A qualitative exploration of burnout prevention and reduction strategies for general practice registrars

High burnout levels among postgraduate medical trainees have prompted research to prevent and reduce this syndrome.¹ However, evaluations of interventions targeting burnout have found weak evidence and small or negligible treatment effects.^2–4 Exploratory research in this area has also largely neglected the impact of group and setting differences. Specifically, medical trainees have typically not been considered separately from experienced doctors,^5,6 despite their distinct work demands and stressors necessitating specific job and person resources.^7,8

The role of context in trainees’ burnout experiences has additionally been overlooked, even among recent qualitative studies.^9,10 Contextual factors, such as specialty, are important to consider given that causes of burnout are predominantly organisational and/or systemic.^1,11,12 Likewise, burnout prevention and management strategies have typically presumed a hospital-based training model.^11,13 Despite similarities in burnout causes between hospital and general practice trainees,¹⁴ applying these strategies to the Australian general practice training context is problematic. For example, improving workplace supportiveness¹³ is difficult when the ‘workplace’ comprises hundreds of independent practices. Similarly, assuming that training organisations have frequent access to trainees⁷ does not hold for Australian general practice trainees.

The aim of the present study was to address these gaps by exploring perspectives and experiences of burnout intervention among stakeholders involved in Australian general practice training. The aim was to identify effective (and ineffective) strategies for burnout prevention and management among this cohort.

Methods

This article is part of a broader study involving semi-structured interviews or focus groups with registrars, supervisors, medical educators (MEs) and program coordinators (PCs) recruited from one regional training organisation (RTO).¹⁵ PCs are non-clinical staff who work with MEs to support, and provide case management for, registrars throughout their training. The sampling frame comprised all registrars in their second term or later (so they could reflect on their earlier and current experiences), all supervisors attending two mandatory RTO supervisor workshops in late 2019, and all MEs and PCs affiliated with the RTO. An invitation, including an information sheet, was emailed to eligible individuals. Participants were reassured that their responses would be kept confidential. Registrars were compensated for their time, and focus groups were held in RTO-paid working time. Overall, a saturation sampling strategy was used to maximise the likelihood of yielding a broadly representative dataset.¹⁶ Saturation was defined as new data sources not yielding new themes. However, given the nature of the study, recruitment was terminated when all individuals who had expressed interest had participated or were either no longer interested or not contactable. This was done so all interested individuals had the opportunity to express their views.

Questions focused on participants’ experiences and perceptions regarding burnout; a subset of questions specific to the present article is reported in Appendix 1. All interviews and focus groups were conducted by SP between August and November 2019. Where participants consented, interviews and focus groups were audio recorded to facilitate transcription, with only SP having access to these recordings. SP reviewed all transcripts to identify general themes, then selected a representative transcript to be separately analysed by himself and TE. Transcripts were iteratively analysed using grounded theory and template analysis underpinned by post-positivism to identify prevention and reduction strategies.^17,18 Grounded theory and post-positivism were chosen in accordance with the exploratory aims of the study to produce generalisable themes of strategies. Analysis was performed using NVivo 12 for Windows (QSR, International) by storing portions of each transcript in codes. NVivo was also used to keep an audit trail of the coding structure development. Following analysis of the initial transcript, both analysts met to fuse their coding structures. SP applied and extended this structure to the remaining transcripts, then reviewed the coding structure for consistency and parsimony, and applied this revised structure to all transcripts again. SP then selected a further two transcripts for TE to code to establish intercoder agreement. The intercoder agreement for the coding regarding the objectives of this study was excellent (98%).

Analysts’ backgrounds were in psychology and medical education research. Both analysts were affiliated with the RTO, providing contextual understanding but also introducing biases. Reflexivity was enhanced through coding comparisons and analysis review by DD, who is not affiliated with the RTO. Power in pre-existing professional relationships between the interviewer and participants favoured participants, reducing the likelihood of participants modifying their responses to please the interviewer.

Ethical approval for this study was granted by the University of Adelaide Human Research Ethics Committee (H-2019-072).

Results

Interviews were conducted with 14 registrars. Focus groups were conducted with 15 MEs, 13 supervisors and five PCs. The average duration of both formats was 56 minutes (range: 38–75 minutes). Most registrars (71.4%) and all PCs were female, whereas most MEs (60%) and supervisors were male (61.5%). Most clinicians practised in urban areas (57.1% registrars; 60% MEs; 61.5% supervisors). Registrars spanned different training terms (second: 35.7%; third: 28.6%; fourth: 35.7%). Thematic saturation was reached with this dataset.

