Qualitative substudy

Aim and objective

The purpose was to explore the barriers and facilitators to recruitment and retention of participants in the PAX-BD trial. Initially, the qualitative component was intended to collect feedback specifically for the pilot phase of the trial. However, due to early trial closure, the applications of the qualitative substudy were broadened to encompass more general insights that could be applied to similar studies in future.

Method

Semistructured interviews were conducted by central researchers via telephone. Topic guides were used to cover areas of interest, as well as allowing openings for emergent topics (see Appendix 5, Qualitative study topic guides for interviews). Interviews were audio-recorded and outsourced to UK Transcription for transcribing. HCP participants verbally consented to interview and patient participants optionally consented at the pre-randomisation and randomisation consent visits.

Sample

A total of 26 participants were interviewed, comprising 15 HCPs (4 PIs and PI delegates, 11 non-clinical research staff including CSOs, nurses and RAs) and 11 randomised patient participants (1 withdrawn prior to week 12 and 10 retained until at least week 12). It was not possible to interview patient participants who were withdrawn in pre-randomisation as some had declined consent or disengaged with trial activities after withdrawal; similar issues were experienced for participants withdrawn prior to week 12.

Analysis

Barriers and facilitators were the key directing themes (Table 22) for interviews and analysis. More focused interpretation of the transcripts was guided by an a priori framework of subthemes (see Appendix 5, A priori framework for analysis of barriers and facilitators to recruitment and retention to PAX-BD trial) from a meta-analysis of studies in depression.83 Interview transcripts were independently coded using NVivo Release 1.6.1 (QSR International, Warrington, UK) by at least one clinical and one non-clinical researcher to generate codes with an expanded span of interpretation. Initial codes were reviewed by a wider collective of clinical and non-clinical researchers, as well as the PPI group. Codes were modified on an ongoing basis to reach consensus and inform subsequent transcript analysis. Once all transcripts had been coded, analysis shifted to the interpretation of indexed codes assisted by the Hughes-Morley et al. framework;83 subthemes were inductively revised, and emergent themes were added per recurring discussions. Homogeneity within the subthemes and heterogeneity across them was established and attributed to the three key themes: Barriers, Facilitators and Suggestions for Future Improvement. A summary framework of themes and codes is presented in Table 22.

Results

The results presented here are a summary of the findings from the qualitative research conducted. Fuller details of all sections, including verbatim quotations from participants, are included in Appendix 5.

Barriers to recruitment and retention

Expression of symptomatology

This subtheme considered the profile of TRBD, interplay from comorbid conditions and the clinical challenges these factors present in identifying eligible participants and engaging them.

Profile of bipolar depression

Healthcare professionals highlighted that BD has a complex profile with symptoms overlapping other disorders such as emotionally unstable personality disorder, and patient presentation can be compounded by lifestyle stressors. These nuances contribute to difficulties in identification of participants for the trial. Subsequent determination of treatment resistance poses further challenges in clinical practice. The exclusion criteria included comorbid substance misuse and impulse control issues. HCPs highlighted that such comorbidities were prevalent among the bipolar patient group, narrowing the pool of eligible participants.

Patient presentation at recruitment

Transcript analysis suggested an interplay between previous treatment failures and chronicity of BD, which appeared to yield a sense of desperation for active treatment. This made it difficult to accept the possibility of placebo. Participants were required to be in a current mood episode, contingent on the periodic nature of BD, and this could change while they were in the pre-randomisation stage. The impact of depressive symptoms and side effects of medication were recognised as barriers to patients’ capacity to process trial material. Similarly, poor motivation could also lead to difficulties completing trial activities and engagement with HCPs. Overall, recruiting patients from the TRBD cohort posed complex clinical challenges in considerations of eligibility and ability to safely participate in the PAX-BD trial.

Determining causality of adverse events

During the randomisation stage, it was difficult to ascertain the causality of AEs due to similarities between common side effects of pramipexole and psychotropic concomitant medications, as well as bipolar symptoms of hypomania, mania and impulse control.

