Embed Size (px)
Citation preview
Journal of Safety Research 42 (2011) 443–451
Contents lists available at SciVerse ScienceDirect
Journal of Safety Research
j ourna l homepage: www.e lsev ie r .com/ locate / j s r
Literature Review
Implementing the evidence for preventing falls among community-dwelling olderpeople: A systematic review
Victoria Goodwin a,⁎, Tracey Jones-Hughes b, Jo Thompson-Coon a, Kate Boddy a, Ken Stein a,b
a PenCLAHRC, Peninsula College of Medicine and Dentistry, University of Exeter, Veysey Building, Salmon Pool Lane, Exeter, UK, EX2 4SGb PenTAG, Peninsula College of Medicine and Dentistry, University of Exeter, Veysey Building, Salmon Pool Lane, Exeter, UK, EX2 4SG
⁎ Corresponding author. Tel.: +44 1392 262745; fax:E-mail address: [email protected] (V. Goo
0022-4375/$ – see front matter © 2011 National Safetydoi:10.1016/j.jsr.2011.07.008
a b s t r a c t
a r t i c l e i n f oArticle history:
Received 30 September 2010Received in revised form 13 July 2011Accepted 28 July 2011Available online 10 November 2011Keywords:Falls preventionImplementationOlder adultsEvidence-based practiceSystematic review
Problem and objective: The translation of the evidence-base for preventing falls among community-dwellingolder people into practice has been limited. This study systematically reviewed and synthesised the effective-ness of methods to implement falls prevention programmes with this population.Methods: Articles publishedbetween 1980 and May 2010 that evaluated the effects of an implementation strategy. No design restrictionswere imposed. A narrative synthesis was undertaken. Results: 15 studies were identified. Interventions thatinvolved the active training of healthcare professionals improved implementation. The evidence aroundchanging the way people who fall are managed within primary care practices, and, layperson, peer or com-munity delivered models was mixed. Impact on industry: Translating the evidence-base into practice involveschanging the attitudes and behaviours of older people, healthcare professionals and organisations. However,there is a need for further evaluation on how this can be best achieved.
© 2011 National Safety Council and Elsevier Ltd. All rights reserved.
1. Introduction
Falls are an increasing public health concern, affecting a third of peo-ple aged 65 and over. It has been estimated that even if age-adjusted in-cidence rates remain stable, the number of hip fracturesworldwidewillclimb from 1.66 million in 1990 to 6.26 million in 2050 (Sambrook &Cooper, 2006). This rising trend exists despite many high quality re-views and clinical guidelines providing evidence for the prevention offalls among community-dwelling older people (American GeriatricsSociety and the British Geriatrics Society, 2010; Gillespie et al., 2009;National Institute for Health Clinical Excellence, 2004). However, oncloser examination it is apparent that this evidence base has not neces-sarily been transferred into clinical practice (Royal College of Physicians,2007; Tinetti, Gordon, Sogolow, Lapin, & Bradley, 2006). As such, fallsand fall-related injuries continue to escalate (Department of Health,2009) with a less than optimal provision of evidence-based healthcare(Goodwin et al., 2010).
One aspect of this problem originates from the lack of understandingon how to effectively implement the evidence-base, particularly whereroutine practice may be in contrast to the experimental conditions ob-served in the original research (Roen, Arai, Roberts, & Popay, 2006).For example, clinicians and patientsmay be required to change behaviorand adopt new practices; and organizationsmay be required to develop
+44 1392 421009.dwin).
Council and Elsevier Ltd. All rights
alternative systems of working across professional and organizationalboundaries (Rose, Alkema, Choi, Nishita, & Pynoos, 2007; Tinetti et al.,2006). Known barriers to implementation of falls prevention strategiesinclude (Tinetti et al., 2006):
• Time;• Lack of knowledge and skills;• Complex health and social issues;• Service organization issues, such as fragmentation or a lack of co-ordination; and
• Financial issues.
Facilitators of successful implementation are (Ganz, Alkema, &Wu, 2008; Greenhalgh, Robert, Macfarlane, Bate, & Kyriakidou, 2004):
• Effective leadership and collaboration;• Strategies adopting simpler interventions;• Benefits of the intervention to be observable by those intending toadopt the intervention; and
• An approach which can be adapted to meet the needs of organiza-tions and practitioners.
We therefore performed a systematic review of studies in whichthe implementation of a falls prevention strategy has been evaluated.We identify and explore the existing evidence base, and attempt toidentify key factors for successful implementation of falls preventionstrategies.
reserved.
1 Accidental Falls/2 (fall or falls or faller$1 or fallen).ti,ab.3 1 or 24 exp Aged/5 (senior$1 or elder* or older or old or oldest).ti,ab.6 4 or 57 3 and 68 (prevent* or reduce* or manage*).ti,ab.9 7 and 810 Program Evaluation/11 Information Dissemination/12 Barrier*.ti,ab.13 evaluat*.ti,ab.14 translat*.ti,ab.15 feasibility.ti,ab.16 integrat*.ti,ab.17 implement*.ti,ab.18 disseminat*.ti,ab.19 adopt*.ti,ab.20 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 1921 9 and 2022 limit 21 to yr="1980 -Current"
Fig. 1. Master search strategy written for Medline (OVID) and adapted for differentdatabases.
444 V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
2. Methods
The systematic review was conducted according to a predefinedprotocol that was developed following consultation with experts inthe field and is available from the authors on request.
