Overview of study objectives
CHEMIST is the first study, to our knowledge, to evaluate the feasibility of conducting research into psychological approaches for depression prevention in a CP setting. The aim of the feasibility study was to adapt an existing depression prevention intervention38 for implementation by CP staff for people with comorbid subthreshold depression and LTCs, and to develop and refine study procedures for use in this setting. The pilot RCT sought to quantify and evaluate important study parameters, including recruitment and retention rates, assessment and data collection methods, and engagement with the ESI. Process evaluations during both the feasibility and the pilot RCT phases aimed to determine the acceptability and appropriateness of the ESI and study procedures, and to explore the barriers to and facilitators of implementation.
Summary of main findings
Feasibility study
Eight CPs participated in the feasibility study and 17 pharmacy staff completed the ESI facilitator training. The recruitment target (20–30 participants) was achieved, with 24 participants recruited across seven of the eight CPs and one general practice over a 9-month period. Initial face-to-face pharmacy recruitment was slower than expected, with pharmacy staff reporting difficulties in finding the time to raise the study with pharmacy customers owing to the busy nature of the pharmacy setting. To achieve the recruitment target, a number of additional recruitment methods were implemented, the recruitment period was extended and additional pharmacies were recruited (see Chapter 5).
Despite these various recruitment methods, the number of study information packs distributed was relatively small (n = 1082), with recruitment activity varying across CPs. Response rates to the study information were lower than anticipated (7.3% via pharmacy methods; 3.5% via general practice database searches) and are lower than those typically seen in primary care studies (18%).38 The highest response rates were for face-to-face recruitment in the pharmacy, although these rates are based on small numbers. Conversion rates (from return of consent form to recruited participant) were similar across pharmacy (34.4%) and general practice recruitment (28.6%), and are comparable to those reported in primary care studies (28%).38
Data completion across all outcome measures was excellent (95–100% of returned questionnaires), indicating that participants found it feasible to complete these. The retention rate at the 4-month follow-up was good (83%) and exceeded the target of 80%; one participant withdrew from the study.
Nine of the 17 trained ESI facilitators delivered the ESI. Engagement with the ESI was good, with 17 out of the 24 recruited participants (71%) commencing the ESI sessions. This level of engagement is comparable to engagement with psychological intervention services in routine care, in which 68% of people commence treatment, although is slightly less than that reported in primary care depression research studies (84%).38,86 For those participants who commenced the ESI, 86% of all possible sessions were completed, with all participants completing at least two sessions, and 10 participants completing all six sessions. This compares favourably with routine psychological intervention services, in which 56% of people who enter the service attend two or more clinical appointments.86
Depressive symptoms (as measured by the PHQ-963) appeared to reduce slightly from baseline to follow-up (mean reduction of 3.5 points), which reflected findings of the nested qualitative study, which found that participants reported positive benefits from the ESI. However, as the sample size is small and randomisation was not employed, this finding must be viewed with caution. Health economic analysis indicated that GP visits were the most common type of health service resource accessed, with only a small proportion of these visits being for low mood.
Data generated from semistructured interviews suggested that study participants felt that the CP was an appropriate place to offer a mental health intervention; that the ESI made sense and was acceptable; and that the ESI supported self-management of mood and LTCs. Study participants reported that they had engaged with the self-help materials and valued the contact with the ESI facilitator. Analysis of data generated by semistructured interviews with ESI facilitators suggested that ESI facilitators found the ESI to be acceptable. Despite initial concerns about delivering the ESI, ESI facilitators reported increased confidence with ESI delivery. The ESI training was reported as acceptable and useful, and suggestions were made to modify this training for the pilot RCT. ESI facilitators and pharmacy staff reported concerns about increased work burden and the practical challenges of study delivery in the CP, which were especially problematic at busy times. Fidelity to the ESI was not assessed as originally planned owing to concerns over the additional burden and anxiety this may cause ESI facilitators.
Learning from the feasibility study and progression to the external pilot randomised controlled trial
The feasibility study generated important learning that led to a number of refinements to the recruitment and study processes during the delivery of the study, and in preparation for the pilot RCT. Analysis of qualitative data informed further refinements to the ESI and the associated training materials (including the ESI training workshop). The participant study materials (including recruitment materials and study questionnaires) were also modified in an attempt to improve recruitment to the pilot RCT. The feasibility study highlighted the complexities of embedding mental health research in the new setting of the CP and the need to adapt study processes.
The findings from the feasibility study demonstrated that the ESI could be delivered in the CP setting and that pharmacy staff could be trained by the research team to deliver the ESI. Although not all of the feasibility study progression criteria were met, the TSC and the funder approved progression to the pilot RCT.
