).

Next steps

• Compare the average number or % of patients who could be offered a 10 minuteappointment with the target which was set.

• If the target is achieved consider whether (a) any of the unforeseenconsequences need to be taken into account, (b) whether there are ways that thistarget can be improved.

• If the target is not achieved consider whether any changes to the factors whichaccounted for variations might help the team top achieve the target. Alsoconsider any impact of the unforeseen consequences.

Measure 6 – Use of out of hours services by patient

Measure = Percentage reduction in the use of private out of hours services by patients in a 6

month

period.

The PHC design team believed that using out of hours services did not providepatients which the best quality service (eg, they are seen by a health professionalwho does not know their medical history). Thus reducing the number of patientsusing out of hours services would reduce number receiving poor quality treatment. Areduction would also lead to more effective use of resources in the team – thesavings made from the reduction in the use of out of hours services could be used toemploy an additional health profession in the team, eg a nurse practitioner.

Steps to clarifying the measure

• Over a 3 month period monitor the use of the out of hours services by patients.

• Note the types of patients who use the service (is it a few frequent users orwidely spread). Are the reasons for the reasons for using the out of hoursservices different for frequent users and occasional users? If the reasons aredifferent, would it be possible to reduce the out of hour usage of these twogroups? If there are fluctuations in usage? Note factors which account for thesevariations.

• Decide the acceptable level of out of hours service usage and the target amountof reduction. Set this target taking account of the factors which are associatedwith increases and decreases in usage (eg, seasonal variations, public holidaysetc). It may be necessary to have separate targets for frequent users andoccasional users.

Using the measure

• Over a six month period monitor the use of the out of hours service.

• If there are weekly of monthly variations in usage, note the factors which mightaccount for these.

Next steps

• Calculate the average number of times the out of hours service has been usedeach month over the six month period and compare this with the target set.

• If the target has been met, using the other information collected, considerwhether the most appropriate patients have been using out of hours service, andwhether there are ways that (a) the usage could be reduced further, and (b)

whether steps could be taken to ensure that the out of hours service is used bythe most appropriate patients.

• If the target was not met, use the information collected to consider changes whichneed to made to help ensure that the target is met in the future. Also use theinformation collected to consider whether the target is realistic.

Measure 7 – Patients have access to an appropriate health professional

Measure = Percentage of patients who have contact with a healthprofessional from the

team at a time and location most appropriate to them and to the professional in a 6 month period.

This measure emerged from discussions the PHC design team had about how toachieve quality of care by ensuring that patients’ needs were met by the healthprofessional most qualified to meet those needs.

This is a complex measure and more work is required to develop a measure whichcan be used to assess effectiveness.

Agreement needs to be reached on the following:

- which health needs can be most effectively met by which health professional.- how health needs are assessed- which health needs can be most effectively met in which location (eg, home, one-

to-one consultation, booked appointment, drop-in, clinic etc.) This needs to takeaccount of both patients’ and the health care professionals’ views.

- what % of patients it might be possible for each health care professional in theteam to see at a time and location most appropriate to them and the patient.

Measure 8 – Patients have access to a home visit from an appropriate healthprofessional.

Measure = Percentage of patients who have a home visit from the mostappropriate

health professional in a six month period.

This measure emerged from a discussion of the use of staff resources within theteam. The PHC design team were considering which team members carried outhome visits, the time of day when it was most convenient to carry out home visits andhow to determine whether home visits were appropriate (ie, some home visits meetsocial rather than health needs). The aim is to ensure that only patients who need ahome visit receive one, and that they are visited by the health professional (DN, GP,pharmacist, HV etc) who has the expertise to meet their needs.

This is a complex measure. The following needs to be determined before it can bedeveloped in a measure of effectiveness.

- which health (social) needs can only be met by a home visit- how to assess these needs- which of these health needs can be most effectively met by which health

professional in the team- what % of patients can realistically be seen at home by the most appropriate

health care professional.

Objective: Enable personal and community responsibility for individual health

Measure 9 - Patients understand the role and function of the PHCT.

Measure = Number of patient requests, use health professionals’ time and PHCTservices which are inappropriate in a 3 month period.

Steps to clarify the measure:

∗ Patient understanding is demonstrated by appropriate use of the healthprofessionals’ and other staff in the team, PHCT services, and appropriaterequests for information.

∗ Define what are judged to be inappropriate uses of: health professionals’ time,and other staff in the team; PHCT services; inappropriate requests forinformation.

∗ Develop a checklist of the above and circulate to team members.

∗ Decide what is an acceptable level of inappropriate uses of: health professionals’time, and other staff in the team; PHCT services; inappropriate requests forinformation.

Using the measure:

∗ Over a two week period all members of the PHCT record the number ofinappropriate uses of health professionals’ and other team members’ time, andinappropriate use of PHCT services and requests for information.

∗ Note type of inappropriate use/request, and type of patient.

∗ After two weeks collate the data from all team members and calculate thenumber of (a) inappropriate uses of health professionals’ time, (b) number ofinappropriate uses of other staff members’ time, (c) number of inappropriate usesof PHCT services by patients, and (d) number of inappropriate requests forinformation.

Next steps:

∗ If the number of inappropriate uses of staff time, PHCT resources and/orrequests for information are unacceptable, develop interventions to reduce thenumber.

∗ Use information on the type of inappropriate use of time/services, and type ofpatients to target the information.

∗ After the interventions have been put in place repeat the measuring process toassess progress.

Objective: Efficient Use of Resources

Measure 10 - Patients able to manage minor illness

Measure = Percentage of patients seen by health professionals in the team who hada minor illness which could have been managed themselves.

Steps to clarify the measure:

∗ Define what is meant by ‘minor’ illness.

∗ Develop a checklist of minor illnesses and circulate to all health professionals inthe team.

∗ Decide what is an acceptable level of patients to see with a minor illness 10% or40%?

∗ Decide whether some groups of patients should be excluded.

Using the measure:

∗ Over a two week period the health professionals in the team log each patientseen, and record which patients attend for minor illness.

∗ Note the type of minor illness, type of client.

∗ After two weeks collate the data from team members and calculate (a) totalnumber of patients seen, (b) total number attending with minor illness. It may beuseful to look at % of patients with a minor illness seen by each type of healthprofessional, and to note which types of minor illness patients attended with, andthe types of patients presenting with a minor illness.

Next steps:

∗ If the measure indicates that an unacceptable % of patients are seen who haveminor illnesses decide on interventions to reduce the %.

∗ The data collected on types of minor illness, which health professionals areseeing these patients, the types of illnesses and types of patients can all be usedto target the intervention/s.

∗ After interventions have been put in place repeat the measuring processes toassess progress.

Instructions for Record Sheet

The data are being collected over 5 working days, starting on XXX. You can continueto collect data in w/c XXX, if you miss any days in the previous week.

Please record on the form information about every patient you have contact with oneach of the 5 days.

Codes for each column are also printed on the bottom of the form.

Type of illness

Column one

MA = minor illness, acute C = chronic illnessMC = minor illness, chronic A = acute illness

Column two

Please describe all types of minor illnesses you have recorded in addition to those inyour leaflet, using medical terms. If the illness conforms to the definitions of minorillness in your leaflet, no further information is required.

Type of contact

Column three - 1 = phone 2 = home visit 3 = consultation

Type of consultation

Column four - 1 = routine 2 = emergency

Type of Patient

Column five - 1 = female 2 = male

Column six - Patient’s age in years

Seen before in last 7 days

has the patient seen another health care professional in the last 7 days for the sameillness as recorded in column 1?

Column seven - record which health professional has seen the patient.

1 = GP 2 = PN 3 = DN 4 = HV 5 = CPN 6 = other

Other comments

Column eight - Please write down any other important information, and, if relevant,note if the patient has been referred inappropriately by other agencies such assecondary care/A&E/social services/dentist, as well as inappropriate internalreferrals.

Please give ALL completed recording forms to the Practice Manager.

Date: Day of week: Name: Job title:

Type of illness1) Code 2) Description ofillness

3) Type ofcontact

4) Type ofconsultatio

n

Type of patient 5) Gender 6)Age

7) Seenbefore inlast week

8) Comments

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1) Minor/Acute = MA 3) 1 = phone 5) 1 = female 7) 1 = GP 8) Please include otherinformation, and Minor/Chronic = MC 2 = home visit 2 = male 2 = PNreferrals from social services and

Chronic = C 3 = consultation 3 = DN secondary care, otherteam members. Acute = A 6) Age in years 4 = HV

5 = Other2) Please describe illness if not 4) 1 = routine minor according to the 2 = emergency agreed definitions.

Measure 11 - Patients/clients who do not attend for an appointment

Measure = Average percentage of total patients appointments not kept in a week (calculatedover a 3

month period).

Steps to clarifying the measure:

∗ Monitor the DNAs for a one month period for each professional group (GP, DN, SN) toestablish the current level in a one week period.

∗ Collect information on DNA levels for other comparable practices (i.e. have a similar typeof practice population).

∗ Decide what is an acceptable DNA level for each professional group.

Using the measure:

∗ Monitor the DNAs for a two month period for each professional group.

∗ Monitor, where possible, the following:⇒ which patients DNA (persistent or across a wide range?)⇒ characteristics of DNAs (age/gender/ethnicity)⇒ when patients DNA (i.e. time of day/a particular GP, PN etc,/regular appointments

booked in advance)⇒ whether DNAs make another appointment⇒ whether DNAs use other services (e.g. out of hours, home visit)

∗ Calculate the average % of patients/clients who DNA for each professional group in aone week period (this is the number of DNAs as a % of the total number of appointmentsmade in each week).

∗ Compare the DNA average with the acceptable level for each occupational group andwith other practices.

∗ Compare the % DNAs for each week in the two month period - does it vary from week toweek? If so, can these differences be explained?

∗ Use the additional information collected to assess whether discernible patterns in theDNAs. Does it happen at certain times of the day, and/or do certain types of patientsDNA more than others? This information can be used to make decisions about how toreduce DNAs. The information about whether DNA patients make another appointmentand/or use other services will provide additional evidence about the cost of DNA to theteam.

Next steps:

∗ If the DNA levels for the team as a whole and/or for specific occupational groups areunacceptably high introduce initiatives to reduce the level, and monitor progress usingthe measure.

∗ The additional information collected will help to establish what types of initiatives might help to reduce DNAs.

∗ It may also be necessary to gather information from patients and clients about the

reasons for DNA (is it because they forgot to attend? Because it is difficult to cancel anappointment? Because they are unaware of the implications to the team of DNA?).

Measure 12 – Efficient use of administrative systems

Measure = Percentage of patients not attending appointments with health professionals in the

team

which result from errors in the administrative system.

The additional work carried out at in the PHCT to explore the reasons why patientsDNA revealed some problems with the administration systems. The result was thatpatients cancelled their booked appointment, but this was not entered on the system.In addition some patients reported that as they had attended the surgery close to abooked appointment they assumed that this had been cancelled.

