Final Report for Developing a Model for Deprescribing in Hospital

Sion Scott1, 2, Michael James Twigg1, Allan Clark3, Carol Farrow2, Janette Guymer4, Helen May5, Noreen Neal4, Martyn Patel5, David John Wright1, Debi Bhattacharya1

1. School of Pharmacy, University of East Anglia, Norwich, NR4 7TJ2. Pharmacy Department, Norfolk and Norwich University Hospitals NHS Foundation Trust

Colney Lane, Norwich, NR4 7UY3. Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ4. Lay advisors, Norwich, UK5. Older People’s Medicine Department, Norfolk and Norwich University Hospitals NHS

Foundation Trust, Colney Lane, Norwich, NR4 7UY

This final report presents research funded in its entirety by Pharmacy Research UK (PRUK reference number: PRUK-2017-PA2-A). The views expressed in this report are those of the author/s and not necessarily that of Pharmacy Research UK.

The University of East Anglia acknowledges the support of the National Institute of Health Research Clinical Research Network (NIHR CRN).

ContentsBackground..............................................................................................................................1

Study 1: Deprescribing admission medication at a UK teaching hospital; a report on quantity and nature of activity................................................................................................................3

Aim.......................................................................................................................................3

Method..................................................................................................................................3

Results..................................................................................................................................4

Discussion............................................................................................................................4

Study 2: Attitudinal predictors of older peoples’ and caregivers’ motivation to deprescribe in hospital.....................................................................................................................................6

Aim.......................................................................................................................................6

Methods................................................................................................................................6

Study sample and setting..................................................................................................6

Recruitment and survey administration.............................................................................6

Sample size......................................................................................................................7

Statistical analysis.............................................................................................................7

Results..................................................................................................................................7

Patients.............................................................................................................................9

Caregivers.........................................................................................................................9

Discussion..........................................................................................................................16

Study 3: A blueprint for deprescribing in hospital: A focus group study with geriatricians and pharmacists using the Theoretical Domains Framework (TDF).............................................17

Aims................................................................................................................................17

Methods..............................................................................................................................17

Sample............................................................................................................................17

Data collection................................................................................................................18

Analysis...........................................................................................................................18

Results................................................................................................................................18

Phase 1: Thematic analysis............................................................................................18

Phase 2: Mapping of all determinants of deprescribing to the TDF and prioritising mapped TDF domains for targeting in a deprescribing intervention...............................21

Phase 3. Identifying the BCTs linked to prioritised TDF domains...................................23

Discussion..........................................................................................................................24

Future recommendations.......................................................................................................25

References.............................................................................................................................26

Patient and public involvement..............................................................................................29

Personal development...........................................................................................................30

BackgroundPotentially inappropriate medicines (PIMs) are those which may be offering more risks than benefits and are associated with adverse outcomes including morbidity, hospitalisation and mortality(1). A prospective study across six European hospitals reported between 34.7% and 77.3% of inpatients ≥65 years were prescribed a PIM on admission(2). There is therefore a need for healthcare practitioners in hospital to review medicines to determine suitability for discontinuation(3,4). The term ‘deprescribing’ has been defined as the “systematic process of identifying and discontinuing drugs in instances in which existing or potential harms outweigh existing or potential benefits…”(5). Accordingly, deprescribing a medication may be in response to an adverse clinical trigger (reactive) or an attempt to reconcile risks of maintaining versus discontinuing (proactive)(6).

While the number of studies investigating clinically significant outcomes associated with deprescribing are limited, deprescribing appears safe and has been associated with positive effects on mortality and falls in certain circumstances(7). Central to ensuring that deprescribing is safe and effective is an accurate medication history and provision for adequate physiological monitoring to observe response to medication withdrawal(5). Given these requirements, an admission to hospital where a medication history is routinely undertaken and physiological parameters are routinely monitored, may provide an appropriate opportunity for a deprescribing intervention. However, deprescribing practice in hospital is poorly understood and there is a need identify the extent to which it currently occurs(8).

Unsurprisingly, patient engagement in decision-making is an essential component of deprescribing(9). It is an opportunity to determine whether deprescribing is appropriate, agree safety-netting measures such as symptoms patients should report to indicate that re-prescribing is necessary and establish motivation to adhere to the deprescribing(9). Trials across multiple settings report up to half of older patients decline deprescribing interventions(10–13). Exploration of predictors for this lack of motivation has focussed on external characteristics such as age, gender and number of medications. Informal caregivers such as family members are increasingly involved in medication decision-making. For patients that are unable to participate in these decisions, such as those living with cognitive impairment, caregivers frequently assume sole responsibility(14,15). Furthermore, caregivers influence engagement with deprescribing by physicians and patients who are able to participate in decision-making(8,16). Despite the wide ranging influence exerted by caregivers on the deprescribing processes, their level of engagement with and attitudinal factors influencing motivation to deprescribe are unknown.

A systematic review and thematic analysis of prescriber views towards deprescribing was undertaken to identify the barriers and enablers to implementation across all healthcare settings(8). Of the 21 included studies, 20 captured the views of primary care physicians and one was based in the hospital setting. Key themes identified were fear of negative consequences, the influence of patients and family and the need for appropriate skills and knowledge to safely deprescribe(8). In seven studies, participants expressed training in geriatric medicine was a facilitator to deprescribing. Geriatrician-led deprescribing in hospital may therefore overcome some of the barriers reported in the primary care setting, however the influencers of this professional group’s deprescribing practice are unknown. The single hospital-based study in the review focused on physicians perceptions of the causes of inappropriate prescribing, whilst the influences on deprescribing were not explored(17). Moreover, despite over half of existing deprescribing interventions in hospital involving pharmacists(18), no studies capturing their views were identified(8). Hospital pharmacists in the UK are responsible for reconciling patients’ medicines and optimising treatments to avoid

1

potential medication-related harms(19). The influencers of pharmacists’ deprescribing behaviour therefore require exploration given their likely pivotal role in hospital deprescribing.

Whilst a number of the barriers and enablers reported from the primary care setting may also apply in hospital(8), there are likely to be others such as those relating to the hospital environmental context, which remain unknown. Moreover, the extent to which these might vary between hospital organisations is also unclear. Large teaching hospitals may benefit from greater resources relative to smaller district general hospitals; however, the former may treat patients with more severe illness, potentially limiting capacity to adopt new models of care(20–22). Accordingly, the nature and relative importance of barriers and enablers to deprescribing may differ across these hospital contexts. There is therefore a need to understand the implementation problems and potential solutions to guide the development of a scalable deprescribing intervention for the hospital setting.

The aim of this programme of work was to identify potentially effective intervention components by drawing on evidence and theory, to support geriatricians and pharmacists to deprescribe for older people in hospital. The programme of work was underpinned by three empirical studies:

Study 1 Deprescribing admission medication at a UK teaching hospital; a report on quantity and nature of activity

Study 2 Attitudinal predictors of older peoples’ and caregivers’ motivation to deprescribe in hospital

Study 3 A blueprint for deprescribing in hospital: A focus group study with geriatricians and pharmacists using the Theoretical Domains Framework

2

Study 1: Deprescribing admission medication at a UK teaching hospital; a report on quantity and nature of activity

AimTo quantify and describe the nature of admission medication deprescribing practice in a large UK teaching hospital.