Pilot randomised controlled trial
Fifteen CPs participated in the pilot RCT, including six CPs that also took part in the feasibility study. An additional 18 pharmacy staff completed the ESI facilitator training. Recruitment remained a challenge and, despite greater engagement with recruitment activities in CPs and via general practice database searches, the recruitment period had to be extended by 6 months.
Forty-four participants (out of the target of 100 participants) were recruited and randomised (24 to the ESI and 20 to usual care) from across 12 of the 15 CPs and from five general practices. As in the feasibility study, recruitment activity varied across pharmacies. The response rate to pharmacy-based recruitment (2.7%) was lower than in the feasibility study (7.3%), although in the case of general practice recruitment response rates were similar (2.9% in the pilot RCT, 3.5% in the feasibility study). In addition, the conversion rate (from return of consent form to randomised participant) was lower in the pilot RCT (20.8% for pharmacy recruitment, 23.9% for general practice recruitment) than in the feasibility study; however, these conversion rates are still comparable to those reported in other primary care studies with similar patient populations (28%).38 By contrast to the feasibility study, the different pharmacy-based recruitment methods yielded similar response rates (although numbers were again small).
As with the feasibility study, data completion of all outcome measures was excellent at baseline and follow-up (100% completion of returned questionnaires), confirming the suitability of these measures for use in this study. Retention at the 4-month follow-up was again excellent (93%), providing further evidence that it is feasible to retain and follow up this patient population. Two participants withdrew from the study.
Eleven ESI facilitators delivered the ESI (including five ESI facilitators who delivered the ESI in the feasibility study). Engagement with the ESI remained positive, with similar levels of engagement to that observed in the feasibility study. Of the 24 participants randomised to receive the ESI, 18 commenced the sessions (75%); this rate is above that seen in standard Improving Access to Psychological Therapy services in the NHS (68%)86 and slightly lower than that observed in similar research studies (84%).38 For those 18 participants who commenced the ESI, 73% of all possible sessions were completed, with 16 participants (90%) completing two or more sessions and nine completing all six sessions. This level of treatment engagement is above that found in routine secondary care and is comparable to that reported in similar research studies.38
There was a slight decrease in depressive symptoms at the 4-month follow-up in both the ESI and the usual-care arms, with a slightly larger decrease observed in the usual-care arm (mean reduction of 2.1 points and 1.1 points on the PHQ-9, respectively); however, the sample size (n = 41 at follow-up) is too small to draw any conclusions and the findings should be interpreted with caution. It is of note that the size of the decrease in depressive symptoms was less than that observed in the feasibility study, although baseline PHQ-9 scores were lower across both treatment arms in the pilot RCT than in the feasibility study.
Data generated from semistructured interviews with study participants, pharmacy staff (including ESI facilitators) and GPs were analysed using NPT.45 The findings demonstrated that study participants found that the intervention made sense and was acceptable, that the support provided was perceived to be useful and that the self-help workbook was accessible and understandable, supporting and extending the qualitative findings from the feasibility study. Most participants who engaged with the intervention reported positive changes in mood and behaviour and viewed the relationship with the ESI facilitator as central to the intervention. Familiarity with pharmacy staff was viewed by some participants as a positive reason to participate in the study, whereas others felt that this may prohibit involvement because of concerns around privacy and confidentiality.
Pharmacy staff and GPs generally provided positive feedback about the intervention. ESI facilitators felt that the ESI training had extended their skill set and those who delivered the ESI reported gaining positive feedback from observing beneficial changes in the participants. Although some pharmacy staff found the study processes feasible to embed within their current pharmacy practice, others reported that the cumulative effect of numerous recruitment methods, and issues with limited availability of facilities for private consultations, impeded the working practices of the pharmacy. Delivering a mental health intervention in the CP was viewed as novel by some pharmacy staff, requiring a change in perception (for pharmacy staff and customers) about the role of pharmacy.
Considerable attempts were made in the pilot RCT to record ESI sessions to assess fidelity to the ESI. This was a new process for the ESI facilitators and the levels of anxiety reported were underestimated. Although some ESI facilitators agreed to record ESI sessions, this was achieved for only one session.