2.1. Literature search and eligibility criteria
By analysis of key studies, we devised a search strategy to identifyrelevant papers capturing the process of implementation in themanage-ment of accidental falls among older people (Fig. 1). No methods filterwas applied. Themaster search strategy was adapted and run in the fol-lowing electronic databases from 1980 toMay 2010: AMED and CINAHL(Using the EBSCO interface); Cochrane Database of Systematic Reviews;CENTRAL; Medline; Embase and Psychinfo (Using the OVID interface);and the Social Sciences Citation Index.We scrutinized the bibliographiesof included studies and of other identified relevant review papers in thesearch for additional articles.
Studies were included if they reported the evaluation of an imple-mentation strategy for the prevention of falls among community-dwelling older adults. Outcomes could include, for example, behaviorchange, attitudes, and uptake of recommendations. Studies were ex-cluded if they only reported health outcomes, such as fractures orhealthcare utilization. There were no restrictions on study design. Ed-itorials, opinion papers, and studies reported only as conference ab-stracts were excluded. Only papers published in the Englishlanguage were included in the review.
Two reviewers independently screened all titles and abstracts. Fulltext manuscripts of any relevant titles/abstracts were obtained andthe relevance of each study was assessed according to the inclusionand exclusion criteria. Studies that did not fulfill the criteria were ex-cluded and their bibliographic details listed with the reason for exclu-sion. Any discrepancies were resolved by consensus and, wherenecessary, a third reviewer was consulted.
2.2. Data extraction and quality assessment
Data were extracted from included papers independently by tworeviewers using a standardized, piloted data extraction form. The fol-lowing data were extracted: study location and setting, study design,implementation method, fall prevention intervention, study popula-tion, outcomes and follow up, analysis and results.
The quality of individual studies were assessed independently bytwo reviewers using the Cochrane risk of bias tool (Higgins & Green,2009). The tool includes six key criteria against which potential riskof bias is judged. These being:
• Was the allocation sequence adequately generated and described toenable the assessment of whether it would produce comparablegroups following randomization?
• Was the allocation adequately concealed anddescribed in enoughdetailto determine whether allocation of research participants could havebeen predicted before or during recruitment by research personnel?
• Were participants, personnel or outcome assessors adequatelyblinded to allocation during the study, what methods were usedand were they successful?
• Were incomplete outcome data, such as exclusions, attrition, ormissingdata reported,with reasons and how thesewere dealtwith in analyses?
• Was the study free of suggestion of selective outcome reporting (e.g., bypre-specifying outcomes and analyses of interest and reporting these)?
• Was the study apparently free from other problems that could put itat risk of bias, such as study design, extreme baseline imbalances?
The results were tabulated by individual reviewers for each studyand compared. Disagreements were resolved through consensus in-volving a third reviewer where necessary.
2.3. Data synthesis
To determine whether effective methods of implementation wereconsistent across studies, data were summarized using evidence ta-bles and synthesized using a narrative approach. Where data allowed,relationships and differences between studies were identified basedon factors such as healthcare system, professions involved, or the na-ture of the implementation method.
3. Results
3.1. Search results and study characteristics
A total of 3,638 unique titles and abstracts were identified from thesearch following removal of duplicates (Fig. 2); 3,530 studies were ex-cluded following a review of titles and abstracts as notmeeting the inclu-sion criteria. A full-text assessment of 108 articles resulted in theexclusion of 93 studies (7 did not target community-dwelling older peo-ple; 76 did not evaluate implementation; 6 were opinion papers, 3 wereonly available as abstracts, and 1 paper was not available in English).The remaining 15 studies met the selection criteria and were includedin the review.
Six studies were undertaken in the United States (Baraff, Lee, Kader, &Penna, 1999; Brown, Gottschalk, Van Ness, Fortinsky, & Tinetti, 2005;Fortinsky et al., 2008; Healy, Haynes, McMahon, Botler, & Gross, 2005;Shah, Maly, Frank, Hirsch, & Reuben, 1997; Wenger et al., 2009), four inAustralia (Barnett et al., 2004; Deery, Day, & Fildes, 2000; McClure et al.,2010; Stackpool, 2006), and one each in Canada (Scott, Votova, &Gallagher, 2006), New Zealand (Gardner, Robertson, McGee, &Campbell, 2002), Sweden (Larsson, Hägvide, Svanborg, & Borell, 2010),Belgium (Milisen, Geeraerts, & Dejaeger, 2009), and Hong Kong (Sze,Lam, Chan, & Leung, 2005). A variety of study designs were utilized in-cluding a non-randomized controlled trial (n=1), cross-sectional studies(n=3), cohort studies (n=4), surveys (n=5), process evaluation(n=1), and a case series (n=1).
3.2. Assessment of study quality
When examining the quality of each study (Table 1), all werefound to be at a high risk of bias. In terms of blinding, six studies
445V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
(Deery et al., 2000; McClure et al., 2010; Scott et al., 2006; Shah etal., 1997; Sze et al., 2005; Wenger et al., 2009) did not provide aclear indication as to whether participants or outcome assessorswere blinded. As all but one study (Wenger et al.) did not includea comparator group, participants were aware of the interventionand, where outcomes were self-reported, this may result in poten-tial reporting bias. In most cases, we were unable to ascertainwhether all collected outcome data were reported. Only one study(McClure et al.) was considered to be free from other sources ofbias such as baseline imbalance between groups.
3.3. Implementation methods and their effects
Table 2 describes each individual study with the corresponding re-sults presented in Table 3. Implementation methods included trainingof healthcare professionals (n=6), changes to primary care/generalpractice management (n=3), peer or lay volunteer-delivered pro-grams (n=3), and community awareness programs (n=3). Thelevel of description of the implementation strategies was mixed,with some studies providing only brief details.