MethodThe study was confirmed as a service evaluation by the University of East Anglia Faculty of Medicine and Health Sciences Research Ethics Committee (Reference: 2016/2017 - 52 SE).

A retrospective analysis of all admission medications prescribed and discontinued at a large UK teaching hospital was undertaken over four weeks in February 2017. Data were extracted from the hospital’s electronic prescribing (e-prescribing) system for all wards and specialities except the Emergency Department and Intensive Care Unit as e-prescribing was not implemented in these areas. Prescriptions newly initiated during the admission and medication recorded as temporarily suspended were excluded because the study was designed to capture the extent to which admission medicines are deprescribed.

Patient sex and age, medication name and the e-prescribing reason for medication discontinuation were analysed.

A sample of 200 medication discontinuations were analysed by reviewing medical records to confirm or refute deprescribing activity and categorise the activity into proactive or reactive deprescribing. This sample size provides a 95% confidence interval of 3.0% around the estimate of the quantity of deprescribing. As there are no estimates of deprescribing prevalence in usual hospital care, the estimate is based on a UK deprescribing intervention trial reporting 8.5% of admission medicines deprescribed(23). Accepting this will be lower in the absence of an intervention, a maximum of 5.0% admission medicines likely to be deprescribed was estimated.

The majority of e-prescribing reasons are unambiguous such as “Acute kidney injury”. However, the reason “No longer clinically necessary” was deemed ambiguous by the local clinical team as in their experience this was often selected by prescribers when a suitable reason could not be identified. Medication discontinuations not assigned an e-prescribing reason were also considered ambiguous. Accordingly, sampling of 200 medication discontinuations was stratified, with a smaller number of discontinuations assigned unambiguous reasons (one-sixth of the total or 100% if three or less occurrences) sampled. Medication discontinuations assigned the ambiguous reason and where no reason was given were evenly sampled for the remaining reviews.

Informed by the existing literature(6), academics, senior hospital clinicians, patients and carers, the following definitions were developed and used to categorise deprescribing behaviour:

Reactive deprescribing: discontinuing a medicine in response to an adverse clinical trigger

Proactive deprescribing: discontinuing a medicine if future gains are unlikely to outweigh future harms

3

One clinical pharmacist extracted the prescriber’s rationale for medication discontinuation verbatim from medical records. Each discontinuation was independently categorised by a clinical pharmacist and consultant geriatrician into proactive, reactive or not deprescribing. Inter-rater reliability was assessed using Cohen’s Kappa, with k=0.6-0.8 considered good and k>0.8 excellent(24). Disagreements were resolved through reviewer discussion and referral to a third reviewer if necessary.

Data from the stratified sample of 200 reviews were extrapolated to the total ‘potentially deprescribing’ discontinuations through multiplying sample deprescribing prevalence within each reason statement by the total number of discontinuations within each reason statement. These were summed to estimate the total proportion and 95% confidence interval (95% CI) of admission medicines deprescribed in hospital and the proportion (95% CI) which were reactive and proactive.

ResultsFrom 24,552 admission medicines prescribed for 2,309 patients, 977 discontinuations were recorded across 415 patients, of which 682 (69.8%) were considered ‘potentially deprescribing’ and included in the analysis according to the e-prescribing reason. Females constituted 228 (54.9%) patients and the median (IQ) age was 79.0 (66.0, 86.0) years.

Unambiguous e-prescribing reasons accounted for 21.0% of the sample. The remaining 158 (79.0%) records were evenly sampled from the ambiguous e-prescribing reason “No longer clinically necessary” and from no e-prescribing reason recorded.

One-hundred and forty-three (71.5%) discontinuations reviewed were not consistent with the definitions for proactive or reactive deprescribing for the reasons; end of life care, treatment escalation or the medication being stopped in error. For a further 13 (6.5%), insufficient information was available for categorisation. The remaining 44 (22.0%) confirmed deprescribing activities were categorised into 7 (15.9%) proactive and 37 (84.1%) reactive. Agreement between reviewers categorising deprescribing activity was excellent (κ=0.872, p<0.01).

Reasons provided in the medical records for medication deprescribed reactively were; side effect (21 (56.8%)), acute kidney injury (8 (21.6%)), treatment failure (5 (13.5%)), swallowing difficulty (1 (2.7%)), allergic reaction (1 (2.7%)) and interaction with other treatment (1 (2.7%)). All proactive deprescribing was in response to resolution of the indication for which the medication was first prescribed as reported by the patient or physiological parameters.

Extrapolation of the 200 stratified sample data to the 682 total discontinuations yielded 22.01% (19.0%-25.2%) consistent with deprescribing, of which 19.2% (12.9%-25.5%) are proactive and 80.8% (75.5%-87.1%) are reactive. This corresponds to 0.6% (0.5%-0.7%) of all admission medications prescribed being deprescribed.

DiscussionVery limited deprescribing activity was identified in this one UK hospital. Dominance of reactive deprescribing suggests that prescribers require the presence of a clinical trigger such as an adverse drug event to prompt deprescribing. The low levels of proactive deprescribing are in accordance with primary care research which reports that practitioners find it challenging to evaluate potential risks and harms associated with medication to inform deprescribing(6). It is conceivable that hospital practitioners may also find this challenging.

4

Findings from the present study endorse this hypothesis, as the observed proactive deprescribing was only in cases with documented evidence of no clinical benefit thus only potential for harm. There was therefore no proactive deprescribing identified as a result from a complex evaluation of risks and benefits.

Accepting the limitations of not assessing the prevalence of PIMs in the present study, given that the deprescribing prevalence was 0.6% it can be concluded that the vast majority of PIMs are unlikely to be being discontinued during the hospital admissions. There may therefore be scope for increasing proactive deprescribing activity in hospital. However, the extent to which this is feasible and acceptable is as yet unknown. A future study should therefore seek to explain low proactive deprescribing activity in hospital and explore the support required for prescribers and patients to increase this activity.

5

Study 2: Attitudinal predictors of older peoples’ and caregivers’ motivation to deprescribe in hospital

AimTo describe the views of older people and caregivers towards deprescribing in hospital and identify attitudinal predictors of motivation to deprescribe using the validated revised Patients’ Attitudes Towards Deprescribing questionnaire (rPATD)(25).

MethodsEthical and governance approval were obtained from the Greater Manchester West Research Ethics Committee (REC reference: 17/NW/0582) and the UK Health Research.

Study sample and settingPatients and visiting caregivers were independently recruited from Older People’s Medicine (OPM) wards at one and two UK hospitals respectively. All inpatients from OPM wards prescribed at least five medicines (polypharmacy[26]) prior to admission were eligible. Inpatients unable to speak or read English, deemed by the healthcare team as inappropriate to approach for recruitment or unable to make informed decisions about medicines were excluded.