Economic analysis demonstrated high completion rates of resource (health service) use and quality-of-life questionnaires, with low levels of missing items. Visits to the GP, nurse and pharmacy were the most common types of resource use accessed. A small number of participants reported inpatient admissions (not related to mental health) during the study; however, owing to high unit cost, these inpatient attendances may have a significant impact on incremental cost and should be carefully considered in any future economic analysis. An imbalance in baseline quality-of-life scores between the two treatment arms was observed. There was some indication that the ESI might reduce the use of primary care and community services and hence costs, although the sample size was insufficient to draw any statistical conclusions. The ESI cost per study participant was £51.40, including the cost of delivering ESI sessions, administrative work and clinical supervision (based on four ESI sessions).
The pilot RCT incorporated progression criteria for a definitive RCT (although this was not part of the funded CHEMIST study) (see Chapter 1). The pilot RCT did not meet the recruitment target of 100 participants; 67% of participants randomised to the ESI completed at least two sessions (lower than the target of 80%, although it is worth noting that 90% of those participants who did commence the ESI completed a minimum of two sessions) and fidelity to the ESI was difficult to assess. However, retention at the 4-month follow-up was 93% (exceeding the target of 80%) and data completeness was excellent (100%).
Discussion of main findings across feasibility study and pilot randomised controlled trial
CHEMIST represented a new experience for the majority of the CPs taking part, both in terms of undertaking research and in terms of undertaking training in and delivering a mental health depression prevention intervention. Recruitment was a challenge; pharmacy staff found it difficult to reliably embed the study recruitment processes into their existing pharmacy routines for a variety of reasons. It is difficult to know whether or not recruitment to similar research studies will improve as CPs become more research active and accustomed to study and recruitment processes. Increased experience of research would facilitate the development of more streamlined processes to better embed research into existing pharmacy workloads and practices.
Response rates were lower than those reported for similar research studies conducted with similar patient populations within primary care. This may be related to the relative novelty of offering and delivering a mental health intervention via the CP, as identified in our process evaluations. The feasibility study demonstrated that the response rate was higher for face-to-face recruitment than for other (non-face-to-face) pharmacy-based methods; this is despite pharmacy staff reporting that this recruitment method was challenging to implement. This perhaps highlights the beneficial effects of existing relationships between pharmacy staff and customers for engagement with the research. The low response rates observed may also indicate that recruitment materials require modification to ensure that these are suitable for this patient population and/or this research setting. It is of note that many of the CPs involved in CHEMIST were located in socially deprived regions. A large proportion of those people who received study information chose not to indicate their interest in the study and the reasons for this are unknown.
Delivery of a mental health intervention in the CP is at this time relatively uncommon. Findings from the process evaluation suggest that preparatory work may be needed to provide context (or to ‘set the scene’) for such mental health support in this setting. This may to some degree explain the lower response rates to study information observed in this study than in primary care, in which such mental health interventions are more routine.
Our qualitative work highlighted that there may be a perception that the role of the CP does not include delivery of a talking treatment for a mental health problem. Pharmacies sit in the communities that they serve and concerns raised as to the confidentiality of information disclosed within this setting suggest that this may not be comparable to the perception of confidentiality in other settings, such as primary care. Some participants who were interviewed indicated a preference to speak to their GP about mental health concerns. Such concerns may have impacted on people’s willingness to take part in the study at the outset or on their willingness to engage with the ESI following recruitment to the study.
There were some indications from our process evaluation that intervention delivery posed some difficulties in the routine activities within the pharmacy. The consulting room is a key and limited resource in most CPs, and was occupied more often owing to the CHEMIST intervention. Further consideration is needed of intervention structures that are suited to the CP setting. Although we found that the CHEMIST intervention (which was based on BA) was acceptable and that we could train pharmacy staff to deliver this, the structure of the intervention placed additional burden on other pharmacy staff and the ability to meet the daily demands posed by a busy CP. If a new intervention impacts on the usual work of the CP, it is likely to encounter barriers to roll out. Our process evaluation indicates that further consideration of this factor is required to support interventions, such as the intervention that we implemented.
The issue of risk raised anxiety for CP staff in relation to both recruitment and intervention delivery activities. These anxieties appeared to ease across time but highlight the need for attention in study set up.
Based on these observations, further preparatory work may be required to help better embed research and related intervention activities, such as those used in CHEMIST, into the CP setting.
An additional factor that may have contributed to the low recruitment activity by the CPs relates to funding cuts that were experienced around the time of CHEMIST. In December 2015, the Department of Health and Social Care revealed that it was planning to reduce CP funding for delivering the contractual framework (essential and advanced services). This had a substantial impact on CPs, with reports that up to one in four pharmacies may face closure causing ‘shockwaves’ through the sector.87,88 It was within this context that CHEMIST was conducted. It is perhaps not surprising that CPs participating in CHEMIST found it challenging to deliver on the additional activity that the study necessitated in an already overstretched workforce. If these funding restrictions were reversed and the funding climate were to improve, it may be that the additional activities required by participation in research may become more feasible for CPs.