3.3.1. Training of health care professionalsSix studies (Baraff et al., 1999; Brown et al., 2005; Fortinsky et al.,
2008; Larsson et al., 2010; Milisen et al., 2009; Scott et al., 2006) uti-lized training and dissemination of evidence to healthcare profes-sionals. For those that reported the duration of training, it variedfrom 30 minutes to one day, targeting a range of staff including
Number of records identified through database searching
n=3701
Number of records screened after duplicates removed
n=3638
Number of full-text articles searched for eligibility
n=108
Number of papers selected for inclusion
n=15
Fig. 2. Flow diagram fo
doctors, nurses, physical and occupational therapists, and healthcaresupport workers.
Two of the studies (Brown et al., 2005; Fortinsky et al., 2008)reported on a comprehensive approach to implementation as part ofthe Connecticut Collaboration for Falls Prevention (CCFP). This pro-gram incorporated training and dissemination of evidence-basedfalls prevention interventions using behavior change strategies, opin-ion leaders, media awareness campaigns, outreach visits to older peo-ple, and patient and provider materials. This collaborative approachresulted in improvements in fall-prevention assessment and manage-ment among physical therapists, community-based rehabilitationtherapists, and nurses. Thirty-eight percent of physical therapistsreported almost always using falls prevention strategies six weeksfollowing training, compared with 14% before training; 68% increasedtheir use of falls prevention strategies in practice with 7% decreasinguse in practice. A year after training, more than 70% of communitybased practitioners reported undertaking assessments of balance,mobility and postural hypotension, with around half assessing homehazards and poly-pharmacy. Around half of home health agencies(HHA) had 100% of their clinical staff following the recommendedfalls assessment and management strategies for mobility, posturalhypotension, polypharmacy, home hazards, and balancemanagement.
Three (Baraff et al., 1999; Milisen et al., 2009; Scott et al., 2006) ofthe remaining four studies using training explicitly evaluated theirimplementation methods in terms of changing clinical practice be-haviors. Baraff et al. (1999) trained medical and nursing staff in
Number of records excluded n=3530
Number of full-text articles excluded n=93
Reason for exclusion: -Full-text not available in English (n=1) -Abstract only (n=3) -Opinion paper (n=6) -Did not address falls prevention in community-dwelling older people (n=7) -Did not evaluate implementation (n=76)
r study selection.
Table 1Quality Assessment of Included Studies using the Cochrane Risk of Bias Tool.
Study Sequencegeneration
Allocationgeneration
Blinding Incompleteoutcome data
Selectiveoutcome reporting
Other sourcesof bias
Baraff et al. (1999) No No No Yes Yes NoBarnett et al. (2004) No No No No Unclear UnclearBrown et al. (2005) No No No Unclear Yes NoDeery et al. (2000) No No Unclear No No NoFortinsky et al. (2008) No No Yes No Yes UnclearGardner et al. (2002) No No Yes Unclear Unclear UnclearHealy et al. (2005) No No No No Unclear NoLarsson et al. (2010) No No Yes Unclear No NoMcClure et al. (2010) No No Unclear Yes Unclear YesMilisen et al. (2009) No No No No Unclear NoScott et al. (2006) No No Unclear Unclear Unclear UnclearShah et al. (1997) No No Unclear No Unclear NoStackpool (2006) No No Yes Yes Yes UnclearSze et al. (2005) No No Unclear Unclear Unclear NoWenger et al. (2009) No No Unclear Unclear Unclear No
Yes=adequately addressedNo=inadequately addressed
446 V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
emergency departments (ED) in a locally developed guideline andreported improvements in documentation for some aspects of historytaking, assessment, and actions. When examining issues around theimplementation of a falls prevention guideline with community-based healthcare staff, Milisen et al. (2009) reported that 88% of prac-titioners considered falls prevention important. However, there wassome disagreement between professions regarding responsibility forthe assessment and management of fall risk factors and how best toimplement the guideline in practice. Only half of nurses thought itwould be feasible to implement guidelines into practice comparedwith between 71% and 89% of GPs, physiotherapists, and occupationaltherapists. Barriers to implementation were identified as time invest-ment without financial compensation, poor patient and family motiva-tion, and a lack of communication/collaboration between professionals.Scott et al. (2006) reported a 25% increase in the fall-related knowledgeamong healthcare support workers following training delivered bynurses and therapists, although it is unclear as to the nature of thisknowledge. Using the Falls Prevention Checklist and Action Plan© theuptake of recommendations by clients was low to moderate, for exam-ple, only 30% who had difficulties balancing whilst in the shower tookaction to reduce the risk.
3.3.2. Changes to primary care managementA total of three studies (Gardner et al., 2002; Shah et al., 1997;
Wenger et al., 2009) evaluated changes to the management of fallswithin primary care organizations as a result of the implementationof a falls prevention strategy. Two of these studies (Shah et al.,1997; Wenger et al., 2009) did so as part of a transformation of theway in which common problems experienced by older people wereassessed and managed. The conditions included urinary incontinence,depression, cognitive impairment, and functional limitations. Wengeret al. (2009) reported improvements in achieving quality indicatorsfor falls, including history-taking, physical assessments, and interven-tions. Adherence to specialist recommendations by primary care phy-sicians and patients was examined in one study by Shah et al. (1997),although only 11% (15/139) of individuals required recommenda-tions for falls. This study reported that recommendations were imple-mented by general practitioners in six out of nine cases. Among theseven patients receiving self-care recommendations, three adhered.The study by Gardner et al. (2002) evaluated the implementation ofprimary care practice nurse training to deliver exercise interventions,in terms of identifying older people for the exercise program and up-take. A perception of an inability to take part in an exercise programwas indicated by both general practitioners and older people. Reasons
for participation included perceived potential benefits in terms ofhealth and well being.
