Using PMOS to Evaluate Safe Care for Adults with Complex Health Needs

Why complex health needs?Adults with complex health needs are those who have more than one illness/disease/condition/disability, or those who need care in more than one setting. Due to our ageing population, improved treatment of long term conditions, and prevalence of lifestyle risk factors like smoking, obesity, alcohol and malnutrition, complex health needs are becoming the norm, and are a growing concern within the NHS.

What were our aims?• To explore the type and frequency of care

experiences reported by patients, carers, and health/social care staff using PMOS.

• To establish if PMOS accurately reflects these care experiences

What did we do?

What is PMOS?• The Patient Measure Of Safety was developed by Sally Giles, Rebecca

Lawton, et al. and was published in BMJ Quality and Safety in 20131.• It was based on the categories defined by the Yorkshire Contributory

Factors Framework2 and contains 12 domains (see Fig. 2).• The patient questionnaire enables patients to identify factors that

contribute to the safety of their care. • It was designed in a hospital setting and has been further tested in

primary care • It has not yet been established whether the tool could represent the

safety concerns of patients with complex health needs.

James Lind Alliance (JLA) Imperial PSTRC ran a Priority Setting Partnership (PSP) to identify the top 10 research questions surrounding safe care for adults with complex health needs, according to patients, carers and health/social care staff. For more information, use the QR code.This process began with a survey distributed to these stakeholders that gathered a rich qualitative data set.

• 70.6% female• 45-64 most

common age • 84.5% English• 83.5% Caucasian

Why is this important?A modified version of the PMOS can accurately reflect the safety concerns of people living with complex needs and those caring for them. These findings suggest areas where further research might be directed to improve patient safety for adults with complex health needs.

Sarah Godlee, Anna Lawrence Jones, Bryony Dean Franklin & Stephanie ArcherNIHR Imperial Patient Safety Translational Research Centre (Imperial PSTRC), Imperial College London

Question 1:Based on your experiences of receiving,

giving or observing healthcare, tell us about any times when something went wrong or

nearly went wrong, or when you felt the care was unsafe?

“”

17 responses excluded:Non-UK (1)Lack of information (16)

441 responses to question 1 transcribed into excel

Data cleansed and anonymised

424 responses imported into NVivo 12

19 responses excluded:No safety issues identified

17 responses excluded:Blank

10% responses double coded with

near perfect agreement(κ= 0.81)

441 responses

Aug 2017-Feb 2018

Via steering group, twitter,

INVOLVE

388 responses analysed and coded:

Using PMOS domains

https://bit.ly/2mPRpFv

Using ‘unclassified’

Responses coded ‘unclassified’

analysed further to identify new themes

Healthcare professionals

34%

Familial carers 18%

Public10%Adults with

complex health needs

25%

Other 5%

Prefer not to say

1%

Blank3%

Employed carer 4%

Contact:Sarah Godleesrbg2@cam.ac.uk

References1. Giles, S. J., Lawton, R. J., Din, I., & McEachan, R. R. C. (2013). Developing a patient measure of safety (PMOS). BMJ Quality and Safety, 22(7), 554–562. 2. Lawton, R., McEachan, R. R. C., Giles, S. J., Sirriyeh, R., Watt, I. S., & Wright, J. (2012). Development of an evidence-based framework of factors contributing to patient safety

incidents in hospital settings: A systematic review. BMJ Quality and Safety, 21(5), 369–380. 3. McEachan, R. R. C., Lawton, R. J., O’Hara, J. K., Armitage, G., Giles, S., Parveen, S., … Wright, J. (2014). Developing a reliable and valid patient measure of safety in hospitals

(PMOS): A validation study. BMJ Quality and Safety, 23(7), 565–573.

15.2%

Holistic Care

7.2%12.6%

Transition of Care

Diagnostic Process

Medication Issues

17.8%

Clinical Negligence

5.2%

Sample

Additional categories were needed to fully represent the specific experiences surrounding complex health needs.Fig. 4 The additional categories identified from the data

and the percentage of responses they were identified in

• All 12 original PMOS domains were represented by the data.

• Communication was the most frequently identified PMOS domain, n=172.

• Communication was also the most frequently identified PMOS domain within every cohort group.

Additional themes Outcomes

What we found

08/10/2019 Word Art

1/1

Fig. 3 The most frequently used words obtained by natural language processing

Fig. 2 A bar chart to show the percentage of responses identifying each PMOS domain

Fig. 1 A pie chart to show distribution of cohort groups