These findings suggest that some reduce in DNA rates would result from developingand improving current systems.

Developing a measure requires the following:

- further work to identify the range of system problems which could be improved by the team- calculate the current number of DNAs which result from system problems- improve and develop the system- continue to monitor the reasons for DNA and assess whether the number resultingfrom system problems decreases / or monitor the DNA rate and if this decreasesattribute this to the improvements and developments in the system.

Measure 13 - Efficient use of GP resources in the teamMeasure = Average number of patients seen by a GPs in a week

The PHCT had introduced a ‘sit and wait’ session; all patients who went to the surgeybetween 10.00 qnd 11.00 were seen by a GP. Introducing the sit and wait sessionenabled the resources of the GPs to be used differently. Two GPs held patientconsultations while a third carried out administrative tasks, dealt with telephonequeries and carried out home visits. The aim was to offer the same number of face-to-face consultations with patients, but with a reduced GP resource.

It was decided, therefore, that maintaining the number face-to-face consultations withpatients was an indication of the success of the sit and wait.

Data from the practice computer showed that there had been a substantialreduction in the number of patients seen by GPs (comparisons were madebetween a 1 week period in 1999 and the same week in 1998).

Discussion of the reasons why GPs might be seeing fewer patients revealed that thisreduction might indicate that the practice was infact using resources more efficiently.

Reasons proposed were:

- Less use of locum doctors for home visits (therefore fewer patients beingadvised to see their own GP after the home visit).

- Nurses seeing more patients (diabetics, blood clinic, hypertensive), nursepractitioner (available on Fridays).

- Patients can call the surgery and talk to a GP and get advice.- The pharmacist visits some patients at home to discuss medication.

This work emphasises the importance of looking at the activities of the team as awhole when assesses the effectiveness of specific aspects.

Objective: Continuous personal and professional development

Measure 14 – Team member access to training

Measure = Percentage of who are satisfied with the extent to which their training needs are

assessed

and met in the previous year.

Steps to clarifying the measure

q Agree what is an acceptable level of satisfaction within the team.

Using the measure

q Each member of the PHCT completes the measure of satisfaction with training.

q Calculate the mean satisfaction with training score for each person (total theresponses from each question ….. and divide by the number of questions). Thencalculate what % of staff report a satisfaction level at, above and below theageed acceptable level.

Next steps

Analysis of the responses to the individual questions in the training satisfactionmeasure can be used to determine the steps which need to be taken to improveaccess to training within the team. For example, these responses can indictatewhether there are concerns about funding available to support training, or if there areissues relating to the identification of training needs. The former could be resolvedby identifying additional sources of funding, while the latter could be tackled via theappraisal system.

Objective: High team member commitment, stress and satisfaction.

Measure 15- Team member commitment and satisfaction

Measure = Percentage of staff in the team who feel committed and satisfied

Steps to clarify the measure:

• Agree what is an acceptable level of commitment within the team • Agree what is an acceptable level of job satisfaction

Using the measure:

• Each member of the PHCT completes the measures of commitment andsatisfaction. It is important that confidentiality is maintained and that it is notpossible for individual responses to be identified

• Calculate the mean job satisfaction for each person (total the responses from

each question, extremely dissatisfied = 1 to extremely satisfied = 7 and divide bythe number of questions, 16). Then calculate what % of staff report a satisfactionlevel at and above the agreed acceptable level for the team

• Calculate the mean commitment for each person (total the responses from each

question, strongly agree = 5, strongly agree = 1, and divide by the number ofquestions, 6). Then calculate what % of staff report a level of commitment at andabove the agreed acceptable level for the team

Next steps

Analysis of the responses to the individual questions in the commitment and jobsatisfaction measure can be used to determine the steps which need to be taken toimprove the overall levels of commitment and satisfaction within the team.

Measure 16 - Team members use each others skills, knowledge and expertiseappropriately

Measure = Percentage of team members who report thatskills, knowledge and expertisewithin the team are used appropriately in 3 monthperiod.

Steps to clarifying the measure

q Agree what is an acceptable level of appropriate use of skills, knowledge andexpertise.

Using the measure:

• Each member of the team completes the questions on the use of each othersskills, knowledge and expertise.

• For each question, calculate the extent to which skills, knowledge and expertise

are used appropriately (total the responses on each dimension = 1,= 5 and then calculate the mean (divide the total by the number of dimensions).

Next steps

If the levels of awareness and appropriate use of skill, knowledge and expertise arebelow the acceptable level the team could improve this by holding more effectivemeetings (when all members are encouraged to contribute to decision-making), bygetting involved in joint projects together, and carrying out work shadowing.

Measure 17 - Effective team working

Measure = Percentage of requests for help and information and referrals from otherteam members which are inappropriate in a 3 month period.

Steps to clarifying the measure:

• Agree what is an acceptable % inappropriate requests for help and information,and level of inappropriate referrals from other team members

Using the measure:

• Over a specified period (one or two weeks) each member of the team logs eachtime another team member requests help and information, and refers a patient.Against each, each team member notes whether this was an appropriate orinappropriate request for help/information or patient referral. It may also beuseful to note which team member made the request for help/information ormade the referral

• At the end of the specified period calculate what percentage of the total number

of requests for information and patient referrals were inappropriate. Comparethis with the acceptable levels agreed.

Objective: Responsiveness to client and community

Measure 18 - Patients’ Experiences of the PHCT service (1)

Measure = Percentage of patients who report that their experiences of the PHCTservices match the standard agreed by the PHCT.

This can be assessed by measuring patients’ experiences of the PHCT.

Steps to developing a measure:

∗ Identify all the aspects of the PHCT’s work and how it is delivered which areknown to be associated with patients satisfaction, e.g. not having to wait, gettingrepeat prescriptions, phone answered quickly, being able to get advice etc.

∗ Develop a checklist for patient asking if they have experienced each of theaspects associated with satisfaction.

∗ Either ask about experiences of the PHCT in general, e.g.⇒ Do you get your repeat prescription within 48 hours? sometimes

always never

∗ Or ask about the contact with PHCT the patient has just had, e.g.⇒ Did you have to wait more than 2 days to get an appointment with the

GP? yes no

∗ Identify other information you would like to collect from patients which might helpyou to use or understand the information you collect on patients’ experiences(e.g. age, gender, number of visits to the surgery in the previous month). It mightbe useful to ask patients to write their own comments.

∗ Consider what would be acceptable and unacceptable responses to the patient’sexperiences questions, e.g. would you expect 90% of patients to report they got arepeat prescription within 48 hours or 10%? Also consider whether some areasare more important than others.

Using the measure:

∗ Over a one week period ask all patients attending the surgery to complete achecklist. Send a % of questionnaires to home addresses, and distribute via DN,HV etc.

∗ Collate the information from patients. Calculate a total score for each patient andthe mean.

It may be useful to look at responses on each item separately (particularly if youconsidered some patient experiences to be more important), and to identify whetherthe views of different types of patients vary.

Next steps:

∗ If patients’ experiences of some aspects of the PHCT work are not as positive asthe team had anticipated develop interventions to improve these. Or you may findexperiences vary across different types of patients and the team want to takesteps to remedy this.

∗ After interventions have been put in place repeat the patient survey to assessprogress.

Measure 19 - Patients’ experiences of the PHCT services (2) (Using the existingmeasure)

Measure = Percentage of patients whose experiences of the PHCT services meetthe standard set by the team.

Steps to clarifying the measure:

∗ On each of the questions in the patient Opinion survey agree the ideal standardthe PHCT wants to achieve for example question 1, the length of time patientswait to get an appointment with a GP, what % of patients does the team aim tosee on the same day/next day/after 2 days/3 days?

∗ On each of the questions in the Patient Opinion survey agree the expectedstandard that the PHCT currently achieves.

Using the measure:

∗ Over a one or two week period distribute questionnaires to patients attending thehealth centre/surgery, attending clinics, and those seen by the HV, DN, CPN andby other professionals carrying out domicillary care. Also survey a sample ofpatients, selected at random from the practice list, who have not been seenduring the week.

∗ On each question calculate⇒ % of patients whose experiences of the PHCT services meet the ideal

standards set by the teams.⇒ % of patients whose experience of the PHCT services meet the expected

standard.

∗ Calculate a total score for % of patients whose experiences meet the idealstandard (total number of questions where patients’ experiences met the idealstandard and calculate this number as a % of the total number of questions).

∗ Calculate a total score for % of patients whose experiences meet the expectedstandard (total the number of questions where patients’ experiences met theexpected standards and calculate this number as a % of the total number ofquestions).

Date -___________________

Patient Opinion Survey

Please could you answer the questions listed below. Your answers will help us toimprove the service we provide for patients.

Age __________ years Male � Female �

1. The last time you wanted an appointment with any of the GPs, how soon didyou get one?

same day �next day �after 2 days �longer ___________

2. The last time you wanted an appointment with the GP of your choice, howsoon did you get one?

same day �next day �after 2 days �after 3 days �longer ___________

3. How could the appointment service be improved?____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

4. The last two times that you phoned the surgery, how long did you wait forthe phone to be answered?

1st time _________ mins Time of day: morning � afternoon �2nd time _________ mins Time of day: morning � afternoon �

5. The last two times that you asked for a repeat prescription, how long did youhave to wait to get it?

1st time ______________ days2nd time ______________ days Not applicable �

6. Have you ever experienced problems/delays with getting a repeatprescriptions?

Yes � No �

7. If yes, please give details of where the delay occurred e.g. at the HealthCentre or at the chemist____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

8. How could the repeat prescription service be improved?____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

9. Have you ever used the services of the Practice Nurse?

Yes � No �

If No, please go to Question 13.

10. Did you have an appointment?

Yes � No �

11. The last two times you had an appointment with the Practice Nurse, howlong after the appointment time did you have to wait to see her?

1st time ____________ mins2nd time ____________ mins

12. If your GP referred you to the Practice Nurse, how long did you sit in thewaiting room until the practice nurse was available?

_____________ mins

Did you know that you could make an appointment to see the Practice Nurse?

Yes � No �

13. Have you required a doctors appointment and had to wait for the PracticeNurse to phone?

Yes � No �

If No, please go to Question 15.

14. The last two times you used this service how long after 2pm did you haveto wait for the Practice Nurse to phone you?

1st time _____________ mins2nd time _____________ mins

15. How could the Practice Nurse services be improved?____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

16. Which of the following services do you think are offered by HealthVisitors? Which have you used?

offered usedAnte natal care � �Post natal advice/support � �Child development assessment � �Behaviour management � �Childcare advice � �Continence advice � �Adult support/advice � �Elderly support/advice � �

If you do not use the services of the Health Visitor, please go to question 20.