3.3.3. Peer or lay-volunteer training to implement programsThree studies (Deery et al., 2000; Healy et al., 2005; Sze et al.,
2005) delivered training to peers (n=1) or lay-volunteers (n=2)in order to deliver health promotion messages, relating to falls preven-tion, to older people. Deery et al. (2000) used peers to deliver educa-tional sessions to groups of older people, although it is unclear as tothe duration or content of their training. The training of lay-volunteersto advise and promote fall-related behavior change among older peoplewas undertaken in two studies (Healy et al., 2005; Sze et al., 2005) withtraining lasting from90 minutes to two days. These three studies exam-ined changes in fall-related knowledge, attitudes, and behaviors and, inthe main, these outcomes improved in the short and longer term, withthe exception of Deery et al. (2000) where control group participantshad greater falls prevention knowledge at three months, although at12 months the reverse was observed.
3.3.4. Community awareness programsThree studies undertaken in Australia used community programs
to raise awareness about falls and promote falls prevention activitiesamong the population, although each of these were evaluated differ-ently (Barnett et al., 2004; McClure et al., 2010; Stackpool, 2006).Barnett et al. (2004) assessed recall and current falls prevention prac-tices of healthcare staff and councils following the four year ‘Stay onyour Feet’ program. Five years after the commencement of the pro-gram, the 321 healthcare staff (GPs, pharmacists, community nurses,occupational therapists, physiotherapists and health promotion staff)took part in a survey. From this, 50% (70/139) of GPs and 30% (16/53)of pharmacists thought the program influenced their practice. Amongthe 129 community staff completing the survey, 48% had been involvedin the program, although many activities had been discontinued (suchas medication checks and exercise classes). Reasons included timelimited resources and a lower priority. Sustainability of activities wasreported to have beenhelped by the adoption of activities as part of nor-mal work, resources, and compatibility with other projects. A follow onfrom this study, by McClure et al. (2010), was undertaken to examinewhether less resource intensive methods would be effective. Althoughthey reported an increased awareness of falls and associated behaviorchange among the older population, no improvements were found interms of fall-related injuries and hospitalization.
The health promotion program utilized by Stackpool (2006) usingcommunity collaboration to promote physical activity among older
447V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
people found a 19% increase in the number of available physical activityclasses for older people and a 16% increase in attendance by older peo-ple over three years.
4. Discussion
There is some evidence to show that the implementation of fallsprevention programs into practice can be successful. Although weidentified a total of 15 studies, heterogeneity in terms of study design,implementation methods and outcomes has limited the extent towhich the identified data could be synthesized. The level of descrip-tion of the implementation strategies included in this review wasoften limited. For example, the papers that report an aspect of theCCFP program had clearly described implementation methods, butthe study by Deery and colleagues failed to describe how the peer-delivered model was developed, and omitted details such as howpeers were identified and trained, and the content of the training.This is in agreement with a review of complex interventions in work-place settings performed by Egan, Bambra, Petticrew, and Whitehead(2009), who found that implementation was frequently referred tobut was poorly described. A clear description of an intervention, albeita treatment or implementation method, is essential for study replica-tion, whether to inform further research or to utilize the findings inclinical practice. Context is also an important factor to describe as dif-ferent healthcare systems and cultural considerations may impact onwhether translating evidence is applicable or feasible.
Successful programs generally included some aspect of training ofhealthcare professionals in order to change clinical practice behaviorsthat have been reported to be a key aspect of implementation (Bero etal., 1998; Tinetti et al., 2006). Peer or lay delivered programs specifical-ly aimed at changing knowledge, attitudes, and fall-related behaviors ofolder people demonstrated some improvements, often related to avoid-ing or removing environmental hazards and extrinsic fall-risk factors.However, none of the non-professionally delivered programs includedtraining in exercise provision, a key element of effective falls preventionstrategies (Gillespie et al., 2009; Sherrington et al., 2008). There is cur-rently a trial underway in the UK comparing the effectiveness of usualcare with a peer-delivered home exercise program, and with a groupexercise intervention delivered by a qualified exercise instructor (Iliffeet al., 2010).
Evidence on changing clinical practice within primary care wasmixed. This may be due to competing priorities with other conditions.Community awareness programs appeared diverse in terms of out-comes and provided no clear picture in terms of the effectiveness ofthis method of implementation. Furthermore, one of the studies(Shah et al., 1997) evaluating impact in this area was publishedprior to 2000 when the evidence for falls prevention interventionswas lesswell established. Falls therefore carried a relatively low prioritywithin healthcare.
There is no general consensus with regards to which outcomesshould be used to examine the impact of implementation, possiblydue to differing interpretations as to what implementation is. Withinthe RE-AIM framework, Glasgow, Vogt, and Boles (1999) suggest theevaluation of implementation programs refers to the fidelity and ad-herence to a program, whereas, Rabin, Glasgow, Kerner, Klump, andBrownson (2010) suggest that evaluation requires a variety of out-comes that should be examined, from those at an individual level(e.g., behavior change of patients or professionals), to organizationallevel data, (e.g., healthcare costs). Policymakers and service commis-sioners are interested in improved outcomes, such as fall-related inju-ries or hospital admissions, which require effective falls preventioninterventions and effective implementation (FPG Child DevelopmentInstitute, 2011). The CCFP program was based upon an effectivemulti-factorial intervention (Tinetti et al., 1994) that has also beenshown to result in a 9% (95% confidence interval [CI] 6 to 12%)
reduction in serious fall-related injuries and an 11% (95% CI 8 to14%) reduction in medical service use (Tinetti et al., 2008).