All visitors self-reporting an unpaid role in managing the medication of an inpatient satisfying the above inclusion criteria were eligible as caregivers. People unable to speak or read English and aged <18 years old were excluded.

Recruitment and survey administrationPatients were screened for eligibility and approached for inclusion by an OPM clinician. Patients expressing an interest were approached by a researcher who provided an information leaflet and answered questions. Written, informed consent was obtained for rPATD administration and collection of demographic information. The rPATD was administered by the researcher at the bedside using an electronic tablet. Patient demographics and the number of admission medicines were recorded.

Visitors of OPM wards were screened for eligibility as caregivers and approached for recruitment by a research nurse. As no identifiable personal information was collected from caregivers, consent was implied through self-completion of the questionnaire. Caregivers who agreed to participant were provided with a questionnaire pack including an anonymous demographic information collection form and the rPATD. Caregivers were invited to self-complete the questionnaire and provide demographic information for themselves and their care recipient in addition to indicating their relationship with the care recipient and the number of prescribed admission medications. Caregivers were instructed to return the pack to a member of ward staff.

Participants were asked to respond to the rPATD reflecting on medication as prescribed prior to admission but in the context of deprescribing in the hospital setting.

6

Sample sizeNo participant data are reported for the rPATD to inform sample size estimation. Participant data from the original PATD indicate a maximum distribution across the response items of 65%:35%[16]. This was reported for the item “I feel that I am taking a large number of medications”. Assuming a similar distribution for the rPATD and anticipated minimal adaptations required for UK use, a sample of 75 participants per population provides a 95% confidence interval (CI) of ±11.0% or smaller around the estimates of agreement with each rPATD item. This sample size is therefore appropriate for ensuring that there is no overlap in CIs between the proportion of respondents agreeing and disagreeing with an rPATD item.

Statistical analysisAnalyses were performed using IBM SPSS Statistics version 23.0 for Windows. Descriptive statistics characterise the participants and rPATD responses. Items are grouped into four factors; medication burden, medication appropriateness, concerns about deprescribing and involvement in decision-making. Global item 1 captures agreement with deprescribing proposed by a doctor and global item 2 captures satisfaction with current medications.

The primary outcome of motivation to consider deprescribing without a doctor’s prompt was the appropriateness question “I would like to try stopping one of my medicines to see how I feel without it” (patients) and “I would like the doctor to try stopping one of my care recipient’s medicines to see how they feel without it” (caregivers). Unlike global question 1, they estimate agreement with deprescribing in the absence of a doctor’s prompt.

Responses to the primary outcome and two global rPATD questions were dichotomised into those in agreement (agree and strongly agree) and those ambivalent or in disagreement (strongly disagree, disagree, neither agree nor disagree).

Backward binary logistic regression was performed between statements in the four factors and the primary outcome. To identify substantive lack of motivation towards deprescribing, agreement between each statement and the primary outcome was dichotomised into those in disagreement (strongly disagree and disagree) and those ambivalent or in agreement (neither agree nor disagree, agree and strongly agree). Variables with less than 5.0% distribution in responses cross-tabulated with the outcome were excluded as it was felt that these had insufficient variability to be reliability modelled.

ResultsFigure 1 summarises recruitment of patients and caregivers; the primary reason for patient ineligibility was being unable to provide informed consent. For caregivers, non-involvement with medicines was the primary reason for exclusion.

7

8

Patients Caregivers

Screened for eligibility

n=215

Eligible

n=107

Excluded (n=108)

71 Unable to provide informed consent

15 Seriously unwell

12 Unable to make decisions about medicines

10 Other

Recruited

n=75

Not recruited (n=32)

8 Discharged 1 Deceased 22 Declined 1 Other

Screened for eligibility

n=215

Eligible

n=113

Excluded (n=102)

45 non-involvement with medicines

29 care recipient not cared for by the OPM speciality

15 care recipient prescribed <5 medicines

7 caregiver <18 5 care recipient

seriously unwell 1 caregiver

unavailable

Recruited

n=76

Not recruited (n=41)

18 declined 11 caregiver

departed without completing questionnaire

11 care recipient deteriorated

1 care recipient objected

Provided complete questionnaire data

n=75Provided complete questionnaire data

n=72

Provided incomplete questionnaire data

n=4

PatientsRecruitment of the target 75 patients from those eligible produced a recruitment rate of 70.1% (95% CI: 52.7, 87.5). The median (IQ) age was 87.0 (83.0, 90.0) years and 34 (45.3%) were female. The median (IQ) number of medications prescribed prior to admission was 8.0 (6.0, 10.0).

Responses to the rPATD questionnaire

Table 1 illustrates patients’ rPATD responses. Agreement with deprescribing proposed by a doctor was high, with 97.4% (95% CI 93.8-100.0) agreeing with global item 1. Conversely, only 29.3% (95% CI 19.0-39.6) agreed with the primary outcome of motivation to deprescribe. A further 92.0% (95% CI 85.9-98.1) agreed with global item 2, indicating high satisfaction with current medications.

Regression analysis

Items from all four factors were entered into the regression analysis. The resulting model predicted 62.9% (Negelkerke R2) of the variance and the Hosmer and Lemeshow goodness-of-fit test implied the model’s estimates fit the data to an acceptable level (p=0.238). rPATD items predicting patients’ lack of motivation to deprescribe are provided in table 2.

CaregiversThe caregiver arm over recruited by one participant producing a recruitment rate of 67.2% (95% CI: 49.9, 84.5) for the 76 caregivers who completed the questionnaire. Thirty-five (46.1%) caregivers were a spouse or partner and the remaining 41 (53.9%) were another relative. The median (IQ) age for caregivers and care recipients were 70.0 (57.0, 83.0) and 86 (83.0, 89.0) respectively. Females constituted 47 (61.8%) and 48 (63.2%) of caregivers and care recipients respectively. The median (IQ) number of medicines prescribed for care recipients prior to admission was 8.0 (6.0, 10.3).

Responses to the rPATD questionnaire

Table 2 illustrates caregivers’ rPATD responses. Agreement with deprescribing proposed by a doctor was high, with 76.3% (95% CI 66.7-85.9) of caregivers agreeing with global question 1. Conversely, only 43.5% (95% CI 32.4-54.6) agreed with the primary outcome. A further 80.3% (95% CI 71.3-89.3) agreed with global question 2, indicating high satisfaction with current medications.

Regression analysis

Question 1 from Burden and 2 from Involvement were not entered into the regression due to insufficient distribution across responses. All remaining questions across the factors were entered into the binary logistic regression analysis with the primary outcome. The resulting model predicted 70.1% of the variance (Negelkerke R2) and the Hosmer and Lemeshow goodness-of-fit test implied the model’s estimates fit the data to an acceptable level

9

(p=0.852). rPATD items predicting caregivers’ lack of motivation to deprescribe are provided in table 2.

