What leads to burnout in general practice supervisors? A qualitative study

This article is part of a series of articles on doctors’ health and wellbeing.

Medicine faces a wellbeing crisis,^1–3 especially among its trainees.^4,5 Research is recognising that the wellbeing of educators (eg faculty, supervisors) is critical for medical trainee wellbeing.^6–8 Indeed, trainees’ burnout correlates as much with their perceived relationships with their educators as it does with their own sense of professional autonomy or dissatisfaction with work–life balance.⁹ Furthermore, educators experiencing poor wellbeing will model this to trainees, who might internalise this as part of their professional identity formation, thereby perpetuating poor wellbeing.^6,8,10,11 Thus, educator wellbeing is a key strategy to support learner wellbeing.¹²

Supporting educator wellbeing will likely introduce broader benefits for medical education. Educator wellbeing is associated with higher education quality.^13–15 Additionally, addressing educator burnout might improve stability within the medical education workforce.^16,17 As practising clinicians, burnout interventions might also improve educators’ clinical practice, benefitting their patients and workplaces.^18–20 Enhancing educators’ wellbeing is therefore a research priority.

Research from the broader doctor wellbeing literature is somewhat applicable to educators. However, educators’ duties differ markedly from those of non-educator clinicians. Educators combine clinical practice with teaching, supervision and assessment of trainees, introducing distinct stressors. Research specifically with educators is needed to understand these nuances. However, there are notable gaps in the educator wellbeing literature. A systematic review focusing on general practice educators identified that research has largely focused on individualistic risk and protective factors for burnout (eg resilience, sense of meaningfulness) rather than considering organisational and systemic risk factors (eg remuneration, medical culture).²¹ Further, there is little research from other specialties to provide additional insights.^22–25

To date, research has focused on ‘faculty’ rather than ‘supervisors’. Unlike faculty, who often hold academic duties alongside their teaching and clinical duties within a large training organisation (eg hospital, university), supervisors practise in small clinics, sometimes being the only supervisor within their clinic, and are tasked with supervising and teaching trainees with often limited connection with training providers. Thus, faculty’s experiences likely differ from supervisors’ experiences. This study aimed to address these literature gaps by exploring the causes of burnout among Australian general practice supervisors. A secondary aim was to understand how these causes could be mitigated to support their wellbeing.

Methods

This manuscript adheres to the Standards for Reporting Qualitative Research Guidelines (refer to Appendix 1, available online only).²⁶ Using a critical realist epistemology, the researchers acknowledged they could only approximate the nature of the phenomena being studied.^27,28 This epistemology was adopted to increase the findings’ longevity and relevance, particularly given the study was undertaken shortly before the transition to national College-led training.

Recruitment and data collection

Interviews with supervisors and focus groups with training organisation-affiliated staff were conducted across two training organisations (GPEx and GPTQ) spanning two states (South Australia and Queensland).

Supervisors could participate if they were actively supervising an Australian general practice training registrar ‘on-site’ (ie not remotely) through either participating training organisation. So that participants could reflect longitudinally on their experiences and observations, supervisors needed at least 3 years’ supervision experience. Supervisors who were also medical educators (ie those responsible for overseeing registrars’ training and delivering out-of-practice education to registrars) or supervisor liaison officers could not participate in interviews, as their views were sought via focus groups. Given the stigma associated with identifying current or historical burnout experiences,⁶ supervisors were not asked whether they personally had experienced burnout.

All supervisors from participating training organisations were emailed invitations to participate in the study and sent one reminder. The emailed survey (refer to Appendix 2, available online only) verified supervisors’ eligibility and collected sociodemographic details. Although purposive sampling based on sociodemographic details was planned, interviews were offered to all supervisors who expressed interest because of low response rates. In the researchers’ experiences, low response rates to research invitations among supervisors are common and were likely compounded by the ongoing COVID-19 pandemic and imminent transition to a new training model. Recruitment occurred between April and June 2022, ceasing once all interested supervisors were given opportunities to participate. Interviews occurred between May and July 2022. Supervisors were provided an honorarium for their participation in line with The Royal Australian College of General Practitioners’ guidelines (AUD 125).