Presenting the trial to participants

Impact of COVID-19 on services

COVID-19 posed significant challenges to the PAX-BD trial. Although many trial activities were conducted remotely, sites were required to obtain consent in-person while adhering to strict guidelines around face-to-face contact. After restrictions became more lenient, there was a lengthy adjustment period post pandemic. Overall, HCPs widely expressed the view that recruitment would have been more successful if not for the COVID-19 pandemic.

Research prioritisation

Some PIs reported use of personal time for conducting research suggesting a workplace ethos that research is not a routine activity integrated into clinical care. HCPs employed at these sites reported limited recruitment and research experience, citing lack of staff capacity, resources and networks with clinical services as barriers to research involvement. Reflecting on the trial, HCPs stated they would reconsider involvement in research, particularly later in their careers, due to mounting clinical pressures.

Trial promotion strategies

Non-medical staff across sites emphasised the importance of engaging clinicians to gain access to eligible participants. This was challenging due to the limited availability of PIs, which shifted the burden of marketing and recruitment to CSOs. They reported that clinicians were less receptive to promotion from non-clinical colleagues; therefore, peer-to-peer trial endorsement was critical in imbuing credibility, especially in relation to a Clinical Trial of Investigational Medicinal Product (CTIMP). Overall, HCPs described the approach to recruitment as proactive and were perplexed when these attempts did not produce more referrals.

Access to eligible patients

Logistical issues in the care pathway structure were widely noted by HCPs who believed that patients with BD were largely under primary care, despite the nature of treatment resistance necessitating management under a community mental health team. HCPs used patient databases and case notes for pre-screening but noted a lack of informative detail. CSOs additionally encountered gatekeeping of clinical caseloads. Ultimately, CSOs focused screening to the caseloads of PIs who were already involved in the trial and with whom they had an existing working relationship. The pool of prospective patients, and scope for professional development, was therefore limited.

The term ‘Treatment resistance’

One patient stated that learning about the ‘treatment-resistant’ label was surprising, indicating this may have been a somewhat confronting experience. Given the nihilistic and stigmatising overtones of the phrase patient-facing documents and clinic discussions might be better using the alternative ‘difficult to treat bipolar depression’, mirroring the use in major depressive disorder.110 Treatment failures prior to the trial appeared to contribute to scepticism about the efficacy of pramipexole. Conversely, one HCP believed patient cost–benefit appraisals did not swing in favour of participation for those who had not exhausted all treatment options.

Uncertainty about research

The profile of research was raised in public consciousness because of the COVID-19 pandemic. Some patients were cautious of partaking in PAX-BD due to the view that research is experimental with unknown drug characteristics and potentially harmful side effects. Some participants expressed general apprehension, particularly around the RCT design.

Relationship with the care team

A patient with a negative view of their care team declined the opportunity for participation. The prospect of closer monitoring by the research team highlighted frustrations around the lack of sufficient support from their clinical team. During the trial, some patient participants reported a lack of direct involvement from their local site teams. Overall, patient views of their local team influenced their general level of satisfaction, perception of support and ability to carry out trial activities with ease.

Trial material

Some participants highlighted the complexity and amount of trial material they received. They felt overwhelmed by it, particularly at the beginning of the trial. Some admitted that rather than refer to the information sheet, they relied on conversations with HCPs leading to an added burden on staff time.

Treatment profile

Clinician attitudes

Clinical Studies Officers noted that sceptical clinicians viewed research as detrimental and/or disruptive to patients’ mental health, thereby withholding access out of desire to protect the vulnerable patient. One PI expressed general anxiety about the experimental nature of pramipexole for the treatment of BD.

Medication changes in pre-randomisation

Prior to the amendment allowing antipsychotics use in the randomisation stage, there was a reluctance to withdraw antipsychotics due to the possibility of destabilising patients and triggering relapse of manic or psychotic symptoms. HCPs reported many patient refusals due to the requirement to withdraw antipsychotics. Patient views suggested that those who were required to make changes to existing medication were likely to express stronger concerns about placebo. The obligation to commence a mood stabiliser was challenging for patients who had experienced intolerable side effects in the past and/or felt they had been ineffective.