17. Who is your named Health Visitor? _________________________________

18. In the last month:

How many morning clinics did you attend? �How many afternoon clinics did you attend? �

How long did you (your child) wait to be seen at each clinic?

morning afternoon__________ mins __________ mins__________ mins __________ mins__________ mins __________ mins__________ mins __________ mins

19. How could the services offered by Health Visitors be improved?____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

20. Have you ever used of the services of District Nurses?

Yes � No �

If No, please go to Question 27.

21. Have you got a named District Nurse?

Yes � No �

22. If yes, who is your named District Nurse? __________________________

23. How long did you have to wait for a visit from a District Nurse for thefollowing:

(i) Urgent condition ________ hours Not applicable � ________days

(ii) Discharge from hospital ________ hours Not applicable �________ days

(iii) Routine referral from GP________ hours Not applicable � ________days

(iv) Nursing home assessment________ hours Not applicable � ________days

24. Did your GP tell you that they would arrange for the District Nurse to call?

Yes � No �

If yes, did this happen within the time period given by the GP?

Yes � No � Not applicable �

25. Hospital discharge (if you have not been discharged from hospital in thepast month, please ignore this section)

Did the hospital tell you that the District Nurse would call to see you?

Yes � No �

Did you have to contact the Health Centre before the District Nurse made a visit?

Yes � No �

How long did you wait following discharge to see the District Nurse? ________days

26. How could the District Nursing services be improved?____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

27. Please write below any other comments you would like to make about thePractice.____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Many thanks for your help.

Other measures – developed in training workshop

Qualityq % of patient records which are: relevant; concise; contemporaneous; legible;

dated; signed and actioned.

q % of previouly agreed conditions for which there is an evidence based protocolwhich is reviewed annually.

q % of evidence based protocols which are audited.

q % of time spent on patient and non-patient contact.

q Number of complaints about access to services.

q Number of inappropriate experiences of access.

Team workingq Number of suggestions which are agreed and acted upon.

q % of time in a month when GPs are available for consultation with other membersof the PHCT.

Administrative efficiencyq % of time spent looking for case notes

q % of time spent preparing repeat prescriptions

Appendix IV

Training Programme -

Tools and Techniques for Assessing

Performance

The Productivity Measurement and Enhancement System (ProMES) was used in the

research to develop effectiveness measures from primary care. Objectives were

developed in workshops with domain relevant experts from primary care (see

Appendix III) and effectiveness measures were developed with representatives two

primary health care teams (see Chapter 3 and Appendix II). ProMES is a

theoretically grounded approach, based on the NPI theory of motivation (Naylor,

Pritchard & Ilgen (1980), that has very practical applications. A critical feature of the

ProMES process is that those people who performance is being assessed are

involved in developing their own measurement and feedback systems. Participants

learn to set clear objectives, to identify ways of measuring whether they are

acheiveing these objectives, and to collect and use information to provide feedback

on performance.

Two key lessons were learnt from the Workshops and the work carried out with

primary health care team members developing measures. Firstly, the ProMES

process provides a valuable learning experience, and secondly, it is possible to

identify common objectives for primary care and to develop measures to assess

performance against these objectives.

Having demonstrated the utility and value of ProMES, and encouraged by the

exepriences of the representatives from primary care who had experienced the

ProMES process, the research team developed a ProMES training programme. The

programme was designed to train primary health care team members, and trust and

health authority staff working with primary care teams, in the ProMES technique.

Letters were sent to all the primary health care teams involved in the research

inviting representatives to attend the training programme. Responses were received

from only two teams and four team members attended the training. Letters of

invitation were also sent to representatives working with primary health care teams in

the local community trust and health authority and four people attended from these

organisations.

Details of the training programme are provided below.

Tools and Techniques for Assessing Performance

Training Plan

Day 1

11.00 - 11.15 - Overall objectives of the training programme:IntroductionsOHP - Aims and objectives• To develop skills which enable the participants to develop performance measures

with primary health care teams.• To develop skills which the participants can use to identify and use evidence to

assess performance in primary health care teams.• To teach participants how to use the measures produced by the research team.

What else would the participants want to include? What specifically do they want toget out of the training?

11.15 - 12.00 - IntroductionThe purpose of this session is to:(a) clarify what is meant by ‘performance’(b) provide participants with an understanding of the contribution that measuringperformance can make to team effectiveness in primary care,(c) raise awareness of the complexities of measuring performance in primary care.

What is ‘performance’?It is important to be clear about what it is we are trying to measure.

Discussion session - What do they understand by ‘the performance of a PHCT’?

The definition to use: The outcomes from a team, what they produce, what they aretrying to achieve.

Introduce the basic principles of ProMES(Handout 1)- productivity- motivation- feedback- prioritising effort.

Basic approach:- Design team- Clarify objectives- Develop measures of the objectives- Gather information with the measures and use this to assess extent to whichmeeting objectives, and to identify changes which can be made to improveperformance.(Handout 2, Handout 3)

Measuring performance and team effectiveness.(i) The importance of effective team working in multidisciplinary groups.

(ii) The principles of effective team working (clarity of objectives and feedback onwhether achieving these objectives, participation, task style, support for innovation,reflexivity).Feedback is only useful in the context of clear goals.To measure performance PHCTs need to have clear objectives. Using measurescan provide feedback on performance. In addition the process of developing andusing measures requires participation and reflexivity. Thus measuring performancecan make a considerable contribution to effective team working.Developing and using measures also enables PHCTs to use evidence basedpractice.(Handout 4, Handout 5)

The difficulties of measuring performance in primary careGiven what we are trying to achieve (measure the outcomes from PHCTs) Why mightthis be difficult?Discussion with the group.- multiple stakeholders, therefore, lack of agreement re. objectives, outcomes,desirable outcomes- outcomes difficult to measure (eg quality of care, patient satisfaction)- lack of sound evidence.- aspects not within the team’s control.

12.00 - 1.00 - Basic Principles for Assessing PerformanceThe purpose of this session is to:(a) Agree the objectives for primary care.(b) Identify potential sources of information which could be used to evaluateperformance against these objectives.(c) Consider what are good and bad measures of performance.(Handout 6)

Step 1 - Setting objectivesOutline the process used to develop the Objective for Primary Care (workshop, workwith teams).Present the group with the Objectives. Rate the objectives in terms of importance(purpose of this is to get them thinking critically). Do they disagree with any of theobjectives?(Handout7, Handout 8)

Critically evaluate the objectives using the ‘criteria for good objectives’.(Handout 9)

Step 2 - Identifying sources of feedback informationWork in pairs, each discuss 2 objectives, and identify sources of information availablewhich could provide feedback on performance on these objectives. (the aim at thisstage is to raise awareness about all the sources of information available, notnecessarily to identify the best sources).

Step 3 - What is a good measure?Discussion of good and bad measures. Introduce the idea that what it is easy tomeasure, is not necessarily the best measure.Discuss the ‘criteria for a good measure’.(Handout 10)

2.00 - 3.30 - Measuring Use of Resources

The purpose of this session is to give participants the opportunity to work through theprocess of developing a measure.The task is for the group to develop a measure / measures which will indicatewhether resources are being used effectively in a team.An example of a measure. In manufacturing organisations wastage of raw materialscan add considerable amount to costs. A measure of performance in relation to theobjective ‘Efficient use of Resources’ could be, therefore, % reduction in the waste ofraw materials.ExerciseThe group/s consider the following questions and develop a measure of use ofresources.

• What are the resources used / available to a PHCT? (eg financial, skills,knowledge, equipment, time, rooms)?

• What evidence is available to indicate that resources are being used efficiently?• What evidence is available to indicate that resources are being use inefficiently?• Which resources is it most critical to use efficiently? (Ie which potentially have

the greatest impact on the performance of the team?)• Decide on one aspect of resources and develop a measure.• Critically appraise the measure using the ‘criteria for measures’.

The groups work on their own with support from the trainers.They need to select a scribe and someone willing to feedback in the plenary session.The group give feedback on the process they went through to develop a measure,and the measure developed(Handout 12)

4.00 - 5.30 Plenary SessionFeedback. What aspects of the process did they find easy/difficult? What problemsdid they encounter? Did they develop a good measure?

Review learning (from the whole day)

Day 2

9.00 - 10.45 - Measuring Quality of CareThe purpose of this session is to give participants more experience of workingthrough the process of developing a measure.iReview the learning points from the session on developing measures on use ofresources (ie what have they learnt and will do differently ?)The task is to develop a measure / measures of in relation to the objective Quality ofCare. This is a difficult exercise so we are providing some materials which mighthelp. Introduce the 5 dimensions of quality.(Handout 13, Handout 14)

Exercise.The group/s consider the following questions.• What is meant by quality of care in primary care? Whose perspective should be

taken into account?• What evidence is available to indicate that good quality of care is being provided

by the PHCT?• What evidence is available to indicate that quality of care is not good?

• Which aspects of the PHCT services / types of conditions is it most critical tofocus on? (ie which would improvements in service / care have the greatestimpact on team performance?)

• Decide on one aspect of quality of care and develop a measure.• Critically appraise the measure using the ‘criteria for measures’.

The groups work on their own with support from the trainers.They need to select a scribe and someone willing to feedback in the plenary session.It is a difficult task, but we will be there to help.We want them to feedback on the process they went through to develop a measure,and the measure developed

11.00 - 11.30 - Plenary sessionFeedback. What aspects of the process did they find easy/difficult? What problemsdid they encounter? Did they develop a good measure?

Review learning.Important to emphasise that they are not now experts in developing measures, butare more aware of the process and the difficulties. With support from me they can dothis with teams.

11.30 - 12.30 - Developing measures with teams / PCGsThe purpose of this session is to provide participants with practical skills in runningmeasurement development sessions with PHCTsRun this as a how to do it information giving session.Hightlight pitfalls and problems. Emphasis thet MUST follow the correct process.MUST allow everyone to have a voice. Must NOT impose their own objectives /agenda. In addition , given the nature of primary care, the process can generateconflict - because it starts to make the implicit explicit and hightlight differences inpersepctives and values.Plus give the group the opportunity to discuss concerns, problems, obstacles thatthey foresee. etc.(Handouts 15 - 22)For example - setting-up the design team, explaining the purpose, agreeingobjectives, starting to develop measures, logging progress, action planning,gathering information, assessing the value of information, using information

1.30 - 3.00 - Using Performance MeasuresThe purpose of this session is to:(a) Familarise the participants with the measures developed by the research team sothat they understand how and when ( to use them. (how to gather evidence)(b) Develop an understanding of how to use the feedback from measures to prioritiseactivities within the PHCT, and to change existing working practices/services. (how touse evidence)

Explain how (the process) and why (to help teams to get feedback on aspects of theirperformance which were considered important to improving overall effectiveness) themeasures were developed and used by the researchers.Provide each participant with a pack of measures (need to think on an interestingway of doing this). Discuss when and how the measures might be useful.(Handouts 23, Handout 24)

Present some feedback data from the measures to the participants and discuss:

(a) how they would interpret this, what else they might need to know?(b) what actions might be taken as a result of the feedback?