The authors also sought the perspectives of those working with supervisors, effectively providing an ‘other-report’. Medical educators, supervisor liaison officers, practice manager liaison officers and nominated training organisation staff from both training organisations were invited to participate in focus groups. Recruitment occurred from April to June 2022, and focus groups were held in June and July 2022. Participants engaged in one focus group each during paid working time.

SP and HM facilitated the interviews. SP facilitated the focus groups. This combination balanced SP’s academic experience in the field and HM’s experience as a supervisor. Recognising the potential power imbalance from HM’s role as a medical educator with one training organisation, SP conducted all interviews and focus groups from HM’s training organisation.

Interviews and focus groups followed the same question schedule (refer to Appendix 3, available online only), adapted from a previous study and reviewed by the project’s steering committee.^29,30 These questions focused on burnout, but construed wellbeing as a continuous spectrum encompassing personal and professional domains, with burnout representing one point on this spectrum.³¹ The schedule was piloted and interviewers met to standardise their approach to questions. Interview questions remained unchanged after piloting.

Interviews and focus groups were conducted via Zoom (Zoom Communications Inc., San Jose, CA, USA) and audio recorded where participants consented. Interviewers de-identified recordings, which were professionally transcribed. Participants were provided a copy of their transcript to review if requested. No transcript changes were made by participants. Subsequently, the respective interviewer removed identifying details and added sociodemographic details (eg gender, duration of supervisory experience) to the transcript, then imported it into NVivo 20 for Windows (Lumivero, Denver, CO, USA).

Analysis

Each transcript was treated as one unit of study. SP, HM and JB independently analysed line-by-line a sample transcript using this structure. Each analyst developed a preliminary coding structure corresponding to each research question. The analysts met to share and fuse their coding structures into a preliminary structure to guide further analysis. Rolling analysis followed; facilitators analysed their transcripts as they were received. The preliminary coding template guided analysis, with analysts adapting and expanding it where required. Analysts kept an audit trail of their coding, noting emerging patterns and describing their coding decisions. Given concurrent data collection and analysis, facilitators adapted their questions using emerging themes. Once all data were analysed, SP and HM compared their coding structures and fused these into a master coding structure. The analysts then agreed that thematic saturation had been reached. SP applied the master structure to the interviews conducted by HM to familiarise himself with the entire dataset. The few focus groups limited the richness of these data, precluding triangulation with interview data. Instead, interview and focus group data were analysed simultaneously within the one coding structure. HM analysed one of SP’s interview transcripts using the master coding structure. Inter-rater agreement for the coding structure reached 100%. The University of Queensland Human Research Ethics Committee granted ethical approval for this study (2021/HE0022612).

Results

Participant recruitment and characteristics

Fourteen supervisors participated in individual interviews (Figure 1). Three focus groups were conducted, comprising five participants in total – three medical educators, one training organisation education administration staff member and one practice manager. One medical educator ‘focus group’ comprised only one medical educator.

Figure 1. Flow diagram of supervisor recruitment.

Most (63.2%) participants were female. Of the 14 supervisors, half practised in an urban area, and most also supervised medical trainees besides general practice registrars (92.9%) and worked in a practice with up to 10 clinicians (85.8%). Most supervisors (57.1%) had been supervising for between 3 and 5 years. The supervisors had worked as general practitioners (GPs) for various lengths of time. Participants were approximately evenly distributed between the two training organisations. Further details are reported in Table 1.

What causes burnout in supervisors?

Themes were categorised using the model of factors affecting clinician wellbeing proposed by Brigham et al.³² These are summarised in Table 2 and discussed below. Given the overlap between organisational factors and the learning/practice environment categories when applied to the Australian general practice setting, these categories were collapsed for the present study.

Personal factors

Supervisors’ personal psychosocial context was discussed considerably. One aspect was personal stressors and responsibilities. This spanned ongoing commitments (eg caring responsibilities) and acute stressful events, such as natural disasters, illness and relationship disruptions. To counterbalance this, participants emphasised having personal social supports to provide practical and emotional support: ‘… having a supportive family is pretty important if you’re going to do any level of supervision, because there are going to be times when you’re unexpectedly going to need to do extra stuff’ (Supervisor).