10

Table 1 Patients’ responses to the rPATD questionnaire

ItemStrongly disagree

%Disagree

%

Neither agree nor

disagree %

Agree %

Strongly agree %

BurdenI feel the National Health Service (NHS) spends a lot of money on my medicines

1.3 5.3 9.3 58.7 25.3

Taking my medicines every day is very inconvenient 29.3 49.3 5.3 14.7 1.3

I feel that I am taking a large number of medicines 8.0 34.7 9.3 36.0 12.0

I feel that my medicines are a burden to me 28.0 50.7 6.7 9.3 5.3

Sometimes I think I take too many medicines 14.7 34.7 9.3 30.7 10.7

AppropriatenessI feel that I may be taking one or more medicines that I no longer need

17.3 38.7 10.7 22.7 10.7

I would like to try stopping one of my medicines to see how I feel without it

18.7 40.0 12.0 21.3 8.0

I would like my doctor to reduce the dose of one or more of my medicines

17.3 42.7 20.0 14.7 5.3

I think one or more of my medicines may not be working 16.0 34.7 29.3 17.3 2.7

I believe one or more of my medicines may be currently giving me side effects

28.0 38.7 5.3 20.0 8.0

ConcernsI would be reluctant to stop a medicine that I had been taking for a long time

9.3 46.7 6.7 28.0 9.3

If one of my medicines was stopped, I would be worried about missing out on future benefits

18.7 37.3 6.7 33.3 4.0

I get stressed whenever changes are made to my medicines 22.7 52.0 9.3 13.3 2.7

If my doctor recommended stopping a medicine, I would feel that he/she was giving up on me

40.0 41.3 2.7 10.7 5.3

I have had a bad experience when stopping a medicine before 24.1 37.9 17.2 13.8 6.9

InvolvementI have a good understanding of the reasons I was prescribed each of my medicines

8.0 4.0 8.0 50.7 29.3

I know exactly what medicines I am currently taking, and/or I keep an

6.7 13.3 4.0 37.3 38.7

11

up-to-date list of my medicinesI like to know as much as possible about my medicines 4.0 13.3 6.7 46.7 29.3

I like to be involved in making decisions about my medicines with my doctors

8.0 25.3 8.0 30.7 28.0

I always ask my doctor, pharmacist or other healthcare professional if there is something I don’t understand about my medicines

1.3 17.3 2.7 44.0 34.7

GlobalIf my doctor said it was possible I would be willing to stop one or more of my regular medicines

1.3 1.3 0 66.7 30.7

Overall, I am satisfied with my current medicines 0 1.3 6.7 65.3 26.7

12

Table 2 rPATD items predicting patients’ and caregivers’ lack of motivation to deprescribe without a doctor’s prompt

Predictor rPATD item Unadjusted odds ratio

(95% confidence

interval)

p-value Adjusted odds ratio

(95% confidence

interval)

p-value

PatientsSometimes I think I take too many medicines

0.072 (0.023-0.231) <0.001 0.195 (0.045-

0.841) 0.028

I feel that I may be taking one or more medicines that I no longer need

0.075 (0.025-0.229) <0.001 0.179 (0.044-

0.726) 0.016

I would like my doctor to reduce the dose of one or more of my medicines

0.066 (0.021-0.206) <0.001 0.199 (0.050-

0.787) 0.021

CaregiversI feel that the person that I care for may be taking one or more medicines that they no longer need

0.092 (0.030-0.279) <0.001 0.044 (0.006-

0.310) 0.002

I would like the doctor to reduce the dose of one or more of my care recipient’s medicines

0.025 (0.007-0.094) <0.001 0.024 (0.004-

0.137) <0.001

13

Table 3 Caregivers’ responses to the rPATD questionnaire

Item Strongly disagree

%

Disagree %

Neither agree nor

disagree %

Agree %

Strongly agree %

BurdenI feel the National Health Service (NHS) spends a lot of money on my care recipient’s medicines

0 6.6 36.8 34.2 22.4

I feel that the person I care for is taking a large number of medicines 5.3 18.4 23.7 40.8 11.8

I feel that my care recipient’s medicines are a burden to them 10.5 48.7 18.4 21.1 1.3

Sometimes I think the person I care for takes too many medicines 10.5 27.6 30.3 27.6 3.9

AppropriatenessI feel that the person that I care for may be taking one or more medicines that they no longer need

5.3 28.9 28.9 32.9 3.9

I would like the doctor to try stopping one of my care recipient’s medicines to see how they feel without it*

9.2 28.9 18.4 38.2 5.3

I would like the doctor to reduce the dose of one or more of my care recipient’s medicines

9.2 28.9 38.2 21.1 2.6

I think one or more of my care recipient’s medicines may not be working

6.6 28.9 42.1 22.4 0

I believe one or more of my care recipient’s medicines may be currently giving them side effects

7.9 32.9 23.7 30.3 5.3

ConcernsI would be reluctant to stop one of my care recipient’s medicines that they had been taking for a long time

2.6 26.3 17.1 46.1 7.9

I get stressed whenever changes are made to my care recipient’s medicines

21.1 36.8 25.0 17.1 0

I feel that if I agreed to stopping one of my care recipient’s medicines then this is giving up on them

19.7 38.2 21.1 17.1 3.9

The person that I care for has had a bad experience when stopping a medicine before

17.1 28.9 7.9 6.6 0

InvolvementI know exactly what medicines the person that I care for is currently taking and/or I have an up-to-date list of their medicines

0 15.8 3.9 53.9 21.1

14

I like to know as much as possible about my care recipient’s medicines

0 3.9 10.5 51.3 28.9

I like to be involved in making decisions about my care recipients medicines with their doctors

2.6 13.2 13.2 43.4 22.4

I always ask the doctor, pharmacist or other healthcare professional if there is something I don’t understand about my care recipient’s medicines

1.3 13.2 9.2 51.3 19.7

GlobalIf their doctor said it was possible I would be willing to stop one or more of my care recipient’s medicines

2.6 1.3 18.4 65.8 10.5

Overall, I am satisfied with my care recipient’s current medicines 2.6 2.6 13.2 67.1 13.2

15

DiscussionEngagement of patients and caregivers is a core component of deprescribing, yet a substantial proportion indicated unwillingness to participate in medication decision-making. Furthermore, the low motivation to consider deprescribing without a doctor’s prompt is in agreement with substantial proportions of participants declining deprescribing in the trial environment(10–13). Deprescribing interventions should therefore focus on increasing patient and caregiver engagement with and motivation to deprescribe. The three diagnostic indicators for establishing motivation are perceptions of the number and necessity of medicines and, a desire for dose reduction. These may also assist physicians with targeting relevant attitudinal predictors during deprescribing discussions.

