Programme Grants for Applied Research (Dec 2023)
The Care Home Independent Pharmacist Prescriber Study (CHIPPS): development and implementation of an RCT to estimate safety, effectiveness and cost-effectiveness
Abstract
Background Medicine prescribing, monitoring and administration in care homes can be significantly enhanced. Effective interventions to improve pharmaceutical care and resident outcomes are required. The enablement of pharmacists to prescribe provides an opportunity for pharmacist independent prescribers to assume responsibility for improving pharmaceutical care, medication-related outcomes and resident safety whilst reducing general practitioner workload. Objective(s) To determine the effectiveness and cost-effectiveness of pharmacist independent prescribing in care homes. Design Development work was undertaken through five work packages before the delivery of the definitive trial. Triads of pharmacist independent prescribers, care home and general practice with responsibility over 20 care home residents were recruited and cluster randomised to intervention or usual care for 6 months. Researchers were blinded at recruitment stage only. Recruitment of 880 residents was required to provide 80% statistical power, to show a 21% reduction in falls over 6 months, assuming 20% attrition. Randomisation was undertaken electronically at triad level, stratified by geographical area. Intention-to-treat analysis undertaken using a negative binomial model. Parameters were estimated using a generalised estimating equation approach. Costs were captured from an NHS perspective. Quality of life (EuroQol; five domain; five level) was collected by proxy to enable cost/quality-adjusted life-year estimation. A concurrent process evaluation was performed. Safety was monitored through a review of pharmacist independent prescriber activities, independent concerns reporting and review of adverse events. Participants Forty-nine triads of general practitioners, pharmacist independent prescribers and care homes were recruited with 454 residents allocated to the intervention arm and 428 to the control arm. Intervention Medication review and care planning, medication reconciliation, staff training, support with care home medication-related procedures, deprescribing and authorisation of monthly prescriptions. Main outcome measure Fall rate per person over 6 months. Results Data for 449 intervention and 427 control residents available for final analysis. The 6-month fall rate ratio in favour of intervention was 0.91 (95% confidence interval 0.66 to 1.26; p=0.58). No significant difference in secondary outcomes was identified except Drug Burden Index (rate ratio 0.83, 95% confidence interval 0.75 to 0.92; p<0.001). No harms were identified. One quarter of medication-related interventions were associated with a reduced risk of falls. The intervention was positively received. Limitations Participant self-selection bias may have affected the generalisability of findings. Open-label cluster randomised controlled trial limited by 6-month follow-up. Potential ceiling effect due to concurrent pharmacist-led interventions. Falls potentially insufficiently proximal to the intervention. Conclusions To enhance effectiveness and acceptance of the proposed model, effective integration into care home and general practitioner teams was identified as a central requirement. A core outcome set and a training package were developed. The final model of care, whilst being safe and well received and resulting in a reduction in drug burden, demonstrated no improvement in the primary outcome of falls. With no improvement in quality-adjusted life-years identified, the pharmacist independent prescriber intervention was not estimated to be cost-effective. Future work To develop and evaluate better models of care for enhancing medication outcomes and safety in care homes or re-test with a longer intervention and follow-up period and a stronger primary outcome. Trial registration This trial is registered as ISRCTN10663852, definitive trial: ISRCTN17847169. Study registration This study is registered as PROSPERO CRD20150907. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-0613-20007) and is published in full in Programme Grants for Applied Research; Vol. 11, No. 10. See the NIHR Funding and Awards website for further award information. Plain language summary The purpose of this study was to explore whether a pharmacist who can prescribe drugs could work with care homes and general practitioners to improve how medicines are prescribed, how they are monitored to see whether they are working or causing problems and how the medicines are then given to the residents. The question was whether this approach was likely to be safe, to improve care for residents and to be a good way for utilising NHS money. The project included six parts: Listening to everyone to help us design a service to create something that was likely to be acceptable and effective Thinking about what the best way would be to capture whether the service worked or not [i.e. what outcome(s) to measure] Thinking about the costs and benefits of the service and how best to capture these to find out whether the service was likely to provide value for money to the NHS Designing a training package for the pharmacists to increase the chances of them being effective in their role Testing the study design to make sure that we had thought about everything Running the main study that involved 882 residents and 72 care homes where half of them received the pharmacist service and half did not, to find out whether the pharmacist service reduced falls (a common side effect of medication). The service presented no