3.15 - 4.00 - Plenary SessionReview of learningNext steps and action plansContacting the research team for supportSupport network/further meetings/next training.

Handout 1

Measuring and Enhancing Effectiveness inPrimary Health Care Teams

Measuring and enhancing system effectiveness is an important element in anyorganisational system, and Pritchard (Pritchard, Jones, Roth, Stuebing & Ekeberg,1988, 1989; Pritchard, 1990) has developed a sophisticated and widely applicableapproach to this - the productivity measurement and enhancement system(ProMES).

The ProMES approach is based on the theory of motivation presented by Naylor,Pritchard, and Ilgen (1980). In this theory, motivation is maximised when people seeclear connections between their efforts and the behavioural “products” or results ofthese efforts, there are clear perceived connections between a person’s products andtheir evaluations, and there are clear connections between these evaluations andvalued outcomes. When these conditions are met, motivation is high. In addition,motivation is maximised when the different evaluators and controllers of rewards in

the person’s environment such as the person himself/herself, peers, differentsupervisors, top management, and union personnel agree as much as possible onwhat should be done in the work and how it should be evaluated. When suchagreement exists, the efforts of the person are more clearly directed and the sameamount of effort results in greater productivity. In addition, stress and wasted effortare reduced.

The ProMES system develops a formal method to measure productivity and usesthese measurements as feedback to people doing the work to help them increasetheir productivity through maximising motivation. The idea is to maximise thevariables indicated in the theory so that motivation will also be maximised. Peopleare given the tools to do the work better and at the same time help them feel a senseof ownership in the resulting system and empowerment in determining importantaspects of their work.

One of the key elements in ProMES is feedback. People doing the work get regularlyoccurring, high quality feedback about how the work unit is doing. The personnel inthe work unit then use this feedback to develop plans for improving productivity.Feedback after this time tells them how well the plans they developed have actuallyimproved productivity. Furthermore, since they are heavily involved in the design ofthe measurement system and resulting feedback system, they have more confidencein its validity and accept it more than systems imposed from above. The ProMESapproach has been applied and evaluated in a wide range of settings (Pritchard,1995) and substantial improvements in performance have been shown as a result ofusing ProMES.

Handout 2

Practical tips and guidelines

Why use this system?

• it makes you think!• it gives you useful information• it indicates where you should focus your resources• it allows you to define your own measurement system - it puts you in control• it increases your participation• you are the first to know about any problems• people report less stress• people know how they are being evaluated• everyone has to agree the priorities - they are not imposed• you get valuable feedback

Handout 3

Practical tips and guidelines

Key implementation principles

• Measurement is the foundation• It takes a lot of work to measure well• What you measure is what you get• Measures for decision making are different from those for motivation• Good measurement makes feedback easy• Good feedback leads to productivity improvements• People want to do a good job• The key is to give them the tools• People want control over their lives• Acceptance of the system is essential for success• Participation leads to acceptance

Handout 4

Teamworking in Primary Care

The idea that teams are important to modern organisations was established about 70

years ago. However, in only the past 15 years has the idea been seized and widely

acted on by large numbers of organisations in the public and private sectors (Guzzo,

1996). But how effective are teams within organisations generally?

Macy and Izumi (1993) conducted an analysis of 131 organisational change studies

in order to determine their effectiveness. Those interventions with the greatest

effects on financially-related measures of organisational performance were team-

related interventions. These also reduced turnover and absenteeism more than did

other interventions, showing that team-oriented practices can have broad positive

effects in organisations. Abblebaum and Batt (1994) offer convergent evidence.

They reviewed the results of a dozen surveys of organisational practices as well as

185 case studies of innovation in management practices. They too found compelling

evidence that teams contribute to improving organisational effectiveness, particularly

increasing efficiency and quality. Other researchers provide evidence of the impact

of team-based work practices on organisational performance. Kalleburg and Moody

(1994) studied over 700 work establishments and found that those in which

teamwork was developed were more effective in their performance than those in

which were not used.

The importance of teamworking has been emphasised in numerous reports and

policy documents on the National Health Service. One recent document (NHSME

1993) particularly emphasised the importance of teamworking if health and social

care for people in local communities were going to be of the highest quality and

efficiency.

‘The best and most cost-effective outcomes for patient and clients

are achieved when professionals work together, learn together,

engage in clinical audit of outcomes together, and generate

innovation to ensure progress in practice and service.’(para 4.3)

Overall, research based evidence of teamworking in primary health care in the UK is

consistent with research in other sectors in suggesting the value of this way of

working for effectiveness and efficiency. Primary care team working has been

reported to improve health delivery and staff motivation (Wood, Farrow and Elliot,

1994) and to have led to better detection, treatment, follow-up and outcome in

hypertension (Adorian, Silverberg, Tomer and Wamosher, 1990). In a longitudinal

study of 68 primary health care teams, Poulton (1995) found a clear relationship

between teamwork and effectiveness. Those teams with high levels of clarity of team

objectives and team members commitment to those objectives were more effective

than those with unclear objectives.

However, despite these encouraging research studies, there is considerable

evidence that the context of primary health care is such that there are substantial

barriers to co-operation and collaboration in the delivery of primary health care.

Bond et al (1985) found little interprofessional collaboration in primary health care

teams in their study of 309 paired professionals. West and Poulton (1995) examined

primary health care team functioning in 68 practice teams and found that on all 4

dimensions of team functioning primary health care teams scored significantly lower

than the other team types. West, Poulton and Hardy (1994) in a study of 9 primary

health care teams identified structural, managerial and employment patterns in

primary care as crucial in undermining the effectiveness of teamworking. These

barriers to co-operation and collaboration need to be removed or reduced for

teamworking to be effective in primary heath care.

There are a number of key elements to effective teamwork (Guzzo and Shea, 1992):

• First, Individuals should feel that they are important to the success of the team.

When individuals feel that their work is not essential in a team, they are less likely

to work effectively with others or to make strong efforts towards achieving team

effectiveness. Roles should be developed in ways which make them

indispensable and essential.

• Individuals roles in the team should be meaningful and intrinsically rewarding.

Individuals tend to be more committed and creative if the tasks they are

performing are engaging and challenging.

• Teams should also have intrinsically interesting tasks to perform. Just as people

work hard if the tasks they are asked to perform are intrinsically engaging and

challenging, when teams have important and interesting tasks to perform, they

are committed, motivated and co-operative (Hackman, 1990).

• Individual contributions should be identifiable and subject to evaluation. People

have to feel not only that their work is indispensable, but also that their

performance is visible to other team members.

• Above all there should be clear, shared team goals with built-in performance

feedback. Research evidence shows consistently that where people are set clear

targets at which to aim, their performance is generally improved. For the same

reasons it is important for the team as a whole to have clear team goals with

performance feedback.

In primary health care, by and large, the first three conditions for effective

teamworking hold true. However, in primary health care teams it is rare for individual

contributions to be measured and feedback on performance given. Moreover,

primary health care teams tend not to have clear, specific objectives and goals and

feedback on performance against those objectives is rarely available. The

development of teamworking in primary healthcare, therefore, needs to focus on

developing clear, shared objectives and on providing feedback on performance.

Handout 5

The difficulties of measuring performance in primary care

• There are multiple stakeholders in primary care (different professional groups,and organisations), therefore, lack of agreement about objectives, what are theoutcomes from primary care teams, and what are desirable outcomes.

• Many of the outcomes are difficult to measure (e.g. quality of care, patientsatisfaction).

• There is a lack of sound evidence.

• Many of the factors which influence outcomes are not within the team’s control(e.g. other agencies, characteristics of practice population).

Handout 6

Practical tips and guidelines

Setting Objectives

• ask the team what it is they are trying to accomplish for their organisation• focus on larger objectives - give the group examples which are as similar to their

work as possible• this stage is typically not difficult• consensus should be easy to reach at this stage• the discussion at this point sets the tone for the future - there needs to be

balance between the facilitator saving the group time and taking control

Handout 7

Core Objectives for Primary Health Care

♦ To promote, maintain and improve health

Provide high quality health care

Accurate identification of individual and population health and care needs

Review and improve the effectiveness of health care provision

Manage illness, injury and disease taking account of agreed standards and

evidence based practice

♦ Enable personal and community responsibility for individual health

Enable patients/clients to make informed decisions about their own health

Proactively encourage positive health behaviour

Implementation of health education and preventative care programmes

♦ Efficient use of resources

Human resources - skills, knowledge, expertise, time

Physical resources - budgets, equipment, premises

♦ Continuous personal and professional development

Individual annual training plans which take account of the plans for the PHCT

Equal access to training/development resources

♦ High team member commitment, stress and satisfaction

Teamworking

Mechanisms for reviewing and acting upon staff dissatifactions, conflicts and

complaints

♦ Responsiveness to clients and community

Mechanisms for gathering information and feedback from clients/community

stakeholders/opinion leaders.

♦ Collaboration and partnership with other relevant organisations

Handout 8

Core Objectives for Primary Health Care Teams

To what extent do you think your team effectively meets the followingobjectives and sub-objectives?

To promote, maintain and improve health Of no Veryimportance important1 2 3 4 5 6 7

- Provide high quality health care 1 2 3 4

5 6 7

- Accurate identification of individual andpopulation health and care needs 1 2 3 4 5 67

- Review and improve the effectiveness ofhealth care provision 1 2 3 45 6 7

- Managing illness, injury and disease takingaccount of agreed standards and evidence 1 2 3 45 6 7based practice

Enable personal and community responsibilityfor individual health 1 2 3 4 5 6 7

- Enable patients/clients to make informed decisionsabout their own health 1 2 3 4 5 67

- Proactively encourage positive health behaviour1 2 3 4 5 6 7

- Implemention of health education andpreventative care programmes 1 2 3 4 5 67

Efficient use of resources Of no Veryimportance important1 2 3 4 5 6 7

-Human resources (skills, knowledge, expertisetime) 1 2 3 4 5 67

- Physical resources (budgets, equipment, premises)1 2 3 4 5 6 7

Continuous personal and professionaldevelopment 1 2 3 4 5 6 7

- Individual annual training plans which takeaccount of the plans for the PHCT 1 2 3 4 5 67

- Equal access to training/developmentresources 1 2 3 4 5 67

High team member commitment, stressand satisfaction 1 2 3 4 5 67

- Teamworking1 2 3 4 5 6 7

- Mechanisms for reviewing and acting uponstaff dissatisfactions, conflicts and complaints 1 2 3 4 5 67

Responsiveness to clients and community1 2 3 4 5 6 7

- Gather information and feedback from clients/community stakeholders/opinion leaders 1 2 3 4 5 67

Collaboration and partnership with otherrelevant organisations 1 2 3 4 5 67

Handout 9

Criteria for Objectives

• stated in clear terms• if exactly that objective was done, the organisation would benefit• the set of objectives must cover all important aspects of the work• objectives must be consistent with the broader organisation• higher management must be committed to each objective• keep the number of objectives manageable, normally 3 to 8

Handout 10

Practical tips and guidelines

Criteria for measures

• the measure must be consistent with the objectives of the broader organisation• if the measure was maximised would the organisation benefit• all important aspects of each objective must be covered by the set of measures• higher management must be committed to the measures• measures must be under control of the staff• measures must be understandable and meaningful to staff• it must be possible to provide information on the measure in a timely manner• the data must be cost effective to collect• the information provided by the measure must neither be too general or too

specific

Handout 11

Measuring Use of Resources

• What are the resources used / available to a PHCT? (eg financial, skills,knowledge, equipment, time, rooms)?