The results are discussed in regard to burnout prevention and reduction strategies. Notably, participants reported fewer ineffective than effective strategies.

Burnout prevention

Strategies were divided between registrars, practices, RTOs and the medical system (Table 1). However, one theme transcended these groups: normalisation. All clinical groups reported that feeling overwhelmed and experiencing imposter syndrome were stigmatised but common. Accordingly, these experiences were hidden, exacerbating registrars’ self-doubt. To avoid these vicious cycles, and potentially burnout, participants advocated for normalising these feelings among registrars through peer and colleague interactions. One registrar noted, ‘once I started talking about [low confidence] with [general practitioner] friends I realised actually this is normal’. However, given the power imbalance and reluctance among registrars, most believed that practices and RTOs should bear the onus to initiate wellbeing-related conversations (eg holidays, workload, supervision access). This could be complemented by experienced general practitioners (GPs), including supervisors, who could role model wellbeing habits and skills. Indeed, one supervisor argued that being ‘good mentors and coaches [and] show[ing] them how we avoid burnout’ was part of a supervisors’ duty. Participants also recognised that normalisation needed to be balanced to avoid registrars ignoring their doubts and acting unsafely.
 

Registrar strategies fell under the broad theme of sustainability, a theme raised by all groups other than MEs. This theme involved a need to create a sustainable balance between the rewards and demands of one’s personal and professional lives, although it was considered a complex process – as suggested by multiple subthemes. A major subtheme was the need to identify and prioritise meaningful and replenishing activities to ensure registrars had a ‘satisfying and meaningful life outside of work’ [Registrar]. Examples included planned holidays, scheduling of non-work/training activities, establishing and maintaining relational supports (both personal and professional) and a healthy lifestyle (eg exercise, diet, effective and adaptive coping styles), as well as having their own GPs. Participants encouraged registrars to optimise their working conditions through workload diversification, clear work–life boundaries (particularly healthy working hours), enhancing their organisational skills and building buffers (eg scheduling regular time slots for administrative tasks). A prerequisite was insight – ‘being able to recognise that you are burning out’ [Registrar] – and the need to prepare a contingency plan to use if they detected symptoms.

Key practice strategies, raised by most groups, fell under the umbrella theme of building a supportive environment, comprising four subthemes. The first related to supporting registrars’ psychological wellbeing. Participants emphasised ‘giving [registrars] the opportunity to debrief’ [Supervisor] and regularly checking in with their registrar(s) to identify challenges and monitor their wellbeing. These tasks could be conducted by the supervisor or others (eg practice manager, nurse). Practices could help registrars manage their workload by adding extra gaps between consultations or promoting registrars’ workload autonomy, provided that this strategy did not overtax other practice staff. The degree of supervisors’ supportiveness was also raised. Supportiveness involved regular two-way communication between the supervisor and registrar so that the supervisor was aware of, and could adapt to, the registrar’s current professional needs and – where appropriate – personal issues. Finally, some suggested that practices’ adherence to training requirements (eg providing mandated protected study time) helped prevent registrar burnout.

RTO-level strategies fell under six themes. The dominant theme, raised by all groups, was engaging with registrars. Participants commonly recommended that RTOs monitor registrars’ wellbeing, potentially through surveys or face-to-face ‘catch-ups’. Regular monitoring involved understanding each registrar’s needs and circumstances to tailor training – an endeavour potentially supported via in-depth interviews with registrars on commencement of training. Some even suggested ongoing contact to ‘see how everyone’s going and … form a profile about … the trainees’ [Registrar]. Participants encouraged RTOs to highlight common stressors and relevant resources or supports to registrars (eg education and workshops on burnout). These strategies required RTOs to establish relationships and maintain contact with registrars throughout training to facilitate approachability.

Another major RTO theme, raised by registrars and PCs, involved optimising the educational experience. More specifically, enhancing practice quality – particularly adequate support and clinical exposure – through standardisation, greater enforcement and incorporating wellbeing considerations into practice accreditation. Some suggested that RTOs develop wellbeing guidelines and training for their practices and supervisors. A related consideration related to placement allocation and prioritising the ‘fit between trainee and supervisor’ [Registrar].