safety concerns. The pharmacists switched and stopped the medication, of which one quarter should have reduced the chances of falls. The service was generally liked. However, there is no evidence to suggest that the service reduced the number of falls or that it represented good value for NHS money. Our public and patient involvement members have helped us at every stage of the process. They were a central part of our final reporting event. Scientific summary Background The predominant UK care home (CH) model for enhancing pharmaceutical care currently consists of a pharmacist visit at least yearly to conduct a medication review and support medication management processes. Evidence for improvements in clinical outcomes resulting from this is inconsistent and lacks consensus. Non-medical independent prescribing provided an opportunity for pharmacist independent prescribers (PIPs) to assume greater responsibility for improving pharmaceutical care for residents in CHs, whilst saving general practitioner (GP) time in prescribing activities. Objectives Identify and describe the components stakeholders would specify in a feasible and acceptable PIP service. Identify the most appropriate outcomes to measure the impact of the Care Home Independent Pharmacist Prescriber Study (CHIPPS) intervention through the development of a core outcome set (COS). Identify and develop the components of a training package, which would prepare PIPs for the role. Test and refine the service specification and proposed study processes to inform a definitive trial. Estimate the effectiveness, cost-effectiveness and safety of a pharmacist independent prescribers assuming responsibility for providing pharmaceutical care to residents in CHs. Methods The study sites were located in Northern Ireland (one site), Scotland (one site) and England (two sites). Ethical approval was received for each work package (WP). WP1 development of service specification Views of residents, relatives, CH staff, CH managers, GPs, and pharmacists with experience of living or working in, or with, CHs regarding how to develop the PIP service were obtained by focus groups and through interviews. Transcripts were iteratively analysed, using the Theoretical Domains Framework, to identify key components, initial codes and themes to inform our study design. WP2 outcomes identification We followed recognised methodology to develop a COS to identify and select potential outcomes for the CHIPPS trial. Phase I: We generated a long list of outcomes through the literature review and stakeholder involvement (semi-structured interviews and focus groups) constituted for WP1, WP2 and WP4. Phase II: We convened a consensus panel to rate the importance of the long list of outcomes and provide additional recommendations for new outcomes. The panel was then asked to re-rate those outcomes for which consensus was not achieved and any newly recommended outcomes. Consensus for inclusion was achieved when 70% of respondents rated an outcome as critical. WP3 health economics Costs were estimated in Great British pounds (£) at 2017/2018 financial year levels. PIPs maintained training and activity logs to enable their time to be costed. Details of healthcare use were obtained from GP/CH records. Unit costs were obtained from published resources. Quality of life was measured using EuroQol Five Dimensions and Five Levels Rating Scale (EQ-5D-5L), which was completed by proxy. Mean quality-adjusted life-year (QALY) scores were estimated for the 6-month follow-up. WP4 training package development A training package was developed through six phases overseen by an expert advisory panel (EAP). Systematic review and narrative synthesis The review was registered with PROSPERO (20150907) and reported according to PRISMA guidelines. Published articles were selected by two independent reviewers if they provided primary empirical data describing a pharmacist intervention in the care home environment, including information regarding education and training. Systematic reviews and abstracts were excluded. Data were extracted to enable description of knowledge requirements for the role. Initial stakeholder engagement Within the interviews and focus groups outlined in WP1, participants were asked for views with respect to education and training requirements for PIPs. Training-specific interviews and focus groups Focus groups were convened with primary care pharmacists, GPs, community pharmacists and care home staff to obtain their views on the first draft of the training package derived by the EAP from Phases I and II. Stakeholder engagement and consensus Panels were held at each site to obtain consensus on the proposed training package. Feasibility testing A focus group was convened with PIPs to obtain feedback on the training package following feasibility testing. Evaluation All intervention PIPs received the training package during the definitive trial (WP6). Evaluation consisted of training day evaluation forms (PIP, n = 21), online survey (PIP n = 16) and semi-structured interviews (PIP, n = 14). The dataset was analysed separately and then triangulated. WP5 feasibility study Four triads (one in each location), consisting of a GP, PIP and care home (providing care to older persons) with 10 residents (65 years and older, prescribed at least one medicine), were recruited. PIPs were trained and provided the intervention for 3 months. Data on recruitment and retention were collated. Outcome measures identified in WP2 were tested for suitability for inclusion within the main trial. A safety assessment process was developed involving an independent review of PIP pharmaceutical care plans (PCPs) and adverse