• What evidence is available to indicate that resources are being used efficiently?• What evidence is available to indicate that resources are being used inefficiently?• Which resources is it most critical to use efficiently? (i.e. which potentially have

the greatest impact on the performance of the team?)• Decide on one aspect of resources and develop a measure.• Critically appraise the measure using the ‘criteria for measures’.

Select a scribe and someone willing to feedback in the plenary session..We want them to feedback on the process they went through to develop a measure,and on the measure/s developed

Handout 12

Practical tips and guidelines

Developing Measures

• ask how they would show that the stated objectives were being met• this is a difficult step for the design team to do• it is frustrating - tell the group they will feel this• you must train the design team to develop and evaluate measures• the design team may not know the answer but they can find out• if someone tells you that you cannot measure what they do then ask them how

they think they are doing?, is it different to last year?. If they have an idea of theirperformance then it can be measured.

Handout 13

Dimensions of Quality

Effectiveness: Is the treatment given the best available in a technical sense,

according to those best equipped to judge?

What is their evidence? What is the overall result of the

treatment?

Acceptability: How humanely and considerately is this treatment/service

delivered? What does the patient think of it? What would/does

an observant third party think of it (“How would I feel if it were

my nearest and dearest?”) What is the setting like? Are

privacy and confidentiality safeguarded?

Efficiency: Is the output maximised for a given input or (conversely) is the

input minimised for a given level of output? How does the unit

cost compare with the unit cost elsewhere for the same

treatment/service?

Access: Can people get this treatment/service when they need it? Are

there any identifiable barriers to service - for example, distance,

inability to pay, waiting lists, and waiting times - or

straightforward

breakdowns in supply?

Equity: Is this patient or group of patients being fairly treated relative to

others? Are there any identifiable failings in equity - for

example, are some people being dealt with less favourably or

less appropriately in their own eyes than others?

Relevance: Is the overall pattern and balance of services the best that could

be achieved, taking account of the needs and wants of the

population as a whole?

Handout 14

Measuring Quality of Care

• What is meant by quality of care in primary care? Whose perspective should betaken into account?

• What evidence is available to indicate that good quality of care is being providedby the PHCT?

• What evidence is available to indicate that quality of care is not good?• Which aspects of the PHCT services / types of conditions is it most critical to

focus on? (ie which would improvements in service / care have the greatestimpact on team performance?)

• Decide on one aspect of quality of care and develop a measure.• Critically appraise the measure using the ‘criteria for measures’.

Select a scribe and someone willing to feedback in the plenary session.

Feedback on the process they went through to develop a measure, and the on themeasure/s developed

Handout 15

Basic ProMES Approach

• Gaining management and staff support • Set-up a design team • Identifying objectives • Develop measures for the objectives • Gather information with the measures • Feedback from the measures • Identify changes which can be made to improve performance.

Handout 16

Practical tips and guidelines

Conditions for success

• Management support - this means public support of the project on a regularbasis, providing the resources of the project regularly, solving project problemsand continuing the project until a clear evaluation can be made

• Trust between management and staff - process needs to be fully explainedincluding advantages and costs, explain why us?, make participation voluntary,explain how the whole team will be involved even though only part will be on thedesign team, explain how the design team will be chosen

• All interested parties approve the project - consider stakeholders, partners,unions

• Values match between management and staff - all see potential improvement asvaluable and have a long range perspective, all see participation/acceptance asessential

• Stable personnel and group structure - are there any major upheavals in staffpremises or technology

Handout 17

Practical tips and guidelines

Resources Needed for Development

• Time for design team meetings• Uninterrupted meeting setting• Full attendance by design team• Access to existing data• Meeting with management for approval of the system

Resources needed for implementation

• Collecting of existing and new data for feedback• Preparation and distribution of feedback results• Regular meetings of the group• Regular meetings of the group members to discuss feedback reports

Handout 18

Practical tips and guidelines

Checklist for starting a project

• All interested constituencies have been involved• Benefits and costs clearly explained to all• Have assessed trust and common values• Have management support (see number 2)• Have staff support (see number 2)

Handout 19

Practical tips and guidelines

Selecting the design team

• people who are respected• must be a cross section of the larger group e.g. practice nurse, senior

receptionist, general practitioner, district nurse, practice manager, health visitor• no-one occupational group should dominate• if someone is unable to attend they must nominate a deputy to represent them• individuals should feel confident about themselves and who/what they represent

and feel able to represent the views of their group

Handout 20

Practical tips and guidelines

Logging progress

• keep an accurate record of the progress made at each meeting• provide the design team with an update of their progress at regular intervals• you may find that between meetings the design team do not any homework and

you will need to review previous sessions at the beginning of new sessions• provide the team with the list of objectives when completed along with their list of

measures• the design team must also report progress to the other members of the larger

group and your updates is a useful way of achieving this• the team will need reassurance that progress is being made even if it is minor to

spur them on to the next stage

Handout 21

Practical tips and guidelines

Action planning

• at the end of each session you will need to plan for the actions to be carried outby the next session

• this may include collecting information and specific data• it must be clear from the outset that this is not the responsibility of the facilitator• members of the team will need to ‘volunteer’ to carry out actions• it is easy to walk away from a session believing that everyone knows what they

are soposed to do - it is more than likely someone does not• check and check until you are satisfied that everyone understands their

responsibilities• if you do this regularly the team will start to check you!• it is a good idea to have a session plan, you must be clear about what you want

to achieve at each session• also record the feedback the team give you at the beginning of the session, it

often proves useful later

Handout 22

Practical tips and guidelines

Possible problems

• lack of commitment from the design team - not present at meetings or lack ofresponse

• getting stuck - cant see how to move forward• design team going off on a tangent - avoiding what can be hard thinking work• disagreement between members - possible personal issues• domineering members - other members passively agreeing• feeling lost with it all - do not know what to do next• not been given the required time and space for the sessions or interrupted• larger team losing interest• management support withheld• no-one willing to collect data• got the data don’t know what to do next

Handout 23

Practical tips and guidelines

Gathering information

• the information gathered must be of good quality• the team must think about the most effective ways to gather information• you will need to discuss their expectations, what do they expect to be the

outcome• they may need to design record sheets or charts or questionnaires which must be

piloted to evaluate their effectiveness• who will distribute, collect, collate, evaluate and present the data to the design

team• these may all be new skills but is an excellent development opportunity• the team will need support in this, from you and from each other Assessing the value of the information • does the data answer the question• are you satisfied that the data is good quality• did you discuss how to asses the data before you began• do you have the resources to assess the data, people, time, technology

Handout 24

Practical tips and guidelines

Using information

• does this data provide you with useful information• you will need to discuss dissemination with the team• you must feedback the results of data gathering if you want people to co-operate

in the future• what next - if you have some useful information you can then use this to plan

ahead and use the information to make your case for change

____________________________________________________________________

BIBLIOGRAPHY___________________________________________________________

Adorian, D., Silverberg, D.S., Tomer, D. & Wamasher, Z. (1990). Group discussionwith the health care team: A method of improving care of hypertension in generalpractice.Journal of Human Hypertension, 4 (3), 265 - 268.

Alderfer, C.P. (1977). Group and intergroup relations in J.R. Hackman and J.L. Suttle(eds) Improving the quality of work life. Pallisades, C.A.: Good year. pp 277 - 296.

Alexander, J. A., Lichtenstein, R. & D’Aunno, T. A. (1996). The effects of treatmentteam diversity and size on assessments of team functioning. Hospital & HealthServices Administration, 41, 37-53.

Allen, N. J. (1996). Affective reactions to the group and organisation. In M. A. West(Ed.), Handbook of Work Group Psychology (pp. 371-396). Chichester: Wiley.

Amabile, T.M. (1983). The social psychology of creativity: A componentialconceptualization. Journal of Personality and Social Psychology, 45, 357-376.

Ancona, D.F. & Caldwell, D.F. (1988). Bridging the boundary: External activity andperformance in organisational teams. Administrative Science Quarterly, 37, 634-665.

Anderson, N.R. & King, N. (1993). Innovation in organisations. In C.L.Cooper &I.T. Robertson (eds). International Review of Industrial and OrganizationalPsychology,Vol 8, Chichester: Wiley.

Anderson, N, & West, M.A. (1994). The Team Climate Inventory: Manual and User’sGuide. Windsor, England: NFER-Nelson.

Anderson, N. & West, M.A. (1998). Measuring climate for work group innovation:development and validation of the team climate inventory: Journal of OrganizationalBehaviour, Vol 19, 235 - 258.

Applebaum, E. & Batt, R. (1994). The New American Workplace. Ithaca, NY: ILRPress.

Audit Commission (1992). Homeward Bound: A New Course for Community Health.London: HMSO.

Bales, R.F., Strodtbeck, F.L., Mills, T.M. & Roseborough, M.E. (1951). Channels ofcommunication in small groups. American Sociological Review, 16, 461-468.

Baumeister, R.F. & Leary M.R., (1995). The need to belong desire for interpersonalattachments as a fundemental motivator. Psychological Bulletin, 117, 497-529.

Berger, J., Fisek, M.H., Norman, R.Z., & Zelditch Jr, M. (1977). Status characteristicsand social interaction. NY: Elsevier.

Berger, J., Rosenholtz, S.J., & Zelditch Jr., M. (1980). Status organizing processes.Annual Review of Sociology, 6, 479-508.

Bhugra, D., Bridges, K., & Thompson, C. (1995). Caring for a community: thecommunity care policy of the Royal College of Psychiatrists. London: Royal Collegeof Psychiatrists.

Billings, R.S., Milburn, T.W. & Schaalman, M.L. (1980). A model of crisis perception:A theoretical and empirical analysis. Administrative Science Quarterly, 25, 300-316.

Blakar, R.M. (1985). Towards a theory of communication in terms of precondition:A conceptual framework and some empirical explorations. In H. Giles and R.N. StClaireds), Recent Advances in Language, Communication and Social Psychology.London: Lawrence Erlbaum.