There were another four minor themes. One of these, emphasised by most groups, involved helping registrars to develop peer support networks. Examples included establishing regular registrar workshops throughout training or offering a peer mentoring system. Another theme reflected transforming the hospital-to-general-practice adjustment into more of a transition. Strategies to achieve this included education about common adjustment issues (eg imposter syndrome, ‘feeling overwhelmed [with] responsibility’ [Registrar]) and allowing pre-semester observation periods to learn the procedures of general practice (eg consultations, workflow, administrative practices). RTOs were also encouraged to be flexible and promote registrars’ autonomy, particularly for part-time registrars. The final RTO theme involved reducing the educational load, particularly the assessment volume, on registrars, as well as helping registrars find placements. RTOs or practices that implemented tokenistic interventions, such as providing supportive messages without supporting or instigating ‘structural changes that address the underlying causes of burnout’ [Registrar], were considered unhelpful by registrars.

At a system level, registrars called for increased clarity about requirements, such as making mandatory reporting laws less ambiguous, or colleges needing to increase transparency regarding examination study requirements. Greater acknowledgement of, and support for, doctors’ wellbeing was endorsed by all groups. Specific strategies included mandating wellbeing courses as part of one’s continuing professional development to ‘[reinforce to] doctors that [they] need to be looking after [themselves] … [and] take the initiative to do it’ [Registrar], and cultural changes to encourage open and supportive discussion of doctors’ wellbeing within medicine. A further theme related to greater financial supports for registrars. Extending the gap between semesters, particularly for rural registrars, who often require extra time for relocating, was one suggestion. Parity between general practice and hospital registrars’ working conditions (eg annual leave) and funding to increase rural registrars’ leave opportunities and reduce financial pressures from relocating were also suggested.

Burnout reduction

Suggested burnout reduction strategies were also allocated between the stakeholder groups (Table 2).
 

Participants believed the first step for registrars was to acknowledge and troubleshoot their burnout. Acknowledgement was seen as crucial to effective management, whereas denial inhibited interventional effectiveness. However, the dominant theme noted by all groups was re-establishing sustainability. Registrars needed to reduce pressures, such as modifying their work structure to ‘[keep] them engaged but … to a level that’s actually comfortable’ [Supervisor] (eg adding extra catch-up appointments), reducing their hours, asking for assessment extensions and actively seeking support from practice staff. A complementary subtheme involved prioritising replenishing activities (ie focusing ‘on recharging [their] battery more effectively’ [Registrar]) and correcting unhealthy lifestyle habits that had developed in response to burnout (eg re-establishing personal–professional boundaries such as leaving work on time, replacing unhealthy eating and drinking habits with healthier alternatives). Participants believed paid leave was required to facilitate this. Registrars also needed to seek support from RTOs, practices, supervisors or mental health professionals. Indeed, one PC stated that registrars should ‘let us know if you’re not coping’.

Among the few maladaptive strategies identified for registrars was persisting or working harder and sacrificing sleep to keep pace with work demands. Similarly, not addressing the cause of burnout – often through internalising problems, withdrawing or unhealthy venting (eg substance use, frequent sick leave) – was seen as counterproductive.

In addition to registrars’ role, participants believed the duty of practices and RTOs included understanding and helping to manage registrar burnout by ‘just having a chat with the registrar and … [asking] “what is going to help you right now”’ [Registrar]. Both groups were encouraged to be flexible. Practices could offer registrars greater supports to manage their workload, while RTOs could delay assessments or offer part-time options. An added suggestion was that RTOs should also deliver resources and support services. Promoting (and potentially offering) mental health support services and providing easier access to Employee Access Programs were two such examples. Instigating a ‘flexible leave arrangement’, whereby registrars suspected of experiencing burnout could, at short notice, receive a brief period of leave with a locum rostered to cover their workload, was another example. Some also endorsed a ‘what to do if you’re suffering from burnout’ [Registrar] guide to provide general solutions, including relevant mental health services details. This support role extended to advocacy, whereby RTOs helped registrars modify their professional stressors, including changing a registrar’s placement.

Problems arose when practices or RTOs failed to acknowledge or understand the registrar’s needs, or did not support the registrar to minimise burnout. Examples included telling a registrar to ‘toughen up’ or not offering extensions for educational assessments. Similarly, being paternalistic could further decrease registrars’ autonomy. One supervisor argued that failing to understand the situation and assuming a work-related cause could lead to unnecessary (and unhelpful) work-based interventions for personal problems. Further illustrative quotes are provided in Appendix 2.

Discussion

The present study explored Australian general practice training stakeholders’ perspectives and experiences regarding burnout prevention and reduction strategies. An overarching concept of normalisation was proposed, consistent with previous suggestions.^6,19,20 Normalisation was viewed as everyone’s responsibility, suggesting a need for strategies that could be enacted by various groups (eg registrars, supervisors, practice staff, MEs, peers) on a daily basis (eg role modelling wellbeing).