To our knowledge, this is the first systematic review that has evalu-ated implementation strategies in relation to falls prevention amongolder people. We conducted an extensive literature search in a rangeof electronic databases and included a range of study designs as werecognize that traditional randomized controlled trials are less feasibleand may not be appropriate when evaluating implementation into clini-cal practice (Medical Research Council, 2000; Rabin et al., 2010).
There are a number of limitations of this review. Firstly, althoughwe were able to identify a reasonable number of relevant papers, po-tential risk of bias was generally high or unclear (Higgins & Green,2009). This was linked to the fact that most study methods did not in-corporate a control element and some studies used surveys. Althoughevidence suggests that the failure to report key quality indicators mayindicate bias, the extent of the size and direction of the impact of thisbias is not always clear. The quality assessment of studies designed toevaluate the implementation of evidence into practice has not beenwell researched and there are no guidelines to assist in the reportingof this type of evaluation. The Cochrane risk of bias tool may not bethe most appropriate tool for evaluating quality in studies of thistype and there may be additional issues such as social desirabilitybias that have not been addressed either in the publications or inthe assessment of their quality. Appraising evaluations of implemen-tation is a relatively new area and further work is required to developappropriate methods (Egan et al., 2009). Secondly, we included onlypapers that were available in English, although based on informationprovided in the abstracts it is unlikely that the non-English languagepapers identified in the search would have met the other selectioncriteria. Thirdly, we did not include grey literature, defined as litera-ture not published in journals, such as conference abstracts andunpublished theses (Higgins & Green, 2009), which may havehighlighted further studies and reports, and finally, we were unableto undertake meta-analyses due to heterogeneity in all aspects ofthe included studies.
A small number of studies in this review employed mixedmethods. Implementation research is particularly ripe for such anapproach in which evidence of qualitative change can be set along-side elucidation of the reasons for such change. The fact that the ma-jority of papers in our reviewwere restricted to quantitative enquirymeans that the influence on implementation efforts at individual(clinician or patient) and organizational levels is constrained.
In summary, there is evidence to support active training andsupport of healthcare professionals in order to implement falls preven-tion evidence into clinical practice. The evidence around changing theway people who fall are managed within primary care practices ismixed, as is the use of community awareness programs and peer orlay-delivered falls prevention programs. Nevertheless, questionsremain about themethods used to report, evaluate, and appraise imple-mentation research, such as developing effective search strategiesand quality appraisal methods. The relative importance of this fieldneeds to be promoted alongside evidence for effective healthcare inter-ventions in terms of funding if evidence is to be translated into policyand clinical practice.
5. Impact on industry
The implementation of falls prevention research into practice in-volves changing the attitudes and behaviors of older people, healthcareprofessionals, and organizations. However, there is a need for furtherevaluation on how this can be best achieved.
Acknowledgement
This work was funded by the National Institute for Health Re-search (NIHR). This report/article presents independent research
Table 2Description of Study Characteristics, Stratified by Type of Implementation Method.
Study Country Setting Study design Study purpose Implementation strategy Falls preventionintervention
Sample Population Outcomesevaluatingimplementation
Follow up
Training of healthcare professionalsBaraff et al.(1999)
USA ED Repeatedmeasures,cohort study
To assess impact ofpractice guideline onprocess of care
Training of physicians (2 hours)and nurses (30 minutes)
Medicationmanagement,vaccinationsandophthalmologyreferral
Not reported 3843 older people(>65 years) attendingED
Documentaryevidence of historytaking, physicalexamination andaction taken
1 year
Brown et al.(2005)
USA Physicaltherapypractices
Survey To describe physicaltherapists knowledge,attitudes and behavioursrelating to fall prevention
CCFP programme comprisingtraining of physical therapists(1 hour)
Multi-factorial 94 physical therapy providersfrom 119 organisations
- Self-reported use offalls preventionstrategies andchange in practice.
6 weeks
Fortinsky etal. (2008)
USA Community Survey To describe extent ofimplementation of EBP bynurses and therapists
CCFP programme comprisingtraining (90 minutes) of homehealth care staff
Multi-factorial 184 nurses and rehabilitationtherapists from 19 homehealth agencies (HHA)
- Self-reported fallspreventionassessment andmanagementpractice
1 year
Larsson et al.(2010)
Sweden Community Repeatedmeasures,cross-sectionalstudy
To evaluate the impact ofthe programme on injuryrates
Training of communitypractitioners (half day);Media campaign 2006/7
Unclear 32 community practitioners;82 members of public
21,898 people aged>55 years
Awareness ofcampaign; use of ahazard reportingtelephone line
1 year
Milisen et al.(2009)
Belgium Community Survey To test feasibility ofimplementing a fallsprevention guideline
Staff training (2 hours) Multi-factorial 23 GPs, 34 nurses, 25 PTs,17 OTs
- Importance,feasibility andpracticality ofguideline.