Some caution should be applied to this message, as whilst the majority of respondents agreed with deprescribing proposed by a doctor, they also reported content with existing medication. This desire to conform may lead to agreement with a doctor’s recommendation to deprescribe despite concerns(26) and reluctance to report adverse outcomes such as return of symptoms(27).

Preference for a passive role by older people(28) is endorsed by both patient and caregiver responses to items in the rPATD involvement factor. Whilst items relating to the passive behaviour of knowledge acquisition regarding prescribed medicines attracted high agreement, the item relating to active engagement with medication decision-making was lower.

16

Study 3: A blueprint for deprescribing in hospital: A focus group study with geriatricians and pharmacists using the Theoretical Domains Framework (TDF)

Aims Understand geriatricians’ and pharmacists’ perceived influencers of deprescribing in

hospital and whether these differ between district general and teaching hospital contexts.

For the influencers of deprescribing which require behaviour change, identify potentially effective intervention components for inclusion in a novel hospital deprescribing intervention for older people were identified.

MethodsEthical and governance approvals were obtained from the Faculty of Medicine and Health Sciences Research Ethics Committee, University of East Anglia (Reference: 2017/2018–59)) and the UK Health Research Authority (Project ID: 231262) respectively.

Theoretical approach

The Medical Research Council guidance on developing complex interventions emphasises the importance of applying theory to understand the processes of change required to adopt a new behaviour such as deprescribing(29). Interventions developed using this methodological approach are more likely to be successfully implemented versus those using an atheoretical design approach(30,31).

This study was underpinned by the Theoretical Domains Framework (TDF), an integrative framework of behaviour change theories organised into 14 theoretical domains(32). The domains represent determinants of behaviour, for example the ‘Knowledge’ domain refers to “an awareness of the existence of something”(32). Identification of the domains which are important to the target behaviour provides the theoretical understanding required to develop a BCI(32). Each domain is linked to a taxonomy of 93 Behaviour Change Techniques (BCTs)(33), which are the ‘active ingredients’ or ‘building blocks’ of interventions that lead to behaviour change.

SampleFour UK hospitals were selected (two large teaching hospitals and two district general hospitals) to explore influencers of deprescribing for older people during a hospital admission and capture any variation arising from differing resources and patient populations. Registrars and consultant geriatricians, and specialist (UK National Health Service Agenda for Change band 7) and senior (UK National Health Service Agenda for Change band 8) hospital clinical pharmacists who worked in the four hospital study sites were eligible for inclusion. Practitioners with a principally non-clinical role (less than two ward-based days per week pro rata) were excluded.

17

To explore similarities and differences in attitudes between geriatricians and pharmacists, and identify hospital characteristics that might influence deprescribing, the study was designed to include eight focus groups. Each hospital site hosted two focus groups: one with geriatricians and one with pharmacists, each consisting of between five and eight participants to allow individuals to contribute but at the same time to foster group discussions around deprescribing (n=40 to 64 participants in total).

To determine whether data saturation had been achieved, the principles for deciding saturation in theory-based qualitative studies outlined by Francis et al. were followed[22]. However, as the study objective to explore any differences in views between district general and teaching hospital contexts, an a priori decision was made to convene all eight planned focus groups irrespective of whether data saturation was achieved prior to conducting all of the groups.

Data collectionA semi-structured topic guide informed by the deprescribing literature was developed iteratively by a team of academics, geriatricians, clinical pharmacists, and patient and family representatives. Guiding questions were designed to illicit participants’ views regarding the following:

1. Perception of existing deprescribing practice2. Barriers to increasing deprescribing practice3. Enablers for increasing deprescribing practice

Focus group data were collected between February and May 2018.

AnalysisFocus group recordings were transcribed verbatim by a research administrator, anonymised and checked for accuracy by a researcher (SS). Data were imported into NVivo 11 (QSR International, Melbourne, Australia) to facilitate the following three phases of data analysis: 1. Thematic analysis to identify determinants of deprescribing for older people in hospital, 2. Mapping of all identified determinants of deprescribing to the TDF and prioritising mapped TDF domains for targeting in a deprescribing intervention 3. Identifying the BCTs linked to these domains. This methodological approach draws on recent theory-based intervention development research(34–36).

ResultsAll geriatricians and pharmacists who were purposively sampled, participated in the focus groups. Fifty-four participants; 28 geriatricians and 26 pharmacists, participated across the eight focus groups. The mean (SD) focus group duration was 55 (5) minutes.

Phase 1: Thematic analysisFour themes were identified: 1. Role of different professionals, 2. The inpatient environment 3. Consideration of outcomes, 4. Attitudes towards medicines. Themes were recurring after the third focus group and no new themes emerged after the sixth focus group.

18

In accordance with the planned analysis, themes 1 and 2 represented novel concepts and are therefore comprehensively reported. Themes 3 and 4 are briefly described because they overlap extensively with the primary care literature, therefore only divergences are highlighted. No discernible differences between participants from teaching and district general hospitals or prescribing and non-prescribing pharmacists were evident.

1. Role of different professionals

There was high motivation to increase deprescribing in hospital, with both geriatrician and pharmacist participants recognising that existing deprescribing practice in hospital is limited and dominated by reactive behaviour.

Geriatricians and pharmacists acknowledged that increasing deprescribing practice aligned with the generalist nature of their professions’ roles and responsibilities. They indicated that this generalist nature meant that they could assume key roles in the deprescribing process. There was also agreed scope for other healthcare practitioners such as nurses and physiotherapists to support deprescribing in hospital. However, the role of practitioners with a restricted focus, such as therapeutic area specialists, was described as potentially incompatible with deprescribing for older people.

Participants indicated that junior healthcare practitioners, including junior geriatricians and pharmacists, lacked the required competencies to participate in deprescribing. This was a recognised consequence of a lack of experience and limited education provided to trainee healthcare practitioners regarding deprescribing.

Whilst the scope for several professional groups working in hospital to contribute to deprescribing was recognised, there was consensus that overall responsibility for deprescribing in hospital rested with one nominated professional group. Participants from both professional groups agreed that geriatricians should be the professional group nominated to assume overall responsibility for deprescribing decision-making in hospital. Furthermore, in the geriatrician focus groups many participants expressed confidence in their ability to weigh up the risks and benefits of deprescribing to inform decision-making. For geriatricians therefore, the principle barriers to deprescribing were the environmental and resource factors in the hospital setting.

Pharmacist participants suggested their skill set better aligns with identifying potentially inappropriate medication and advising on deprescribing, which was endorsed by geriatrician participants. Pharmacist participants were reluctant to assume overall responsibility for deprescribing because of an expressed lack of confidence in decision-making.

However, existing working patterns and priorities in hospital was reported to limit pharmacists’ capacity to assume any role in deprescribing.

2. The inpatient environment

The influence of the inpatient environment on deprescribing was discussed in relation to the interacting dynamics of the clinical picture, communication and access to resources. Both geriatrician and pharmacist participants acknowledged that many older people in hospital were prescribed inappropriate medication.