Burnout prevention and management strategies for specific groups were also identified. For registrars, the overarching message was to achieve sustainability (eg prioritising replenishers, optimising working conditions). This aligns with previous research,^5,21–23 suggesting that while contextual factors (eg specialty, training model, seniority) are influential,^7,8 individual-level strategies (eg problem-focused coping, not persisting with work when experiencing symptoms) remain similar. However, developing a ‘burnout contingency plan’ appears to be novel. Given the centrality of exhaustion to burnout,²⁴ registrars experiencing burnout may struggle to develop effective recovery strategies. Thus, having an individualised recovery plan may help prevent burnout from worsening and could complement individualised wellness ‘learning plans’.²⁵ Development of self-management tools to guide preparation of such plans could be further explored.

Practices’ duty was seen as building a supportive environment, entailing strategies such as debriefing and workload management. These strategies mirror the resilience-promotion techniques identified by Zwack and Schweitzer.⁶ Notably, these techniques were raised as practices’ responsibility, rather than trainees’. This finding may reflect registrars’ lower professional autonomy when compared with consultants, emphasising the distinction between burnout prevention strategies for these two groups. Additionally, while supervisor supportiveness was predominantly expressed in terms of communication in the present study, other aspects are pivotal for registrars’ training (eg bond, role agreement).²⁶ When implementing practice-level strategies, one must also balance registrar wellbeing with that of other practice staff as well as the practice’s sustainability.

Some RTO strategies (eg optimising educational experiences, facilitating peer supports, increasing autonomy) align with experts’ recommendations.^11,13 However, suggestions of how to ease the adjustment to general practice in the training context have not previously been explored in the postgraduate medical trainee literature.²⁷ Moreover, while monitoring and raising awareness of burnout has been raised,²⁸ the focus on better understanding registrars could reflect training organisations’ relatively low direct contact with trainees in Australian general practice training. Training organisations might consider implementing strategies such as induction interviews and opportunities for trainees to raise and discuss mental health issues.

Notably, most of the identified strategies focused on organisations and systems rather than individuals, aligning with research on the causes of burnout.^11,12 To this end, the delineation between practice and training organisation strategies raised in the present study may help translate interventions from centralised training models. Specifically, practice strategies largely focused on workload and immediate support (eg debriefing), whereas RTO-level strategies focused on structural changes to training. This distinction can be applied to established wellbeing strategies. For example, Shanafelt and Noseworthy’s¹³ notions of ‘promoting flexibility and work–life integration’ and ‘harnessing the power of leadership’ predominantly relate to practices. Conversely, ‘acknowledging and assessing the problem’ and ‘providing resources to promote resilience and self-care’ can be enacted by the training organisation.¹³

The finding that broader systemic-level strategies focused on increasing awareness, emphasising wellbeing and offering financial supports aligns with previous calls to increase the supportiveness of medical culture.¹⁹ Barriers to support-seeking and burnout detection are a further consideration.²⁹ Elements of the medical culture, particularly the stigmatisation of imperfection, can be detrimental for doctors’ wellbeing.^19,20 There remains an urgent need to destigmatise support-seeking in the medical profession through raising awareness of the effectiveness of psychological and (particularly for trainees) educational interventions.^30,31

The present findings must be considered in light of certain factors that may influence their transferability. First, participants were recruited from one RTO, potentially limiting the breadth of strategies suggested. Secondly, the focus on one specialty limits the transferability of the findings to other training contexts. Future research could incorporate different stakeholders across specialty and training model contexts to ensure a comprehensive understanding of universal and setting-specific considerations. Additionally, the interviewer’s affiliation with the RTO may have influenced some participants to provide favourable responses regarding the RTO. Finally, the temporal context of the present study must be acknowledged. During this research, training was jointly managed by RTOs and the colleges. The delineation of roles meant that most registrar direct contact during their training program was with RTOs, reflected by most themes centring on the RTO rather than the colleges. Given the current transition of the Australian general practice training program, it is anticipated that many RTO themes will also be relevant to the colleges.

Conclusion

Investigation of burnout prevention and reduction strategies must be sensitive to training models and trainees’ specialty. The authors encourage general practice trainers to consider the delineation between practices and training organisations when attempting to apply interventions from centralised training models. Understanding and recognising this distinction will facilitate the translation of intervention research. To ensure effectiveness, strategies also should be tailored or adapted to address the specific burnout causes for general practice trainees.