Unclear
Scott et al.(2006)
Canada Community Repeatedmeasures,cohort study
To evaluate the impact oftraining on knowledge,practice, falls and relatedinjuries
Training of communityhealthcare support workers(1 day)
Multi-factorial 57 community healthcaresupport workers
87 people requiringhome help support
Change inknowledge, uptakeofrecommendations
Sixmonths
Changes to primary care practicesGardner et al.(2002)
NewZealand
Primarycare
Process andimpactevaluation ofa non-randomisedtrial
Applicability andfeasibility of a primary carenurse-delivered exerciseprogramme
Nurse training (1 week) Strength andbalance training
61 general practitioners in36 practices;3 nurses
330 exerciseparticipants aged>80 years
Recruitment issues;fidelity andadherence
1 year
Shah et al.(1997)
USA Primarycarepractices
Case series To examineimplementation of CGArecommendations
Communication betweengeriatrician, primary carephysician and patient
Multi-factorial(individuallytailored)
Not reported 150 people >65 yearswith urinaryincontinence, falls,depression orfunctional impairment
Physicianimplementationand patientadherence rates
3 months
448V.G
oodwin
etal./
JournalofSafetyResearch
42(2011)
443–451
Table 2 (continued)
Study Country Setting Study design Study purpose Implementation strategy Falls preventionintervention
Sample Population Outcomesevaluatingimplementation
Follow up
Wenger et al.(2009)
USA Primarycarepractices
Non-randomisedtrial
To examine effect ofACOVE-2 intervention onprocess of care
Changes to practice processesand training of primary carephysicians (3 hours)
Unclear 2 practices;40 physicians 644 people aged>70 yearsexperiencing falls,urinary incontinenceor cognitiveimpairment
% of qualityindicators satisfied
13 months
Peer or lay volunteer delivered programmesDeery et al.(2000)
Australia Community Matchedcohort withrepeatedmeasures
To assess impact of peereducation on fall-relatedknowledge, attitudes andbehaviours
Peer-presented education sessions.Training of peers unclear.
Education Not reported 361 peopleaged>60 years(education) and 174age and sex matchedcontrols
Fall-relatedattitudes,knowledge andbehaviours
3 and12 months
Healy et al.(2005)
USA Community Repeatedmeasures,cohort study
To examine whether a CBTprogramme ‘a Matter ofBalance’ can be effectivelydelivered by volunteers
Training of lay volunteers (2 days) Risk behaviourchange
Not reported 349 older adults(51–95 years)
Fidelity to theprogramme;changes in fall-related self-efficacyand behaviours.
6 weeks,6 months,1 year
Sze et al.(2005)
HongKong
Community Survey To evaluate impact of aneducation and trainingprogramme on awarenessand knowledge of fallprevention
Training programme-communitycentre staff and lay volunteers(90 minutes); Educational seminarfor older people;
Education andhome hazardmodification
34 staff and 312volunteers
5114 older people Knowledge andawarenessregarding fallsprevention
Unclear
Community awareness programmesBarnett et al.(2004)
Australia Community Surveys To assess sustainability of acommunity SOYF fallsprevention programme
Awareness raising, communityeducation, policy development,engaging health professionals (1992to 1996)
Multi-factorial 321 healthcareprofessionals); 9 shirecouncils and 8 shire accesscommittees
80,000 people aged>60 years
Recall of SOYF,involvement andcurrent fallspreventionactivities
5 years
McClure et al.(2010)
Australia Community Repeatedmeasures,cross-sectionalstudy
To evaluate whether apopulation basedprogramme reduces fallsand injuries
(a) Peer health promotion of fallsprevention activities, or (b) healthpromotion officers delivering andsupporting physical activity. 2002 to2006.
Multi-factorial 1,600 older people (a) 43,821, (b) 58,722 Fall-relatedbehaviour change
4 years
Stackpool(2006)
Australia Community Repeatedmeasures,cross-sectionalstudy
To establish viability ofcollaborative model topromote physical activityamong older people
Collaborative management model(2000 to 2003)
Physical activity 6 area Health Promotion uni Not reported Availability anduptake of physicalactivityprogrammes
3 years
ED Emergency Department; CFFP Connecticut Collaborative Falls Prevention; CGA Comprehensive Geriatric Assessment; ACOVE-2 Assessing Care of Vulnerable Elders; CBT C nitive behavioural therapy; SOYF Stay of your Feet
449V.G
oodwin
etal./
JournalofSafetyResearch
42(2011)
443–451
ts
og
Table 3Individual Study Results, Stratified by Implementation Method.
Study Results
Training of healthcare professionalsBaraff et al.(1999)
Improvement in 6 out 10 items on history taking; 2 out of 4 items on physical examination; 2 out of 6 items on actions taken
Brown et al.(2005)
Most physical therapists reported an increased use of falls prevention strategies in practice.
Fortinsky et al.(2008)
Most community healthcare staff used recommended guidance for assessment and management of falls in practice
Larsson et al.(2010)
Low awareness of campaign in the community (20%); 72% of fall prevention agents aware. 29 reports of community hazards in 6 months
Milisen et al.(2009)
Disagreement between different professionals as to feasibility and roles in using falls prevention guideline.
Scott et al.(2006)
Increased knowledge of staff, high use of checklist and action plan by staff, mixed uptake of recommendations by clients.
Changes to primary care practicesGardner et al.(2002)
Reasons for exclusion: being medically unwell; physical frailty; considered incapable of exercise. Reasons for participation: doctor recommendation, health/functional benefits, prevent falls. Reasons for declining: already active, too frail/unwell, commitment too long; not interested.
Shah et al.(1997)
6/9 physician recommendations implemented and all adhered to by patients. 3/7 self-care recommendations adhered to
Wenger et al.(2009)
44% of intervention group and 23% controls met quality indicator for falls.
Peer or lay volunteer programmesDeery et al.(2000)
Greater changes in attitude reported for intervention group; Intervention group has lower knowledge at 3 months but greater at 12 months compared withcontrols; intervention group made more environmental changes and changed behaviour at 3 and 12 months.
Healy et al.(2005)
Significant improvements in self-efficacy and fall management.
Sze et al.(2005)
Older people and volunteers reported gaining knowledge about falls prevention. Almost all community centre staff had set up falls prevention activities.