events, i.e. hospitalisations and deaths, plus enablement of independent reporting of concerns through a specific email address. Post-intervention face-to-face semi-structured interviews were held with participants and a focus group was held with the PIPs. Proceedings were digitally recorded and transcribed verbatim. WP6 definitive trial with internal pilot This was a cluster randomised controlled trial (cRCT) involving the training of PIPs to collaborate with the care home residents’ GP, care home staff and residents, to assume responsibility for medication management over 6 months. Recruitment General practices, CHs and PIPs were recruited in triads according to the following inclusion and exclusion criteria: Inclusion criteria Pharmacists had to be an independent prescriber (PIP). General practices had to be caring for sufficient care home residents to enable recruitment of 20 eligible participants. Care homes had to be caring for adults aged ≥65 years and associated with a participating general practice. Care home residents had to be under the care of a participating general practice, ≥65 years old and prescribed at least one regular medicine. Exclusion criteria Care homes in regular (monthly or more) receipt of a medication-focused review service or under formal investigation by a regulator. Care home residents receiving end-of-life care or participating in another study. Identification Pharmacist independent prescribers were identified using local networks. Care home(s) were approached by participating GPs. General practices identified and screened residents against the inclusion and exclusion criteria. Eligible residents were provided with invitation packs by care home managers to obtain verbal permission for a researcher to approach for consent. For residents without capacity, consent packs were posted to the resident’s next of kin. Randomisation and blinding Randomisation was at a general practice (triad) level and was stratified by the four geographical areas, using a web-based electronic system. Research associates were blinded to allocation during recruitment. PIPs, CHs and GPs were unblinded. Training Intervention PIPs undertook the training package, revised after WP5, over 6 weeks. Time-zero for the intervention was 6 weeks post-randomisation. Intervention The PIP intervention, 4 hours allocated per week per 20 residents, included developing and implementing PCPs assuming prescribing responsibilities supporting ordering, prescribing and administration processes medication management training of care home staff liaison between all parties to optimise medication-related activities. Control Pharmacist independent prescribers allocated to the control arm did not participate. Care homes and residents received usual care. Sample size Overall, 880 participants (440 in each arm) were required to detect a 21% decrease in fall rate from 1.50 per resident over 6 months with 80% power, at the 5% significance level and with an intra-class correlation coefficient of ≤0.05. We assumed a 20% loss to follow-up. Primary outcome measure Fall rate per person from care home falls record over 6 months. Secondary outcome measures Quality of life (EQ-5D-5L) Barthel Index Drug Burden Index (DBI) Hospital admissions GP visits Mortality Statistical methods Intention-to-treat analysis was performed with between-arm comparison of falls using a negative binomial model. Analyses were conducted using the generalised estimation equation approach adjusted for the clustered design. The final model included baseline fall rate, prognostic variables and arm as fixed factors. Safety reporting of serious adverse events Serious adverse events (SAEs), defined as hospitalisation or death, identified through routine monitoring, were assessed for association with PIP activity by residents’ GPs. An independent email address was created to enable concerns regarding the intervention to be confidentially reported. A random 20% sample of PCPs and associated resident documents were assessed by study geriatricians during the intervention period for clinical appropriateness and safety. Process evaluation Quantitative (surveys of care home staff, GPs and PIPs, PIP activity logs, PCP review and trial outcomes) and qualitative (interviews with care home staff, residents, GPs and PIPs) data were collected. Results WP1 development of service specification Twenty-seven pharmacists (care home and community), 24 GPs, 6 care home managers, 9 care home staff and 14 residents and relatives provided input. The concept was broadly welcomed by all. Potential barriers were identified as pharmacists’ knowledge of chronic disease management and older people’s medication, the care home environment, requirement to provide clarity with respect to the role and the need to become integrated both socially and professionally within CHs and general practice teams. Participants agreed that the PIP should assume responsibility and provide support for all elements of medication management within the home. WP2 outcomes identification Sixty-three outcomes were identified in Phase I (22 from the literature and 41 from stakeholders). Twelve outcomes met consensus criteria for inclusion in round 1, with 17 achieving no consensus and the remainder achieving consensus for exclusion. Two outcomes were further included after round 2, yielding a final list of 14 potential outcomes. WP3 health economics The estimated mean per resident cost of the PIP intervention was £323 (including training). The mean (per resident) incremental cost of the PIP intervention, after considering other NHS costs and personal social services (PSS) and adjusting for any differences between arms, was £280. The mean incremental