19

The scope to increase deprescribing was complemented by the necessary resources and capacity available in the hospital environment to safely trial deprescribing and monitor patients’ responses to medication withdrawal.

Furthermore, geriatrician and pharmacist participants were reassured by the network of other specialist healthcare practitioners working in the hospital setting available to support the deprescribing process.

Conversely, significant challenges to deprescribing in the hospital setting were acknowledged. The principal barrier identified was limited information regarding patients’ medications available to hospital practitioners. Strength, dose and formulation of patients’ usual medications were routinely ascertained by pharmacists on admission through discussion with patients/family, examining of medications brought into the hospital or through accessing electronic medication records. However, key information required to determine whether deprescribing was appropriate, such as the reason for prescribing a medication and the duration of treatment, were described as rarely known to hospital practitioners.

Recent advances including some hospitals gaining access to primary care-held comprehensive medication records were regarded as a potential solution to this problem by some participants. However, discussions across both professional groups were dominated by scepticism towards navigating records in search for the required information. This was described as a time consuming, impractical and often unsuccessful exercise.

The acute nature of a hospital admission, focussing on prioritising patients’ problems requiring immediate action was considered a barrier to proactive deprescribing. This was confounded by patients’ artificial lifestyle whilst in hospital, including acute immobility, scheduled meals and medications being managed by healthcare practitioners. These factors were perceived to potentially distort the assessments undertaken in hospital to inform long-term deprescribing decisions.

In response, participants asserted a clear need to establish a safety net through sharing information with primary care providers responsible for ongoing care after patients are discharged from hospital. It was suggested that such correspondence could include directions for primary care practitioners to monitor for changes which may indicate re-prescribing is necessary. However, participants agreed that the existing transfer of information between care settings is poor and may undermine deprescribing efforts.

Similarly, pharmacist participants voiced concerns regarding communicating medication changes with patients in hospital. Barriers identified included patients being unable to participate in decision-making because of ill health and deprescribing being regarded by patients as a low priority relative to the acute condition responsible for the admission. However, the availability of family members as both sources of medication information and participants in decision-making were facilitators identified by pharmacists and geriatricians.

3. Consideration of outcomes

The perceived risks and potential benefits of deprescribing versus continuing to prescribe were identified by participants as key factors influencing deprescribing behaviours, with decisions predicated on finely balancing the medication, the patient’s clinical condition and their preferences. Potential patient orientated positive outcomes arising from deprescribing were reduced medication burden and incidence of adverse drug events leading to improved quality of life. A reduction in unnecessary medication expenditure, reduced treatment costs

20

associated with adverse drug events and rationalising use of health resources were suggested as potential benefits to healthcare systems. In turn, patient and health system benefits were proposed to lead to individual practitioner benefits, with geriatrician and pharmacist participants suggesting deprescribing may lead to reduced workload.

There was divergence in views between geriatrician and pharmacist participants regarding whether deprescribing was perceived to carry greater risk than continuing potentially inappropriate medicines (PIMs). Pharmacist participants felt that on balance, passively continuing to prescribe a medicine in the absence of an immediate need to deprescribe was safer than proactively deprescribing. Conversely, geriatricians felt that both deprescribing and continuing to prescribe were active decisions, with no inherent differences in risk between the two decisions providing both are carefully considered.

4. Attitudes towards medicines

Patient, family, healthcare provider, and wider societal attitudes towards medicines were reported to influence geriatrician and pharmacist participants’ deprescribing behaviour. Some patients and family attachment to long-term medication was a barrier identified to deprescribing, which participants believed would be difficult to overcome despite explaining the risks and limited benefits of continued use. Conversely, participants characterised a significant proportion of older people who disliked taking medication and are amenable to deprescribing. Healthcare culture’s attitudes towards medication use was acknowledged as changing in favour of deprescribing efforts, with the burden of inappropriate medication use increasingly recognised, leading to calls for medicines optimisation initiatives. Nevertheless, healthcare culture’s positive attitudes towards deprescribing were not perceived by participants to be reflected in prescribing recommendations, with overemphasis on commencing treatment and little consideration of ongoing reviews and deprescribing.

Phase 2: Mapping of all determinants of deprescribing to the TDF and prioritising mapped TDF domains for targeting in a deprescribing interventionFigure 1 provides the TDF domains which were mapped onto the four themes. TDF domains prioritised for targeting in an intervention are indicated. ‘Social professional role and identity’ and ‘Knowledge’ are not prioritised domains because both geriatricians and pharmacists asserted that deprescribing aligns with existing perceptions of their complementary roles and knowledge. Similarly, ‘Beliefs about capabilities’ is not represented because both professions had confidence in their ability to undertake the roles that they had defined for themselves. It can be seen that the ‘Emotion’ domain, which was exclusively expressed with barriers by pharmacists being fearful of assuming responsibility for any negative consequences of deprescribing, is not represented. This is in recognition of all participants agreeing that the geriatrician should be the professional assuming overall responsibility for deprescribing in hospital.

21

Figure 1 Themes (inner four quadrants) mapped to nine TDF domains (outer ring). TDF domains prioritised for targeting in an intervention are emboldened in yellow text.

22

Phase 3. Identifying the BCTs linked to prioritised TDF domainsTable 1 provides the 44 BCTs linked to the prioritised TDF domains(37).

Table 1 Potentially effective BCTs for a hospital deprescribing intervention for older people

Prioritised theoretical domain Linked BCTs

Social influence

Social comparisonSocial support or encouragement (general)Information about others’ approvalSocial support (emotional)Social support (practical)Vicarious reinforcement*Restructuring the social environment*Modelling or demonstrating the behaviourIdentification of self as role modelSocial reward*

Beliefs about consequences

Emotional consequencesSalience of consequencesCovert sensitizationAnticipated regretSocial and environmental consequencesComparative imagining of future outcomesVicarious reinforcement*Threat*Pros and consCovert conditioning

Environmental Context and Resources

Restructuring the physical environmentDiscriminative (learned) cuePrompts/cuesRestructuring the social environment*Avoidance/changing exposure to cues for the behaviour

Goals

Goal setting (outcome)Goal setting (behaviour)Review of outcome goal(s)Review behaviour goalsAction planning (including implementation intentions)

Reinforcement

Threat*Self-rewardDifferential reinforcementIncentiveThinningNegative reinforcementShapingCounter conditioningDiscrimination trainingMaterial rewardSocial reward*Non-specific rewardResponse costAnticipation of future rewards or removal of punishmentPunishmentExtinctionClassical conditioning

23

DiscussionBoth geriatricians and pharmacists perceive that deprescribing is a part of their generalist role, that it is feasible within the hospital environment and worthwhile. Barriers to implementation are the patient’s acute problems and lack of comprehensive communication about medicines. The potential for adverse outcomes was perceived by pharmacists to be a more significant barrier to deprescribing than by geriatricians. A hospital deprescribing intervention for older people should therefore be led by geriatricians and supported by pharmacists. The intervention specification should address the five prioritised TDF domains through selecting from the 44 potentially effective BCTs for deprescribing in hospital, those which are appropriate for the health system contexts.