Community awareness programmesBarnett et al.(2004)
Culprit medication checked by more than half of GPs/Pharmacists most of time.Around half of community staff ran exercise classes. No councils had acomprehensive falls prevention policy. No access committees maintained falls prevention activities.
McClure et al.(2010)
Increased awareness of falls. Behaviour change of older people in relation to falls prevention.
Stackpool(2006)
Increase in availability of exercise classes and enrolment.
450 V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
commissioned by the NIHR. The views expressed in this publicationare those of the author(s) and not necessarily those of the NHS, theNIHR or the Department of Health.
References
American Geriatrics Society and the British Geriatrics Society (2010). Clinical practiceguideline: Prevention of falls in older persons. New York: Author.
Baraff, L. J., Lee, T. J., Kader, S., & Penna, R. D. (1999). Effect of a practice guideline on theprocess of emergency department care of falls in elder patients. Academic EmergencyMedicine, 6, 1224–1231.
Barnett, L. M., Van Beurden, E., Eakin, E. G., Beard, J., Dietrich, U., & Newman, B. (2004).Program sustainability of a community-based intervention to reduce falls amongolder Australians. Health Promotion International, 19, 281–288.
Bero, L. A., Grilli, R., Grimshaw, J. M., Harvey, E., Oxman, A. D., & Thomson, M. A. (1998).Closing the gap between research and practice: an overview of systematic reviewsof interventions to promote the implementation of research findings. BMJ, 317(7156), 465–468.
Brown, C. J., Gottschalk, M., Van Ness, P. H., Fortinsky, R. H., & Tinetti, M. E. (2005).Changes in physical therapy providers' use of fall prevention strategies followinga multicomponent behavioural change intervention. Physical Therapy, 85, 394–403.
Deery, H. A., Day, L. M., & Fildes, B. N. (2000). An impact evaluation of a falls preventionprogram among older people. Accident Analysis and Prevention, 32, 427–433.
Department of Health (2009). Fracture prevention services: an economic evaluation. London:Author.
Egan, M., Bambra, C., Petticrew, M., & Whitehead, M. (2009). Reviewing evidence oncomplex social interventions: appraising implementation in systematic reviewsof the health effects of organisational-level workplace interventions. Journal ofEpidemiology and Community Health, 63, 4–11.
Fortinsky, R. H., Baker, D., Gottschalk, M., King, M., Trella, P., & Tinetti, M. E. (2008). Extentof implementation of evidence-based fall prevention practices for older patients inhome health care. Journal of the American Geriatrics Society, 56, 737–743.
FPG Child Development Institute (2011). State implementation and scaling up ofevidence-based practices (SISEP). Chapel Hill, NC: Authorwww.scalingup.org
Ganz, D. A., Alkema, G. E., &Wu, S. (2008). It takes a village to prevent falls: reconceptualiz-ing fall prevention and management for older adults. Injury Prevention, 14, 266–271.
Gardner, M. M., Robertson, M. C., McGee, R., & Campbell, A. J. (2002). Application of afalls prevention programme for older people to primary health care practice. Pre-ventative Medicine, 34, 546–553.
Gillespie, L. D., Robertson, M. C., Gillespie, W. J., Lamb, S. E., Gates, S., Cumming, R. G., &Rowe, B. H. (2009). Interventions for preventing falls in older people living in thecommunity (Review). Cochrane Database of Systematic Reviews, 2.
Glasgow, R. E., Vogt, T. M., & Boles, S. M. (1999). Evaluating public health impact ofhealth promotion interventions: the RE-AIM framework. American Journal of PublicHealth, 89, 1322–1327.
Goodwin, V., Martin, F. C., Husk, J., Lowe, D., Grant, R., & Potter, J. (2010). The nationalclinical audit of falls and bone health - secondary prevention of falls and fractures:a physiotherapy perspective. Physiotherapy, 96(1), 38–43.
Greenhalgh, T., Robert, G., Macfarlane, F., Bate, P., & Kyriakidou, O. (2004). Diffusion ofinnovations in service organisations: systematic review and recommendations.Millbank Q, 82, 581–629.
Healy, T. C., Haynes, M. S., McMahon, E. M., Botler, J. L., & Gross, L. (2005). The feasibilityand effectiveness of translating a Matter of Balance into a volunteer lay leadermodel. Journal of Applied Gerontology, 27, 34–51.
Higgins, J. P. T., & Green, S. (2009). Cochrane handbook for systematic reviews of inter-ventions Version 5.0.2. Chichester, UK: John Wiley and Sons Ltd.
Iliffe, S., Kendrick, D., Morris, R., Skelton, D., Gage, H., Dinan, S., Stevens, Z., Pearl, M., &Masud, T. (2010). Multi-centre cluster randomized trial comparing a communitygroup exercise programme with home based exercise with usual care for peopleaged 65 and over in primary care: protocol of the ProAct 65+ trial. Trials, 11(1), 6.
Larsson, T. J., Hägvide, M. -L., Svanborg, M., & Borell, L. (2010). Falls prevention throughcommunity intervention: a Swedish example. Safety Science, 48, 204–208.
McClure, R. J., Hughes, K., Ren, C., McKenzie, K., Dietrich, U., Vardon, P., Davis, E., & New-man, B. (2010). The population approach to falls injury prevention in older people:findings of a two community trial. BMC Public Health, 10(79).
Medical Research Council (2000). A framework for development and evaluation of RCT'sfor complex interventions to improve health. London: Medical Research Council.
Milisen, K., Geeraerts, A., & Dejaeger, E. (2009). Use of a fall prevention practice guide-line for community-dwelling older persons at risk for falling: a feasibility study.Gerontology, 55, 169–179.