The depression symptom severity of study participants is worthy of note. A structured diagnostic interview (MINI)48 was administered by the study team at screening to confirm diagnosis of subthreshold depression. Despite this, participant PHQ-9 scores at baseline in both the feasibility study and the pilot RCT indicated the presence of moderate and severe depression. Although the overall mean baseline PHQ-9 score was lower in the pilot RCT than in the feasibility study (mean PHQ-9 scores of 11.0 and 13.5, respectively), this score is still higher than what would be expected, with 57% (25/44) of participants having a PHQ-9 score indicative of moderate or more severe depression. This degree of depressive symptoms is somewhat higher than reported in research studies with similar patient populations.38
The small sample size in the pilot RCT indicates uncertainty as to whether or not all potential confounders were distributed equally across the ESI and usual-care arms; however, the two arms appeared equal at baseline in terms of key characteristics. Differences observed between the two treatment arms in the primary outcome (PHQ-9) at the 4-month follow-up were minimal and less than seen in similar research studies.38,76 Even if the sample size of 100 participants had been achieved in the pilot RCT, the study was not powered to test for intervention effects and, for this reason, the findings should be viewed with caution.
Promising levels of engagement with the ESI were demonstrated across both the feasibility study and the pilot RCT. Process evaluations indicated that the ESI was acceptable to both study participants and those ESI facilitators delivering the ESI; both study participants and ESI facilitators reported positive benefits. Across both phases of the study, data (questionnaire) completeness and participant follow-up rates were good. These findings suggest that it would be possible to collect the relevant information to conduct analysis in a definitive RCT of the ESI delivered within the CP setting, provided that the issues discussed above regarding recruitment and intervention implementation could be resolved.
Strengths and limitations
CHEMIST explores the feasibility of a psychological intervention (BA) for people with comorbid subthreshold depression and LTCs, and is delivered in the CP setting.
It demonstrated that CPs are interested in mental health research and are keen to participate, despite the challenges they face with funding cuts and an often understaffed workforce. A key strength of the study is the finding that CP staff can be successfully trained to deliver a psychological intervention (BA). This builds on previous work exploring the capacity for this simple treatment to be delivered by non-mental health specialists and has potential implications for the roll out of psychological interventions, such as BA, by non-specialist groups.38–40
Qualitative work permitted an in-depth exploration of a range of aspects important to CHEMIST. These data demonstrated that the ESI was often found to be acceptable to both those delivering and those receiving the intervention, although with some important implications for the routine activity in the CP setting.
People with comorbid subthreshold depression and LTCs were willing to engage in CHEMIST. They were able to complete study questionnaires and were willing to be followed up. Those participants who were offered the ESI generally engaged well with this, despite the finding that some concerns relating to confidentiality existed.
A further strength is the learning that has been gained from conducting this study. This was critical throughout the duration of the study and allowed procedures and materials to be developed and refined for use in a CP setting. This learning will also be crucial in the design and delivery of future research studies within this setting.
The study benefited from the input of key stakeholders (including pharmacy and public health stakeholders) and PPI from the inception of the study and throughout its management. This allowed issues to be discussed as these arose. Involvement of the PPI AG was instrumental in the refinement of the ESI and study materials, and ensured that the study remained focused on supporting people with mental and physical health-care needs, and supporting CP staff to engage effectively with the study.
We adopted usual care as our comparator. This was seen to be suitable in this study and reflects what is usually provided to people with LTCs and subthreshold depression. There is ongoing debate about the use of attention placebos in psychological therapy research and no gold standard is available.89 The approach taken in CHEMIST is consistent with other large-scale research in similar areas38,76 and was considered to be appropriate.
The main limitation is the small sample size in the pilot RCT. Although no formal statistical comparisons were planned, the smaller than expected sample size precludes the drawing of any conclusion from the comparison of primary outcome data. It also limits any reliable sample size estimates for a definitive RCT.
Although a range of interviews were conducted, the small sample size of some participant groups may have limited the range of views gathered. The views of those people who decided not to participate in the study are unknown. A further limitation is the predominantly white patient population in both phases of the study.