Blau, Peter M. (1977). Inequality and Heterogeneity. New York: Free Press.

Bobko, P., & Colella, A. (1994). Employee reactions to performance standards:A review and research proposition. Personnel Psychology, 47, 1-29.

Bond, J., Cartilidge, A.M., Gregson, B.A., Philips, P.R., Bolam, F., & Gill, K.M. (1985).A study of interprofessional collaboration in primary health care organisations.Report No 27 (2), Newcastle-upon-Tyne, Health Care Research Unit, University ofNewcastle-upon-Tyne.

Borrill, C.S. & West, M.A. (1997). Effectiveness in primary health care. CAIPEbulletin.No 14.

Borrill, C.S. & West, M.A. (1998). Strain in primary health care. Unpublished report.Institute of Work Psychology, University of Sheffield, England.

Borrill, C.S., Wall. T.D., West, M.A., Hardy, G.E., Shapiro, D.A., Haynes, C.E., Stride,C.B., Woods, D. and Carter, A.J. (1998) Stress among staff in NHS Trusts. Instituteof Work Psychology, University of Sheffield, Psychological Therapies ResearchCentre, University of Leeds.

Bottger, P.C. & Yetton, P.W. (1987). Improving group performance by training inindividual problem solving. Journal of Applied Psychology, 72, 651-657.

Bowers, D.G. & Seashore, S.E. (1966). Predicting organisational effectiveness with afour-factor theory of leadership. Administrative Science Quarterly, 11, 238-263.

Brewer, N., Wilson, C. & Beck, K. (1994). Supervisory behavior and teamperformance amongst police patrol sergeants. Journal of Occupational andOrganizational Psychology, 67, 69-78.

Brown, R.J. (1988). Group Processes: Dynamics Within and Between Groups.London: Blackwell.

Bryk, A. & Raudenbush, S. (1992) Hierarchical Linear Models: Applications and dataanalysis methods. In J. Deleeuw (Ed.) Advanced quantitative techniques in thesocial sciences series. Newbury, CA: Sage Publications.

Burns, T., & Stalker, G.M. (1966). The Management of Innovation. London: TavistockPublications.

Carletta, J., Garrod, S., & Fraser-Krauss, H. (1998). Communication and Placementof Authority in Workplace Groups — The Consequences for Innovation. Small GroupResearch, 29(5), 531-559.

Carter, M.F., Evans, K.E., Crosby, C., Prendeergast, L.A., & De Sousa Butterworth,K.A. (1997). The all-Wales community mental health team survey. Bangor: HealthServices Research Unit.

Campion, M.A., Medsker, G.J. & Higgs, A.C. (1993). Relations between work groupcharacteristics and effectiveness: Implications for designing effective work groups.Personnel Psychology, 46, 823-850.

Campion, M.A., Papper, E.M. & Medsker, G.J. (1996). Relations between workteam characteristics and effectiveness: A replication and extension. PersonnelPsychology, 49, 429-689.

Cant, S. & Killoran, A. (1995) Team tactics: a study of nurse collaboration in generalpractice, Health Education Journal, 52, 203 - 8.

Coch, L. & French, J. R. (1948). Overcoming resistance to change. HumanRelations,1, 512-532.

Cohen, S.G. & Bailey, D.E. (1997) What makes teams work: Group effectivenessresearch from the shop floor to the executive suite. Journal of Management, No 3.239 - 290.

Connolly, T., Conlon, E.J. & Deutsch, S.J. (1980). Organizational effectiveness: amulti-disciple-constituency approach, 98, 310 - 357.

Cott, C. (1997). “We decide, you carry it out”: A social network analysis ofmultidisciplinary long-term care teams. Social Science & Medicine, 45 (9), 1411-1421.

Cowan, D.A. (1986). Developing a process model of problem recognition. Academyof Management Review, 11, 763-776.

D'Zumla, T.J. & Goldfried, M.R. (1971) Problem solving and behaviour modifications.Journal of Abnormal Psychology. 78, 107 - 126.

Davenport, T.H. (1993). Process Innovation: Re-engineering Work ThroughInformation Technology. Cambridge, MA: Harvard Business School Press.

Deming, W.E. (1986). Out of the Crisis. Cambridge, MA: Center for AdvancedEngineering Study, Massachusetts Institute of Technology.

Department of Health (1990). The Care Programme Approach for people with amental illness referred to the specialist psychiatric services. London, Department ofHealth, HC (90) 23/LASSL (90) 11.

Department of Health (1995). Building bridges: a guide to arrangements for inter-agency working for the care and protection of severely mentally ill people. London:HMSO.

Department of Health (1997). The Patient's Charter: mental health services. London,Department of Health.

Department of Health (1999). National Service Framework for Mental Health: Modernstandards and service models. London: Department of Health.

Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (2000). Workforce diversity: Implicationsfor the effectiveness of health care delivery teams. Social Science & Medicine, 50,1403-1414.

Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (1999a). Communication patterns andgroup composition: Implications for patient-centred care team effectiveness. Journalof Healthcare Management, 44, 252-268.

Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (1999b). Key indicators of nursing teamperformance: Insights from the front line. The Health Care Supervisor, 17, 70-76.

Drolen, C.S. (1990). Current community mental health center operations:Entrepreneurship or business as usual? Community Mental Health Journal, 26, 547-558.

Drory, A. & Shamir, B. (1988). Effects of organizational and life variables on jobsatisfaction and burnout. Group and Organization Studies, 13 (4), 441-455.

Eden, D. (1990). Pygmalion without interpersonal contrast effects: Whole groups gainfrom raising manager expectations. Journal of Applied Psychology, 75, 394-398.

Eggert, G.M., Zimmer, J.G., Hall, W.J. & Friedman, B. (1991). Case management:A randomised controlled study comparing a neighbourhood team and a centralizedindividual model. Health Services Research, 26 (4), 471-507.

Faulkner, A. (1997). Knowing our own minds: A survey of how people in emotionaldistress take control of their lives. London: Mental Health Foundation.

Fay, N., Garrod, S., & Carletta, J. (2000). Group discussion as interactive dialogue orserial monologue: The influence of group size. Psychological Science, 11(6), 487-492.

Field, R. & West, M.A. (1995). Teamwork in primary health care. Two Perspectivesfrom practices. Journal of Interprofessional Care, 9, 2, 123-130.

Freeman, M., Miller, C. & Ross, N. (2000). The impact of individual philosophies ofteamwork on multi-professional practice and the implications for education. Journal ofInterprofessional Care, 14 (3), 237-247.

Galbraith, J. R. (1993). The business unit of the future. In J. R. Galbraith, E.E. LawlerIII & Associates (Eds), Organizing for the Future: The New Logic for ManagingComplex Organizations. San Francisco: Jossey-Bass.

Galbraith, J. R. (1994). Competing with Flexible Lateral Organisations (2nd edn).Reading, MA: Addison-Wesley.

Galbraith, J.R., Lawler, E.E. III & Associates (1993). Organizing for the Future: TheNew Logic for Managing Complex Organizations. San Francisco: Jossey-Bass.

George, J.M. (1989). Mood and absence. Journal of Applied Psychology, 74, 317-324.

George, J.M. (1990). Personality, affect, and behavior in groups. Journal of AppliedPsychology, 75, 107-166.

George, J.M. (1995). Leader positive mood and group performance: The case ofcustomer service. Journal of Applied Social Psychology, 25, 778-794.

George, J.M. (1996). Group affective tone. In M.A. West (Ed.), Handbook of WorkGroup Psychology (pp. 77-94). Chichester: Wiley.

George, J.M. & Bettenhausen, K. (1990). Understanding pro-social behaviour, salesperformance and turnover. A group-level analysis in a service context. Journal ofApplied Psychology, 75, 698-709.

Gladstein, D. (1984). Groups in context: A model of task group effectiveness.Administrative Science Quarterly, 29, 499-517.

Goldberg, D.P. & Williams, P. (1991). A user's guide to the General HealthQuestionnaire. Windsor: NFER-Nelson.

Goldberg. D.P. (1972). The detection of minor psychiatric illness by questionnaire.Oxford: Oxford University Press

Goni. S. (1999). An analysis of the effectiveness of Spanish primary health careteams. Health Policy, 48, 107-117.

Grusky, O. (1995). The organization and effectiveness of community mental healthsystems. Administration & Policy in Mental Health, 22, 361-388.

Guzzo, R.A. (1996). Fundamental considerations and about work groups. In M.A.West (ed) Handbook of Work Group Psychology. Chichester: John Wiley.

Guzzo, R. A. & Dickson, M. W. (1996). Teams in organisations: Recent research onperformance and effectiveness. Annual Review of Psychology, 46, 307-338.

Guzzo, R.A., Jette, R.D., & Katzell, R.A. (1985). The effects of psychologically basedintervention programs on worker productivity: A meta-analysis. PersonnelPsychology, 38, 275-291.

Guzzo, R. A. & Salas, E. (Eds) (1995). Team Effectiveness and Decision Making inOrganisations. San Francisco: Jossey-Bass.

Guzzo, R.A. & Shea, G.P. (1992). Group performance and intergroup relations inorganisations. In M. D. Dunnette and L. M. Hough (Eds), Handbook of Industrial andOrganizational Psychology, (Vol 3, pp. 269-313). Palo Alto, CA: Consulting

Psychologists Press.

Hackman, J.R. (1987). The design of work teams. In J.W. Lorsch (ed) Handbook of

Organisational Behaviour. Englewood Cliffs, NJ: Prentice-Hall. Pp 315 - 342.

Hackman, J.R. (1990). (Ed), Groups That Work (and Those That Don't): CreatingConditions for Effective Teamwork. San Francisco: Jossey-Bass.

Hackman, J.R., Brousseau, K.R. & Weiss, J.A. (1976). The interaction of task designand group performance strategies in determining group effectiveness.Organizational Behavior and Human Performance, 16, 350-365.

Hackman, J.R. & Morris, C.G. (1975). Group task, group interaction process, andgroup performance effectiveness: A review and proposed integration. In L. Berkowitz(Ed.), Advances in Experimental Social Psychology, (Vol. 8). New York: AcademicPress.

Hannigan, B. (1999). Joint working in community mental health: prospects andchallenges. Health and Social Care in the Community, 7, 25-31.

Hardy, G. E., Shapiro, D.A., Hayes, C.E. & Rick, J.E. (1999). Validation of theGeneral Health Questionnaire using a sample of employees from the health careservices. Submitted.

Haynes, C.E., Wall, T.D., Bolden, R.I. & Rick, J.E. (1998). Measures of perceivedwork characteristics for health service research: test of a measurement model andnormative data. Submitted.

Hedburgh, B.L.T., Nystrom, P.C. and Starbuck, W.H. (1976) Company on seesaws:prescriptions for a self designing organisation. Administrative Science Quarterley,21, 41 - 65.