National Institute for Health Clinical Excellence (2004). Falls: The assessment and pre-vention of falls in older people. CG21. London: Royal College of Nursing.
Rabin, B. A., Glasgow, R. E., Kerner, J. F., Klump,M. P., & Brownson, R. C. (2010). Dissemina-tion and implementation research on community-based cancer prevention. AmericanJournal of Preventative Medicine, 38, 443–456.
451V. Goodwin et al. / Journal of Safety Research 42 (2011) 443–451
Roen, K., Arai, L., Roberts, H., & Popay, J. (2006). Extending systematic reviews to includeevidence on implementation: methodological work on a review of community-based initiatives to prevent injuries. Social Science & Medicine, 63, 1060–1071.
Rose, D. J., Alkema, G. E., Choi, I. H., Nishita, C. M., & Pynoos, J. (2007). Building an infra-structure to prevent falls in older Californians. Annals of the New York Academy ofSciences, 1114, 170–179.
Royal College of Physicians (2007). National clinical audit of falls and bone health in olderpeople. London: Author.
Sambrook, P., & Cooper, C. (2006). Osteoporosis. The Lancet, 367(9527), 2010–2018.Scott, V. J., Votova, K., & Gallagher, E. (2006). Falls prevention training for community
health workers: strategies and actions for independent living (SAIL). Journal ofGerontological Nursing, 32, 48–56.
Shah, P. N., Maly, R. C., Frank, J. C., Hirsch, S. H., & Reuben, D. B. (1997). Managing geriatricsyndromes: what geriatric assessment teams recommend, what primary care physi-cians implement, what patients adhere to. Journal of the American Geriatrics Society, 45,413–419.
Sherrington, C., Whitney, J. C., Lord, S. R., Herbert, R. D., Cumming, R. G., & Close, J. C. T.(2008). Effective exercise for the prevention of falls: a systematic review andmeta-analysis. Journal of the American Geriatrics Society, 56, 2234–2243.
Stackpool, G. (2006). 'Make a move falls prevention project: an area health service col-laboration. Health Promotion Journal of Australia, 17, 12–20.
Sze, P. C., Lam, P. S., Chan, J., & Leung, K. S. (2005). A primary falls prevention programmefor older people in Hong Kong. British Journal of Community Nursing, 10, 166–171.
Tinetti, M. E., Baker, D., King, M., Gottschalk, M., Murphy, T. E., Acampora, D., Carlin, B. P.,Leo-Summers, L., & Allore, H. G. (2008). Effect of dissemination of evidence in reduc-ing injuries from falls. The New England Journal of Medicine, 359, 252–261.
Tinetti, M. E., Baker, D., McAvay, G., Claus, E. B., Garrett, P., Gottschalk, M., Koch, M. L.,Trainor, K., & Horwitz, R. I. (1994). A multifactorial intervention to reduce therisk of falling among elderly people living in the community. The New England Jour-nal of Medicine, 331(13), 821–827.
Tinetti, M. E., Gordon, C., Sogolow, E., Lapin, P., & Bradley, E. H. (2006). Fall-risk evaluationand management: challenges to adopting geriatric care practices. The Gerontologist,46, 717–725.
Wenger, N. S., Roth, C. P., Shekelle, P. G., Young, R. T., Solomon, D. H., Kamberg, C. J.,Chang, J. T., Louie, R., Higashi, T., Maclean, C. H., Adams, J., Min, L. C., Ransohoff,K., Hoffing, M., & Reuben, D. B. (2009). Practice-based intervention to improve pri-mary care for falls, urinary incontinence and depression. Journal of the AmericanGeriatrics Society, 57, 547–555.
VlaE
ictoria Goodwin, PhD, is a Senior Research Fellow for PenCLAHRC (Peninsula Col-boration for Leadership in Applied Health Research and Care) at the University ofxeter, UK and a physiotherapist for Torbay Care Trust. She has recently completed a doc-
torate evaluating an exercise intervention to reduce falls among people with Parkinson'sdisease. She is involved with the British Geriatrics Society specialist section for Falls andBone Health and is former national chair of AGILE (Chartered Physiotherapists workingwith Older People). Her research interests are the rehabilitation of older people and thosewith long term conditions.
Tracey Jones-Hughes, PhD, is an Associate Research Fellow for PenTAG (Peninsula Tech-nology Assessment Group), currently working on Health Technology Assessment. She hasa diverse background, ranging from nursing to earning a PhD in environmental chemistryat Plymouth University. However, more recently she became involved in project facilita-tion for PenCLAHRC, focusing on translation of research into clinical practice. Linking withthe varied nature of her career, Tracey's current research interests include systematic re-views of environment and human health related issues.
Jo Thompson-Coon, PhD, is a Research Fellow for PenCLAHRC as part of the evidencesynthesis team. Her background is in pharmacology and she has worked in the respira-tory and complementary medicine fields. Her current role involves identifying andprioritising potential local research projects and producing systematic reviews to in-form evidence-based practice.
Kate Boddy, MSc, is an Information Specialist at PenCLAHRC where she has been work-ing since 2009. She has been working in health services research since 2004 and re-ceived her MSc in Library and Information Management from the University of theWest of England in 2009. She has worked on numerous systematic reviews providinginformation support and has a particular research interest in the ways in which differ-ent search interfaces can affect search results.
Ken Stein, MD, is Professor of Public Health with a background as a physician in generalpractice. He directs a multi-disciplinary research group which undertakes evidence syn-theses and economic evaluation on a wide range of health technologies and is deputy di-rector of the PenCLAHRCwhich aims to improve the influence of research on NHS practicein the UK.