Despite variations in practice, organisational structure and resource availability, no discernible differences in the influencers of deprescribing were identified between small district general and larger teaching hospitals across the four emergent themes. The generalisability between barriers and enablers from district general and teaching hospitals and the transferability of two of the four themes with the primary care context affords some confidence in the BCTs identified being relevant to other hospital contexts(8,38). Capturing the perspectives of the two professions primarily responsible for prescribing decisions for older people in hospital has allowed exploration of a wide range of potential barriers and enablers. The potentially effective BCTs therefore offer the efficiency of aligning with the needs of these two pivotal professional groups.

The influencers of whether geriatricians and pharmacists deprescribe in the hospital setting that require targeting are the prioritised TDF domains of ‘Social influence’, ‘Beliefs about consequences’, ‘Environmental context and resources’, ‘Goals’ and ‘Reinforcement’. The perception of geriatricians and pharmacists that they should routinely be deprescribing as part of their role and have the required knowledge makes the hospital setting a strong candidate for developing a novel deprescribing intervention. For the five prioritised TDF domains, 44 evidence based BCTs are available for selection, providing a blueprint for developing and implementing a hospital deprescribing intervention. Selection of BCTs from this blueprint for operationalising may vary across different health system contexts. For example, the BCTs restructuring the physical and social environment by implementing deprescribing in an outpatient hospital clinic context may be may be practical for some health systems but not for others.

24

Future recommendationsThis programme of work has identified that there is scope for the development of a novel hospital deprescribing intervention for older people, targeted at geriatricians and pharmacists. Engagement with the key stakeholders relevant to deprescribing in hospital and the application of behaviour change theory has guided identification of potentially effective intervention components to facilitate deprescribing in hospital. This provides a blueprint of intervention components for selection when developing a novel hospital deprescribing intervention for older people.

Future research should focus on selection of intervention components from the blueprint prior to definitive intervention testing. Selection of components from the blueprint for operationalising will vary across different health systems contexts according to local factors such as resource availability. The APEASE criteria offer those wishing to draw from the blueprint a systematic and tailored approach to selecting the intervention components most appropriate for the health system context by considering the six factors of Affordability, Practicability, Effectiveness and cost-effectiveness, Acceptability, Side effects/safety and Equity, which are all equally relevant to intervention success(39).

Subsequent research should then model and refine the draft intervention through a series of cycles, followed by a feasibility study to ensure acceptability to patients, caregivers and healthcare practitioners and, to ensure sufficient information to design a trial. Initially, the intervention should be piloted to determine whether all selected intervention components function together as expected, followed by a definitive trial.

25

References1. Spinewine A, Schmader KE, Barber N, Hughes C, Lapane KL, Swine C, et al.

Appropriate prescribing in elderly people: how well can it be measured and optimised? Lancet. 2007;370(9582):173–84.

2. Gallagher P, Lang PO, Cherubini A, Topinková E, Cruz-Jentoft A, Montero Errasquín B, et al. Prevalence of potentially inappropriate prescribing in an acutely ill population of older patients admitted to six European hospitals. Eur J Clin Pharmacol. 2011;67(11):1175–88.

3. Woodward MC. Deprescribing: achieving better health outcomes for older people through reducing medications. J Pharm Pract Res. 2003;33(4):323–8.

4. Hoffmann T, Jansen J, Glasziou P. The importance and challenges of shared decision making in older people with multimorbidity. PLoS Med. 2018;15(3):e1002530.

5. Scott IA, Hilmer SN, Reeve E, Potter K, Le Couteur D, Rigby D, et al. Reducing inappropriate polypharmacy: the process of deprescribing. JAMA Intern Med. 2015;175(5):827–34.

6. Anderson K, Foster M, Freeman C, Luetsch K, Scott I. Negotiating “Unmeasurable Harm and Benefit”: Perspectives of General Practitioners and Consultant Pharmacists on Deprescribing in the Primary Care Setting. Qual Health Res. 2017;27(13):1936–47.

7. T. PA, M. CR, Kathleen P, Darren S, D. EC. The feasibility and effect of deprescribing in older adults on mortality and health: a systematic review and meta‐analysis. Br J Clin Pharmacol. 2016;82(3):583–623.

8. Anderson K, Stowasser D, Freeman C, Scott I. Prescriber barriers and enablers to minimising potentially inappropriate medications in adults: a systematic review and thematic synthesis. BMJ Open. 2014;4(12).

9. Reeve E, Shakib S, Hendrix I, Roberts MS, Wiese MD. Review of deprescribing processes and development of an evidence‐based, patient‐centred deprescribing process. Br J Clin Pharmacol. 2014;78(4):738–47.

10. Beer C, Loh P-K, Peng Gee Y, Potter K, Millar A. A pilot randomized controlled trial of deprescribing. Ther Adv drug Saf. 2011;2(2):37.

11. Luymes CH, Poortvliet RKE, van Geloven N, de Waal MWM, Drewes YM, Blom JW, et al. Deprescribing preventive cardiovascular medication in patients with predicted low cardiovascular disease risk in general practice – the ECSTATIC study: a cluster randomised non-inferiority trial. BMC Med [Internet]. 2018;16(1):5. Available from: https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-017-0988-0

12. Potter K, Flicker L, Page A, Etherton-Beer C. Deprescribing in frail older people: a randomised controlled trial. PLoS One. 2016;

13. Komagamine J, Sugawara K, Hagane K. Characteristics of elderly patients with polypharmacy who refuse to participate in an in-hospital deprescribing intervention: a retrospective cross-sectional study. BMC Geriatr. 2018;18(1):96.

14. Bogardus ST, Bradley EH, Williams CS, Maciejewski PK, Gallo WT, Inouye SK. Achieving Goals in Geriatric Assessment: Role of Caregiver Agreement and Adherence to Recommendations. J Am Geriatr Soc. 2004;52(1):99–105.

15. Look KA, Stone JA. Medication management activities performed by informal

26

caregivers of older adults. Res Soc Adm Pharm. 2017;14(5):418–26.

16. Reeve E, To J, Hendrix I, Shakib S, Roberts MS, Wiese MD. Patient barriers to and enablers of deprescribing: a systematic review. Drugs Aging. 2013;30(10):793–807.

17. Spinewine A, Swine C, Dhillon S, Franklin BD, Tulkens PM, Wilmotte L, et al. Appropriateness of use of medicines in elderly inpatients: qualitative study. BMJ. 2005;331(7522):935.