Hill, F. (1998). Trying to catch a cloud: organizational climate in the NHS.Unpublished PhD Thesis, Institute of Work Psychology, University of Sheffield.

Hirokawa, R.Y. (1990). The role of communication in group decision-making efficacy:A task-contingency perspective. Small Group Research, 21, 190-204.

Hoffman, L.R. & Maier, N.R.F. (1961). Sex differences, sex composition, and groupproblem-solving. Journal of Abnormal and Social Psychology, 63, 453-456.

Hogg, M. & Abrams, D. (1988). Social Identifications: A Social Psychology ofIntergroup Relations and Group Processes. London: Routledge.

Hughes, S.L., Cummings, J., Weaver, F., Manheim, L., Brawn, B. & Conrad, K.(1992).

A randomised trial of the cost effectiveness of VA hospital-based home care for theterminally ill. Health Services Research, 26 (6), 801-817.

Jackson, G., Gater, R., Goldberg, D., Tantam, D. Loftus, L. & Taylor, H. (1993). Anew community mental health team based in primary care: A description of theservice and its effect on service use in the first year. British Journal of Psychiatry,162, 375-384.

Jackson, L.A., Sullivan, L.A. & Hodge, L.N. (1993). Stereotype effects onattributions, predictions and evaluations: No two social judgements are quite alike.Journal of Personality and Social Psychology, 65 (1), 69-84.

Jackson, S.E. (1996). The consequences of diversity in multidisciplinary work teams.InM.A. West (Ed.), Handbook of Work Group Psychology, pp. 53-76, Chichester: Wiley.

Jacobs, D. & Singell, L. (1993). Leadership and organizational performance:Isolating links between managers and collective success. Social Science Research,22, 165-189.

Janis, I.L. (1982). Groupthink: A Study of Foreign Policy Decisions and Fiascos, 2nded. Boston: Houghton Mifflin.

Jansson, A., Isacsson, A. & Lindhom, L.H. (1992). Organization of health care teamsand the population’s contacts with primary care. Scandinavian Journal of HealthCare, 10,257-265.

Jervis, I.L. (1976). Perception and Misperception in International Politics. Princeton,NJ: Princeton University Press.

Jones, R.V.H. (1992). Teamworking in primary care:how do we know about it?Journal of Interprofessional Care, Vol 6, p25-29.

Juran, J.M. (1989). Juran on Leadership for Quality. New York: Free Press.

Jussim, L. (1986). Self-fulfilling prophecies: A theoretical and integrative review.Psychological Review, 93 (1), 429-445.

Jussim, L., Coleman, L.M. & Lerch (1987). The nature of stereotypes: A comparisonand integration of three theories. Journal of Personality and Social Psychology, 52(3), 536-546.

Kalleburg, A.L. & Moody, J.W. (1994). Human Resource Management andOrganisational Performance. American Behaviourist Scientist, 37, 948 - 962.

Kanter, R.M. (1983). The Change Masters: Corporate Entrepreneurs at work.New York: Simon & Schuster.

Kiesler, S. & Sproull, L. (1982). Managerial responses to changing environments:perspectives in problem solving from social cognition. Administrative ScienceQuarterley, 27, 548 - 570.

Kimble, C.E., Marsh, N.B. & Kiska, A.C. (1984). Sex, age and cultural differences inself-reported assertiveness. Psychological Reports, 55, 419-422.

King, N. (1990). Innovation at work: The research literature. In M. A. West & J. L.Farr (Eds), Innovation and Creativity at Work: Psychological and OrganisationalStrategies, pp. 15-59, Chichester: Wiley.

King, N., Anderson, N.R. & West, M. (1991). Organizational innovation in the UK: Acase study of perceptions and processes. Work and Stress, 5 (4), 331-339.

King, R., Le Bas, J., & Spooner, D. (2000). The impact of caseload on the personalefficacy of mental health case managers. Psychiatric Services, 51, 364-368.

Kinnunen, J. (1990). The importance of organizational culture on developmentactivities in a primary health care organisation. International Journal of HealthPlanning and Management, 5, 65-71.

Klimoski, R. & Mohammed, S. (1994). Team mental model: Construct or metaphor?Journal of Management, 20, 403-437.

Komaki, J.L., Desselles, M.L. & Bowman, E.D. (1989). Definitely not a breeze:Extending an operant model of effective supervision to teams. Joumal of AppliedPsychology, 74, 522-529.

Koshuta, M. & McCuddy, M.K. (1989). Improving productivity in the health careindustry:An argument and supporting evidence from one hospital. The Health CareSupervisor.

LaFrance, M. & Mayo, C. (1978). Moving Bodies: Nonverbal Communication inSocial Relationships. Monterey, C.A.: Brooks/Cole.

Landsberger, H.A. (1955). Interaction process analysis of the mediation oflabormanagement disputes. Journal of Abnormal and Social Psychology, 51, 522-528.

Latham, G.P., Erez, M. & Locke, E.A. (1988). Resolving scientific disputes by thejoint design of crucial experiments by the antagonists: Application to the Erez Lathamdispute regarding participation in goal setting. Journal of Applied Psychology, 73 (4),753-772.

Lawler, E.E. & Hackman, J.R. (1969). Impact of employee participation indevelopment of pay incentive plans: A field experiment. Journal of AppliedPsychology, 53, 467-471.

Lawrence, P.R. & Lorsch, J. (1967). Organization and Environment. Cambridge,NIA:Harvard University Press.

Macy, B.A. & lzumi, H. (1993). Organizational change, design and work innovation:A meta-analysis of 131 North American field studies-1961-1991. Research inOrganizational Change and Design (Vol. 7). Greenwich, CT: JAI Press.

Maier, N.R.F. (1963). Problem-solving Discussions and Conferences: LeadershipMethods and Skills. New York: McGraw-Hill.

Maier, N.R.F. (1970). Problem Solving and Creativity in Individuals and Groups.Monterey, CA: Brooks/Cole.

Maier, N.R.F. & Solem, A.R. (1962). Improving solutions by turning choice situationsinto problems. Personnel Psychology, 15, 151-157.

Main, J. (1989). At last, software CE0s can use. Fortune, 13 March, 77 - 83.

Markiewicz, L. & West, M.A. (1996). Team-based Organisation. Aberdeen:Grampian/ECITB.

Markiewicz, L. & West, M.A. (1996). Team-based Organisation. Aberdeen:Grampian/ECITB.

Mathison, D.L. & Tucker, R.K. (1982). Sex differences in assertive behaviour:A research extension. Psychological Reports, 51(3), 943-948.

Maxwell, R.J. (1992). Dimensions of quality revisited: from thought to action. QualityHealth Care 1, 171-177.

Maznevski, M.L. (1994). Understanding our differences: Performance in decisionmaking groups with diverse members. Human Relations, 47 (5), 531-552.

McClure, L.M. (1984) Teamwork, myth or reality: community nurses effectivenesswith general practice attachment. Journal of Epidemiology and Community Health, 21(1),68 - 74.

McGrath, J.E. (1984). Groups: Interaction and Performance. Englewood Cliffs, NJ:Prentice-Hall.

Miceli, M.P. & Near, J.P. (1985). Characteristics of organisational climate andperceived wrong-doing associated with whistle-blowing decisions. PersonnelPsychology, 38,525-544.

Milliken, F.J. & Martins, L.L. (1996). Searching for common threads: Understandingthe multiple effects of diversity in organizational groups. Academy of ManagementReview, 21(2), 402-433.

Mistral, W., & Velleman, R. (1997). Community mental health teams: Theprofessionals' choice? Journal of Mental Health, 6, 125-140.

Mitrot, I.I. & Featheringham, T.R. (1974) On systematic problem solving and the errorof the third kind. Behavioural Science, 19, 383 - 393.

Mohman, S.A., Cohen, S.G. & Mohrman, A.M., Sr (1995). Designing Team-BasedOrganizations. San Francisco: Jossey-Bass.

Mullarkey, S., Wall, T.D., Warr, P.B., Clegg, C.S. & Stride, C. (1999). Measures ofjob satisfaction, mental health and job-related well-being: A bench-marking manual.Sheffield, England: Sheffield Academic Press Ltd.

Mumford, M.D. & Gustafson, S.B. (1988). Creativity syndrome: Integration,application and innovation. Psychological Bulletin, 103, 27-43.

Myer, C. (1993). How to Align Purpose, Strategy and Structure for Speed. NewYork: Free Press.

Naylor, J.C., Pritchard, R.D., & Ilgen, D.R. (1980). A theory of behaviour inorganisations. New York:Academic Press.

Netten, A. & Dennett, J. (1997). Unit lists of health and social care. London: PSSRU.

Nievaard, A.C. (1987). Communication climate and patient care: Causes and effectsof nurses’ attitudes to patients. Social Science and Medicine, 24 (9), 777-784.

Onyett, S. (1995). Responsibility and accountability in community mental healthteams. Psychiatric Bulletin, 19, 281-285.

Onyett, S. (1997). The challenge of managing community mental health teams.Health and Social Care in the Community, 5, 40-47.

Onyett, S., Pillinger, T. & Muijen, M. (1995). Making community mental health teamswork. London: Sainsbury Centre for Mental Health.

Pearce, J.A. & Ravlin, E.C. (1987). The design and activation of self-regulating workgroups. Human Relations, 40, 751-782.

Peck, E., & Norman, I.J. (1999). Working together in adult community mental healthservices: Exploring inter-professional role relations. Journal of Mental Health, 8,231-243.

Peck, E., & Parker, E. (1998). Mental health in the NHS: Policy and practice 1979-98.Journal of Mental Health, 7, 241-259.

Peiro, J.M., Gonzalez-Roma, V., & Romos, J. (1992) The infleince of work teamclimate on role stress, tension, satisfaction and leadership perceptions. EuropeanReview of Applied Psychology, 42 (1) 49-46.

Pincus, H.A., Zarin, D.A. & West, J.C. (1996). Peering into the 'black box': measuringoutcomes of managed care. Archives of General Psychiatry, 53, 870-877.

Podsakolf, P.M. & Todor, W.D. (1985). Relationships between leader reward andpunishment behavior and group processes and productivity. Journal of Management,11, 55-73.

Porac, J. F. & Howard, H. (1990). Taxonomic mental models in competitor definition.Academy of Management Review, 2, 224-240.

Poulton, B. C. & West, M. A. (1993). Effective multidisciplinary teamwork in primaryhealth care. Journal of Advanced Nursing, 18, 918-925.

Poulton, B. C. & West, M. A. (1994). Primary health care team effectiveness:Developing a constituency approach. Health and Social Care, 2, 77-84.

Poulton, B.C. & West, M.A. (1997). A failure of function: teamwork in primary healthcare. Journal of Interprofessional Care, 11, No 2, 1997.