effect was estimated to be −0.004 QALYs. As the PIP intervention was estimated to be associated with higher costs and no improvement in quality of life, it was not estimated to be cost-effective. WP4 training package development Literature provided therapeutic and clinical knowledge requirements. Qualitative work added the importance of understanding local cultures and requirement to integrate into teams. Recognising that PIPs started from different baselines, the final training package consisted of face-to-face training for 2 days underpinning knowledge pack personal development framework personal development planning and implementation with mentor time to integrate into teams and understand local cultures oral competency assessment by a mentor and an independent medical assessor. Pharmacist independent prescribers reported that all elements were useful and appropriate, enhancing their confidence for the role. Additional training on how to build effective relationships was recommended. WP5 feasibility study Four triads, each with 10 residents, were successfully recruited; 30% of the residents in CHs were found unsuitable following screening and 30% of those invited declined to participate or failed to reply. Two outcome measures [Mini-Mental State Examination (MMSE) and QUALIDEM] were removed following testing. Differences in outcome measures pre- and post-intervention suggested that the intervention had the potential to enhance care. No adverse events related to PIP activity occurred, and no major concerns were identified from reviewing PCPs. Qualitative feedback confirmed acceptability of the intervention and identified potential for the intervention to improve patient care and safety and to save GP and care home staff time and effort. WP6 definitive trial with internal pilot Recruitment Forty-nine triads (49 GPs and PIPs and 72 CHs) were randomly allocated to the intervention (n = 25) and control (n = 24) arms. Additionally, 454 residents were recruited to the intervention arm and 428 to the control arm. Baseline characteristics Baseline characteristics were largely similar between arms. The control arm had a greater proportion of CHs with nursing residents (59% vs. 42%) and lower performance with respect to activities in daily living. Residents in the intervention arm had a higher mean number of falls in the 3 months before service implementation (0.78 for intervention vs. 0.57 for control). Primary outcome A greater number of falls were recorded in the intervention arm in the 6-month follow-up. Once adjusted for differences at baseline, the result was non-significantly in favour of the intervention [rate ratio (RR): 0.91, 95% confidence interval (CI) 0.66 to 1.26; p = 0.99]. Secondary outcome measures The intervention reduced residents’ DBI by 25% compared with that by 15% in the control arm (RR 0.83, 95% CI 0.75 to 0.92; p<0.001). It had no effects on mortality, hospitalisation, activities of daily living or quality of life in the intervention and control arms. Safety No SAEs were related to PIP activity. Independent review of care plans revealed no safety concerns, and no concerns regarding PIP activity were received through email. Process evaluation Pharmacist independent prescribers largely adhered to service specification, varying provision according to need. Two-thirds of PIP time was spent on resident-related activities, of which approximately 24% was face-to-face. They spent 24% of their time on other general activities in the care home and 10% of their time travelling. Five PIPs stated that 4 hours per week was not enough, eight PIPs found it sufficient and three PIPs found it too much. Overall, 668 interventions were recorded, of which 566 were clinical, including 379 for medication discontinuation or dose reduction, 86 for medication initiation or dose increase, 49 for medication change and 52 for monitoring. Among the 566 interventions, 189 (33.5%) involved medication for treating diseases of the central nervous system. Among the 189 interventions, 148 reduced the likelihood of falls, 37 increased it and 4 were unclassifiable. Moreover, 179 interventions reduced drug burden and 10 increased it. The service was valued by most of the stakeholders. This was believed to be more effective where there was good communication, a readily accessible PIP, confidence on PIP competence, stable care home management and resonance between PIP activities and GP and care home needs. Conclusions Service specification, COS, training package and study design were delivered for use within a definitive trial designed to evaluate this model of care. The trial recruited and retained to target. Three PIPs (12%) failed to deliver any part of the intervention. The intervention was well received and believed to enhance resident care and safety. Several medicines were discontinued or stopped, resulting in a significantly reduced drug burden. However, no difference in falls, the primary outcome measure, was identified. The PIP intervention was also not estimated to be cost-effective, as it was associated with an increase in costs and no improvement in QALYs. Given the evidence of suboptimal prescribing in CHs, further work is recommended to develop interventions that improve resident clinical outcomes and are likely to be cost-effective. Trial registration This trial is registered as ISRCTN10663852, definitive trial: ISRCTN17847169. Study registration This study is registered as PROSPERO CRD20150907. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-0613-20007) and is published in full in Programme Grants for Applied Research; Vol. 11, No. 10. See the NIHR Funding and Awards website for further award information.
Keywords