Placebo

The possibility of being randomised to placebo appeared to give many patients pause when considering participation in PAX-BD. The resistance to placebo appeared to be exacerbated by the length of the trial.

Pramipexole

Healthcare professionals expressed major concerns about impulsivity and mania as potential side effects of pramipexole in bipolar patients. For some clinicians, this was compounded by the view that they could not intervene per usual clinical care due to restrictions in the first 12 weeks of the trial. Hence, a risk-averse approach was employed with patients perceived to be more vulnerable but not definitively ineligible.

View of trial processes and procedures

Administrative duties

During the trial, HCPs reportedly found it difficult to keep abreast of the numerous protocol changes and maintain administrative files, particularly in line with pandemic guidelines. PIs struggled to complete research tasks within trial timelines, especially when accounting for their restricted availability.

Eligibility criteria

Although reflective amendments were made to the eligibility criteria, some HCPs stated this did not significantly increase recruitment. HCPs thought the eligibility criteria were limiting and did not accurately represent the normative bipolar patient. Criteria around comorbid conditions, manic and psychotic symptoms, and treatment resistance were repeatedly flagged as stringent and restrictive.

Medication processes

Healthcare professionals expressed concerns about the capability of patients to independently manage study medication, particularly if they were in remission from substance misuse and/or experiencing impulse control issues related to BD or as a side effect of pramipexole. In these cases, they appeared to regard the independent management of medication as a risk factor. For participants in the trial, challenges were oriented around managing the large supplies of trial medication. Challenges in the 4-week titration phase related to the complexity of dosing.

Safety monitoring

Remote trial monitoring in PAX-BD was a barrier for patients who were reportedly lacking computer literacy or access, fearful of technology and/or had anxieties around answering the telephone. Central RAs observed that if participants experienced issues with the digital platform on first use, frustration developed at an early stage which meant they were likely to be less tolerant of continuing faults. Patient participants reported that the frequency and volume of questionnaires were time consuming. Some participants felt that the questions were not representative of their experiences. Weekly calls were unappealing to some patients.

Healthcare professionals noted difficulties engaging participants, and the reliance on a remote design led to gaps in crucial safety monitoring which resulted in recommendations to withdraw participants if they did not re-engage.

Facilitators to recruitment and retention

Marketing of the trial

Promoting to clinicians

The central research team regularly contacted sites about PAX-BD, which served as ongoing reminders and helped HCPs maintain recruitment goals. CI involvement in promoting the trial was highlighted by HCPs who expressed the effectiveness of a credible expert outlining limitations in the treatment of BD and emphasising the subsequent salience of trial objectives. Locally, PI networks seemed to be especially important in raising the profile of the trial. Research-active trusts had dedicated teams, and HCPs at such sites seemed better able to promote the trial. HCPs at less established research sites mitigated lack of prioritisation by proactively using special interest time for the trial. An additional appeal of PAX-BD was the opportunity for professional development from upskilling research and clinical competencies.

Promoting to patients

Healthcare professionals found the ability to promote the trial in clinical appointments efficient and effective. Some PIs adopted the strategy of framing PAX-BD as a 12-week trial and found this approach was more palatable to patients compared to the somewhat daunting prospect of 48-week participation. Patients generally reported positive impressions of the trial design and that although the trial processes were vast and complex, the language in the documents was clear.

Trust

Trial design

One patient response indicated that the research focus on treatment for BD validated their experiences of chronicity and debilitation; HCPs also recognised the salience of this in a professional capacity. Pramipexole has an existing profile of product characteristics, side effects and previous research and is also prescribed off-licence for mood disorders.27 This evidence base was accessible to HCPs and patients; combined with trial information and discussions with HCPs, this may have perpetuated a perception of pramipexole as relatively low risk.

Central research team

During the trial, the central team organised regular teleconferences attended by multiple sites to discuss progress and provide updates. They were understanding of recruitment challenges and collaborated with sites to implement alternative strategies, rather than applying idealistic pressure. This approach was particularly beneficial to less-experienced sites. Reflective amendments to trial processes and the eligibility criteria in response to challenges at site demonstrated that HCP feedback was valued and tangibly implemented.