Pritchard, R. D., Jones, S. D., Roth, P. L., Stuebing, K. K. & Ekeberg, S. E. (1988).Effects of group feedback, goal setting, and incentives on organizational productivity.Journal of Applied Psychology, 73, 337-358.

Pritchard, R.D. (ed) (1995) Productivity Measurement and Improvement:Organisational Case Studies. New York: Praeger.

Pritchard, R.D., Jones, S.D., Roth, P., Stuebing, K.K., & Ekeberg, S.E. (1989). Theevaluation of an integrated approach to measuring organisational productivity.Personnel Psychology, 42 (1) 69-115.

Pritchard, R.D. (Ed) (1995). Productivity Measurement and Improvement:Organisational Care Studies: New York: Praeger.

Pritchard, R.D., Jones, S.D., Roth, P.L., Stuebing, K.K. & Ekeberg, S.E. (1988). Theeffects of feedback, goal setting, and incentives on organisational productivity.Journal of Applied Psychology Monograph Series, 73 (2), 337-358.

Prosser, D., Johnson, S., Kuipers, E., Szmukler, G., Bebbington, P., & Thornicroft, G.(1996). Mental Health, 'Burnout' and job satisfaction among hospital and community-based mental health staff. British Journal of Psychiatry, 169, 334-337.

Quinn, R.E. & Rohrbaugh, J. (1983). Predicting sales success through handwritinganalysis: an evaluation of the effects of training and handwriting sample content.Journal of Applied Psychology 68 (2), 212 - 17.

Redmond, M.V. (1989). The functions of empathy (decentering) in human relations.Human Relations, 42, 593-605.

Redmond. M.V. (1992). A multi-dimensional theory and measure of decentering.Unpublished manuscript.

Reed J. (1995). Leadership in the mental health service: What role for doctors?Psychiatric Bulletin, 19, 67-72.

Rees, A., Stride, C.B., Shapiro, D.A., Richards, A. and Borrill, C.S. (in press).Psychometric properties of the Community Health Team Questionnaire. (CMHTEQ)Journal of Mental Health.

Richards, A. & Rees, A. (1998) Developing Criteria to measure the effectiveness ofcommunity mental health teams. Mental Health Care, 2, 14 - 17.

Roberts, J.M. (1995) The History of the World. Hammondsworth, Middlesex:Penguin.

Ross, F., Rink, E. & Furne, A. (2000). Integration or pragmatic coalition? Anevaluation of nursing teams in primary care. Journal of Interprofessional Care, 14 (3),259-267.

Rousseau, D.M. (1985) Issues of level in organisational research: multi-level andcross-level perspectives. Research in Organisational Behaviour, 7, 1 - 37.

Sainsbury Centre for Mental Health (1997). Pulling together: The future roles andtraining of mental health staff. London: Sainsbury Centre for Mental Health.

Schober, M.F., & Clark, H.H. (1989). Understanding by addressees and overhearers.Cognitive Psychology, 21, 211-232.

Schwenk, C.R. (1988). The Essence of Strategic Decision-making. Cambridge, NIA:Heath.

Senge, P. (1990). The Fifth Discipline: The Art and Practice of the LearningOrganization. New York: Doubleday Currency.

Shaw, M.E. (1976). Group Dynamics: The Psychology of Small Group Behavior. NewYork: McGraw-Hill.

Shaw, M.E. (1981). Group Dynamics: The Psychology of Small Group Behavior.New York: McGraw-Hill.

Sluyter, G.V. (1995). Mental health leadership training: A survey of state directors.Journal of Mental Health Administration, 22, 201-204.

Smircich, L. (1983). Organization as shared meaning. In L.R. Pondy, P. Frost, G.Morgan & T. Dandridge (Eds), Organizational Symbolism (pp. 55-65). Greenwich,CT: JAI Press.

Smith, K.G., Locke, E.A. & Barry, D. (1990). Goal setting, planning andorganizational performance: An experimental simulation. Organizational Behaviorand Human Decision Processes, 46, 118-134.

Sommers, L.S., Marton, K.I., Barbaccia, J.C. & Randolph, J. (2000). Physician, nurseand social worker collaboration in primary care for chronically ill seniors, Archives ofInternal Medicine, 160, 1825-1833.

Stein, M. (1996). Unconscious phenomena in work groups. In M.A. West (Ed.),Handbook of Work Group Psychology, pp. 143-157 Chichester: Wiley.

Stemberg, R.J. & Lubart, T.I. (1990) Defying the Crowd. Cultivating Creativity in aCulture of Conformity. New York: Free Press.

Stevens, M.J., & Campion, M.A. (1994). The knowledge, skill and abilityrequirements for teamwork: Implications for human resource management. Journal ofManagement, 20,503-530.

Stevens, M.J. & Campion, M.A. (1999). Staffing Work Teams: Development andValidation of a Selection Test for Teamwork Settings. Journal of Management, 25,No 2 207 - 228.

Sundstrom, E., De Meuse, K.P. & Futrell, D. (1990). Work teams: Applications andeffectiveness. American Psychologist, 45, 120-133.

Tajfel, H. (1978). Differentiation Between Social Groups: Studies in the SocialPsychology of Intergroup Relations (European Monographs in Social Psychology,No. 14). London:Academic Press.

Tajfel, H. & Turner, J.C. (1979). An integrative theory of intergroup conflict.In W. G. Austin and S. Worchel (Eds), The Social Psychology of IntergroupRelations. Monterey, CA: Brooks/Cole.

Tannenbaum, S.I., Beard, R.L. & Salas, E. (1992). Team building and its influenceon team effectiveness: An examination of conceptual and empirical developments.In K. Kelley (Ed.), Issues, Theory and Research in Industrial/OrganizationalPsychology (pp. 117-153), London: North Holland.

Tannenbaum, S.I., Salas, E. & Cannon-Bowers, J.A. (1996). Promoting teameffectiveness. In M.A. West (Ed.), Handbook of Work Group Psychology, (pp 503-529)Chichester: Wiley.

Taylor, M.F. (Ed), with Brice, J., Buck, N. & Prentice, E. (1995). British HouseholdPanel Survey User Manual. Colchester: University of Essex.

Tjosvold, D. (1982). Effects of approach to controversy on superiors' incorporation ofsubordinates' information in decision making. Joumal of Applied Psychology, 67,189-193.

Tjosvold, D. (1985). Implications of controversy research for management. Journalof Management, 11, 21-37.

Tjosvold, D. (1991). Team Organisation: An Enduring Competitive Advantage.Chichester: Wiley.

Tjosvold, D. & Field, R.H.G. (1983). Effects of social context on consensus andmajority vote decision making. Academy of Management Journal, 26, 500-506.

Tjosvold, D. & Johnson, D.W. (1977). The effects of controversy on cognitiveperspective-taking. Journal of Educational Psychology, 69, 679-685.

Tjosvold, D., Wedley, W.C. & Field, R.H.G. (1986). Constructive controversy, theVroom-Yetton model, and managerial decision making. Journal of OccupationalBehavior, 7,121 -138.

Toon, P.D. (1994) What is Good General Practice? A Philosophical Study of theConcept of High Quality Medical Care.

Tziner, A.E. (1988) Effects of team composition on ranked team effectiveness. SmallGroup Behaviour, 19, 363 - 378.

Vroom, V.H. (1964). Work and Motivation. New York: Wiley.

Wall, T.D. & Lischeron, J.H. (1977). Worker Participation: A Critique of the Literatureand some Fresh Evidence. Maidenhead, UK: McGraw-Hill.

Walsh, J.P. & Fahey, L. (1986). The role of negotiated belief structures in strategymaking. Journal of Management, 12, 325-338.

Walsh, J.P., Henderson, C.M. & Deighton, J. (1988). Negotiated belief structuresand decision performance: An empirical investigation. Organizational Behavior andHuman Decision Processes, 42, 194-216.

Watson, W.E., Kumar, K. & Michaelsen, L.K. (1993). Cultural diversity's impact oninteraction process and performance: Comparing homogeneous and diverse task

groups. Academy of Management Journal, 36, 590-602.

Weldon, E. & Weingart, L.R. (1993). Group goals and group performance. BritishJournal of Social Psychology, 32, 307-334.

West, M.A. (1990). The social psychology of innovation in groups. In M.A. West &J.L. Farr (Eds), Innovation and Creativity at Work (pp. 309-333). Chichester: Wiley.

West, M.A. (1994). Effective Teamwork. Leicester: British Psychological Society.

West, M.A. (1996a). The Handbook of Work Group Psychology. Chichester: Wiley.

West, M.A. (1996b). Reflexivity and Work Group Effectiveness: A conceptualIntegrationin M.A. West (Ed) Handbook of Work Group Psychology (555 - 580), John Wiley &Sons Ltd.

West, M.A. (1997). Developing Creativity in Organisations. Chichester: Wiley.

West, M.A. (2000). State of the art: Creativity and Innovation at work. Psychologist,13, 9 460 - 464.

West, M.A. & Anderson, N.R. (1996). Innovation in top management teams. Journalof Applied Psychology, 81(6), 680-693.

West, M.A., Borrill, C.S. and Stride, C.B. (1998). Strain as a moderator of therelationship between work characteristics and work attitudes. Journal of OccupationalHealth. Vol 4,No 1, 3 - 14.

West, M.A., Borrill, C.S. & Unsworth, K. (1998). Team Effectiveness in Organisations.In C.L. Cooper & I.T. Robertson (eds) International Review of IndustrialOrganisational Psychology. Vol 13. Wiley & Sons: Chichester.

West, M.A. & Field, R. (1995). Teamwork in primary health care. Perspectives fromorganisational psychology. Journal of Interprofessional Care, 9, 2, 117-122.

West, M.A., & Poulton, B.C. (1995). Primary health care teams: Rhetoric versusreality. Paper submitted for publication. Institute of Work Psychology, University ofSheffield.

West, M.A. & Slater,J.A. (1996). The Effectiveness of Team Working in PrimaryHealth Care. London:Health Education Authority.

West, M.A. & Wallace, M. (1991). Innovation in health care teams. European Journalof Social Psychology. Vol 21, 303 – 315.

West, M.A. & Pillinger, T. (1995). Innovation in UK manufacturing (Research report).Institute of Work Psychology, University of Sheffield.

Wood, N., Farrow, S., & Elliott, B. (1994). A review of primary health careorganisation. Journal of Clinical Nursing, 3(4), 243-250.

Worchell, S., Wood, W. & Simpson, J.A. (eds) (1992). Group Processes andProductivity. Newbury Park, CA: Sage.

Yeatts, D.E. & Seward, R.R. (2000). Reducing turnover and improving health care innursing homes: The potential effects of self-managed work teams. The Gerontologist,40 (3),358-363.

Zimmer, J.G., Eggert, G.M. & Chiverton, P. (1990). Individual versus team casemanagement in optimising care for chronically ill patients with dementia. Journal ofAging and Health, 2 (3), 357-372.