18. Thillainadesan J, Gnjidic D, Green S, Hilmer SN. Impact of Deprescribing Interventions in Older Hospitalised Patients on Prescribing and Clinical Outcomes: A Systematic Review of Randomised Trials. Drugs Aging [Internet]. 2018 Apr;35(4):303–19. Available from: https://doi.org/10.1007/s40266-018-0536-4

19. Royal Pharmaceutical Society. Professional Standards for Hospital Pharmacy [Internet]. 2017 [cited 2018 Aug 20]. Available from: https://www.rpharms.com/Portals/0/RPS document library/Open access/Professional standards/Professional standards for Hospital pharmacy/Hospital Standards-2017.pdf?ver=2017-12-21-132808-697

20. Medicine IWP to P and RA. Academic medicine: the evidence base. BMJ Br Med J. 2004;329(7469):789.

21. Papanikolaou PN, Christidi GD, Ioannidis JPA. Patient outcomes with teaching versus nonteaching healthcare: a systematic review. PLoS Med. 2006;3(9):e341.

22. Berman RA, Green J, Kwo D, Safìan KF, Botnick L. Severity of illness and the teaching hospital. J Med Educ. 1986;61(1):1–9.

23. Marvin V, Ward E, Poots AJ, Heard K, Rajagopalan A, Jubraj B. Deprescribing medicines in the acute setting to reduce the risk of falls. Eur J Hosp Pharm. 2017;24(1):10–5.

24. McHugh ML. Interrater reliability: the kappa statistic. Biochem medica. 2012;22(3):276–82.

25. Reeve E, Low L-F, Shakib S, Hilmer SN. Development and Validation of the Revised Patients’ Attitudes Towards Deprescribing (rPATD) Questionnaire: Versions for Older Adults and Caregivers. Drugs Aging. 2016;1–16.

26. Belcher VN, Fried TR, Agostini J V., Tinetti ME. Views of older adults on patient participation in medication-related decision making. J Gen Intern Med. 2006;21(4):298–303.

27. Wrede-Sach J, Voigt I, Diederichs-Egidi H, Hummers-Pradier E, Dierks M-L, Junius-Walker U. Decision-Making of Older Patients in Context of the Doctor-Patient Relationship: A Typology Ranging from “Self-Determined” to “Doctor-Trusting” Patients. Int J Family Med. 2013;2013:1–10.

28. Benbassat J, Pilpel D, Tidhar M. Patients’ preferences for participation in clinical decision making: A review of published surveys. Behav Med. 1998;24(2):81–8.

29. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. Developing and evaluating complex interventions: the new Medical Research Council guidance. Bmj [Internet]. 2008;1655(September):a1655. Available from: http://www.bmj.com/lookup/doi/10.1136/bmj.a1655

30. Albarracín D, Gillette JC, Earl AN, Glasman LR, Durantini MR, Ho M-H. A test of major assumptions about behavior change: a comprehensive look at the effects of passive and active HIV-prevention interventions since the beginning of the epidemic.

27

Psychol Bull. 2005;131(6):856.

31. Noar SM, Zimmerman RS. Health Behavior Theory and cumulative knowledge regarding health behaviors: Are we moving in the right direction? Health Educ Res. 2005;20(3):275–90.

32. Cane J, O’Connor D, Michie S. Validation of the theoretical domains framework for use in behaviour change and implementation research. Implement Sci. 2012;7(1):37.

33. Atkins L, Francis J, Islam R, O’Connor D, Patey A, Ivers N, et al. A guide to using the Theoretical Domains Framework of behaviour change to investigate implementation problems. Implement Sci [Internet]. 2017;12(1):77. Available from: http://implementationscience.biomedcentral.com/articles/10.1186/s13012-017-0605-9

34. Dyson J, Lawton R, Jackson C, Cheater F. Does the use of a theoretical approach tell us more about hand hygiene behaviour? The barriers and levers to hand hygiene. J Infect Prev. 2011;12(1):17–24.

35. Alexander KE, Brijnath B, Mazza D. Barriers and enablers to delivery of the Healthy Kids Check: an analysis informed by the Theoretical Domains Framework and COM-B model. Implement Sci. 2014;9(1):60.

36. Patey AM, Islam R, Francis JJ, Bryson GL, Grimshaw JM, Driedger M, et al. Anesthesiologists’ and surgeons’ perceptions about routine pre-operative testing in low-risk patients: Application of the Theoretical Domains Framework (TDF) to identify factors that influence physicians’ decisions to order pre-operative tests. Implement Sci. 2012;7(1):1–13.

37. Cane J, Richardson M, Johnston M, Ladha R, Michie S. From lists of behaviour change techniques (BCTs) to structured hierarchies: Comparison of two methods of developing a hierarchy of BCTs. Br J Health Psychol. 2015;20(1):130–50.

38. Smith B. Generalizability in qualitative research: misunderstandings, opportunities and recommendations for the sport and exercise sciences. Qual Res Sport Exerc Heal [Internet]. 2018;10(1):137–49. Available from: https://doi.org/10.1080/2159676X.2017.1393221

39. Michie S, Atkins L, Robert W. The Behaviour Change Wheel: A Guide to Designing Interventions. First edit. Silverback Publishing; 2014.

28

Patient and public involvementOn receipt of this award NN (patient representative) and JG (caregiver representative) were appointed to the project management group as lay representatives. Both NN and JG have had significant input at all critical stages of the research, from developing and refining the projects to interpretation of results and preparation of manuscripts for publication. In addition to attending formal research management group meetings, the author and lay representatives were in regular contact via telephone and email. A subsequent research application has been developed in collaboration with the lay representatives to progress and build on the findings of the present programme of work.

Additionally, one of the three projects were dedicated to exploring the views of patients and caregivers towards deprescribing in hospital. The survey of 151 patients and caregivers and the results supporting deprescribing in addition to endorsement by the lay representatives in hospital further supports this programme of work.

29

Personal developmentThe research and training activities undertaken as part of this award has enabled the researcher to develop professionally in several aspects.

Perhaps one of the most important and rewarding developmental aspects of this award for the author has been experience managing a research grant and chairing a project management committee at a very early research career stage. Several soft skills including organisation and time management, financial management and communication have been developed, equipping the author for an academic research career. The opportunity to work with meaningful patient and public involvement is relatively rare for PhD level research. The author has thoroughly enjoyed working alongside NN and JG through this award, who have both had a significant positive impact on the research, the author’s professional development and PhD.

Through attending the pre-planned PRUK-funded training at the Centre for Behaviour Change, University College London, the researcher has gained the required knowledge and skills to apply relevant behaviour change theory to the development of behaviour change interventions. This has not only supported the researcher in the present programme of work and their PhD, it has also fostered collaborative opportunities with other research projects and researchers who have sought the expertise of the author.

The research award has also enabled wider collaboration with National Health Service institutions through adoption of two of the three projects onto the National Institute of Health Research portfolio. Subsequently, the author has been able to work closely with the Clinical Research Network to develop recruitment strategies. This experience will be of significant benefit in relation to future research projects and the relationships with these organisations will potentially provide further collaborative research opportunities.

The author also acknowledges the benefit of the PRUK application process to their PhD, specifically, the experience of proposing projects to a peer review panel and defending these at award interview. This experience has accelerated the researcher’s development and significantly improved the quality of work produced at this early research career stage.

30