Owing to the problems with the delivery of the study as described above, not all those ESI facilitators trained to deliver the ESI had the opportunity to support participants through the intervention; ESI facilitators may have found this frustrating and this may have affected their engagement with the study. However, it is worth noting that, of the 35 pharmacy staff who attended the ESI facilitator training, only 21 were available to deliver the ESI during the intervention periods (reasons for unavailability included ESI facilitators leaving their pharmacy role, concerns over existing workloads and personal circumstances). The problems experienced with recording ESI sessions prohibited an independent assessment of fidelity to the ESI. This posed a further limitation in assessing the feasibility of the intervention for future research.
Recruitment via general practice database searches was limited across both the feasibility study and the pilot RCT owing to research-active general practices not being close or linked to those CPs participating in the study. Recruitment may have been improved if a larger number of general practices within the vicinity of the recruiting CPs were research active.
Conclusions
The CP is a relatively new setting for mental health research and often constitutes a novel setting in which to conduct depression prevention studies, such as CHEMIST.
The findings from the feasibility study and the pilot RCT suggest that there would be challenges to conducting a large-scale definitive RCT of a structured psychological intervention for people with comorbid subthreshold depression and LTCs in the CP. Although key aspects of the study were promising (e.g. suitability and completion of outcome measures, retention rates and engagement with the ESI), a significant issue was recruitment to the study.
Importantly, CHEMIST has provided important reflections on how to effectively embed research and study processes in this busy public health setting, and the barriers that may be encountered. The learning acquired from conducting the feasibility study was critical in refining study procedures, training and materials for use in the pilot RCT. Such findings will also be useful in the design and delivery of future studies within a CP setting.
The CP would appear to be ideally placed to provide support to people with mental–physical multimorbidity. Prevention of depression in people with long-term health problems is important because they are at increased risk. The Community Pharmacy Contractual Framework for 2019/20 to 2023/24: Supporting Delivery for the NHS Long Term Plan24 encourages CPs to play a greater role in clinical service delivery and in helping people to stay well. The CHEMIST intervention would seem well placed to reflect this new focus for the CP, which has prevention acknowledged as a core role. The limitations highlighted in this study suggest that more research is required to consider effective methods to recruit people to studies and successfully embed interventions within the CP setting. The evidence base for such activities, especially in relation to depression, is lacking43 and further definitive research is required. Targeted approaches, such as the CHEMIST intervention, require more research.
Clinical Research Networks are keen to expand the delivery of research in this important public health setting and the establishment of primary care networks, with the aim of fostering wider relationships with their partners to deliver integrated care, which will further facilitate important collaborative and interdisciplinary working. The CP has been identified as one such key partner, especially in relation to public health and prevention. It may be that embedding CP research within a primary care network may offer an effective platform to better implement research within this setting and help explore ways to improve and facilitate recruitment.
Implications for health care
Mental–physical multimorbidity poses a significant challenge for the future. The identification of a suitable depression prevention intervention that can be offered and delivered in the CP setting could offer an important option to communities, particularly in areas of socioeconomic deprivation. The ability to train CP staff to deliver the CHEMIST intervention does indicate that further work may be warranted to make a robust evaluation of clinical effectiveness and cost-effectiveness possible.
Recommendations for future research
Further research is required to inform the design and delivery of a definitive RCT to test the clinical effectiveness and cost-effectiveness of depression prevention interventions within the CP setting.
Such research should seek to explore and address barriers to recruitment in this setting. This could include further exploration to understand the issues of confidentiality and disclosure of mental health concerns to CP staff, and to gain further knowledge about people’s perceptions of the CP as a setting to deliver a psychological intervention for people with comorbid mental and physical health conditions. Future research may also benefit from consideration of joined up working across CP and primary care with respect to mental health support, which may promote engagement in mental health studies conducted within the CP setting.
The factors promoting effective implementation of a depression prevention intervention in the CP setting need more detailed investigation. Consideration of intervention design and the impact of intervention delivery on the usual activities within the CP is recommended to help overcome some of the barriers to recruitment and study/intervention implementation described in this report. Such research will facilitate a better understanding of the organisational, operational and cultural factors to consider when designing research for implementation within this setting.
Future work should also seek to involve a more ethnically diverse sample. Further work to review the numbers of people with subthreshold depression who can be identified and who would be willing to engage in a mental health intervention delivered within the CP setting may be needed.
Consideration may need to be given to developing and testing a more pragmatic approach to study inclusion criteria and assessment that uses a depression scale rather than a diagnostic interview. Such processes may improve recruitment rates and reflect what may be adopted in the future in this type of setting, although the methodological limitations of such approaches would need to be considered.
Future research may also benefit from preparatory cross-disciplinary learning and training between research and CP environments to better facilitate more effective embedding of research and intervention processes and structures within the CP setting.