Relationship with care team

Positive and trusting patient relationships with secondary care teams was a key facilitator of recruitment and engagement throughout the trial.

Informed consent

A strong sense of autonomy was emphasised by patient participants, achieved by extended discussions with an emphasis on their right to exercise agency and withdraw consent at any time. Access to support systems outside the trial enabled some trial participants to join the trial and engage with trial activities.

Attitude to risks

Some participants de-emphasised potential risks of the trial, adopting the stance that consequences of participation would be either neutral or beneficial. Participants felt they had fostered acute awareness of their response to drugs and felt confident in their ability to quickly identify side effects to prevent escalation. In terms of antipsychotic and mood stabiliser adjustments, some indicated that they were more receptive to changes as the current medication was ineffective. Many patients also appeared to have an implicit belief in HCPs’ duty of care, diminishing excessive concerns about trialling an experimental drug or the severity of side effects.

Central research assistants

Constructive relationships and rapport were established with weekly calls during pre-randomisation and the first 12 weeks of randomisation. Demonstrations of reliable assistance and patient-centred care acquired participant trust.

Perceived benefits

Access to treatment

PAX-BD provided clinicians the opportunity to access a new medication option and medical expertise. HCPs anticipated that demonstrating treatment efficacy would strengthen the evidence base for pro-dopaminergic strategies. For participants, a strong component of hope was present.

Altruism

Many participants expressed altruistic motives for participation.

Safety monitoring and support

Compared to routine clinical practice, follow-up in PAX-BD was more robust, frequent and personalised, providing quick intervention in matters of patient safety. This was reassuring to participants and HCPs. Central RAs flagged any cause for concern to PIs indiscriminately, which inadvertently facilitated regular holistic clinical input.

Treatment response

A large component of participant retention and motivation for continuation in the trial centred on treatment response. HCPs similarly indicated that observing a decline in symptoms reinforced the possible benefits of the trial, potentially encouraging them to continue recruitment. Some participants ascribed improvements in facets of quality of life to trial participation.

Ability to conduct trial

Trial information

Many HCP participants cited the trial protocol and Clinician Manual as comprehensive base reference resources. Other trial resources such as training videos, newsletters, information sheets, participant diaries and the trial website were similarly highlighted as useful source material.

Reflective protocol amendments

The PAX-BD protocol was reflectively amended in response to feedback from sites; this resolved some eligibility constraints and expanded the pool of prospective participants, as well as mitigating clinician and patient concerns around medication changes.

Remote design

Prior to the COVID-19 pandemic, the PAX-BD trial utilised remote means of data collection and monitoring. Post-COVID-19 remote monitoring additionally increased the feasibility of the trial. Participants described the ease of online self-reported data compared to clinic visits.

Medication processes

Healthcare professionals indicated that regular updates from central RAs provided comprehensive oversight of trial participants, thus prescribing processes felt well informed. The use of the central CNTW pharmacy was noted to expand feasibility for sites, as this streamlined trial set-up and dispensing processes. Participants additionally highlighted the convenience of receiving trial medication directly to their homes, rather than travelling to collect them. The option for CSOs to collect unused medication and empty bottles provided an additional adherence check and assurance that participants were managing medication safely.

Flexibility in trial processes

The role of flexibility was identified as a pragmatic facilitator, for instance the ability to make dose adjustments and lenient guidelines for concomitant medication and psychotherapy post week 12 made the trial more acceptable to patients and HCPs.

Healthcare professional attributes and attitudes

Clinical experience with bipolar depressed patients and successful recruitment in previous studies appeared to expedite knowledge of effective strategies to identify, engage and recruit participants. HCPs with prior experience in research were also better able to grasp nuanced trial requirements to avoid deviations and manage administrative tasks effectively. HCPs widely reported the value of CSOs assisting screening. Working values of collaboration, proactivity and dedication to patients emerged as facilitators of participant engagement.