Another risk factor could be supervisor’s psychological traits. For example, perfectionism combined with low uncertainty tolerance could mean a supervisor had ‘… high demands for our registrars and (themselves) … (leading to) an additional strain ...’ (Supervisor). Similar issues could arise ‘… if you’re not able to give up autonomy and you’re very much a controlling person’ (Supervisor). Another example was a supervisor’s sense of duty towards others, which would encourage self-sacrificial behaviours. Supervisors’ coping mechanisms also featured; disengaging from work and supervision when off-duty was important to facilitate boundaries. Conversely, supervisors who were reluctant to seek support from others could be at elevated risk of burnout. Psychological support for unhelpful traits (eg perfectionism) and developing awareness of one’s values were considered beneficial, potentially supported by the practice or training system.

A further theme was the supervisor’s financial situation. Background financial stress – either personal or business – could be exacerbated by loss of income from supervising, because ‘… even though you’re teaching, it’s not enough to pay your pay bills’ (Supervisor). The impact of financial stress from teaching varied by career stage:

Junior supervisors are far more in need of the financial input because they’re usually supporting a family … to take an hour out a week, and then to take regular time off, and decrease the amount of patients you’re seeing, can impact on them financially. Whereas those of us who are older … it doesn’t impact … as much. (Supervisor)

Skills and abilities

Participants’ discussions of supervisors’ skills and abilities pertained exclusively to supervision, particularly duration of supervision experience. The large learning load facing new supervisors could make them feel ‘… that you don’t have the skills to be able to … bring them (registrars) to where they need to be …’ (Supervisor), particularly for registrars requiring extra support. Similarly, struggling registrars could prompt self-doubt: ‘Am I the reason that they’re struggling? Am I not giving enough education or answering their questions …’ (Medical Educator). Supervisors who had recently commenced independent practice were at greater risk still: ‘… if (I’m) not even confident in (my) own skills … then how do I teach a registrar how to do that (procedure) as well?” (Supervisor).

Conversely, experienced supervisors’ preconceptions when judging a registrar’s competence could induce stress when a registrar was not meeting their internal benchmarks. Likewise, skewed experiences with under- or over-performing registrars could distort these benchmarks and introduce ‘a risk … of not just seeing them (the registrar) for who they are’ (Supervisor). Experienced supervisors, by virtue of their career stage, likely held greater professional responsibilities (eg leadership roles in the practice), creating additional demands.

To mitigate these risks, supervisors recommended aligning one’s supervision workload with one’s supervision experience to minimise becoming overwhelmed. More formalised opportunities to develop supervisors’ supervision skills was viewed as key.

Healthcare role

Participants acknowledged the inherent stressors of general practice. Some commented on the draining nature of clinical work, including constant decision making, high load of mental health presentations (which ‘… (puts) an extra load on you as a doctor and as a person’ (Supervisor)), experiencing vicarious trauma, and risks of negative clinical outcomes. This was compounded by the large clinical and administrative workload facing GPs, exacerbated by workforce shortages and the COVID-19 pandemic.

Overlaying this was the ‘burden of supervision’. Participants largely commented on trainee characteristics, such as the stress and effort required to manage registrars with competency difficulties or who were disengaged from learning, overconfident and/or unreceptive. Critically, these experiences could accumulate: ‘if they (the registrar) don’t put in the effort, they bring you down. Then another one brings you down a bit further. Eventually, you go, “oh well I guess they’re all like that”’ (Supervisor). Some supervisors also highlighted the extra work required to dismantle poor practice habits that registrars had acquired in previous placements. Another aspect of this supervisory burden was a sense of loss: ‘… despite you essentially nurturing and raising a really great registrar, … there’s things that are outside of our control, like their partners and their jobs, that they then leave and you’re like, “oh, is it worth it?”’ (Supervisor). At a systemic level, supervisors were concerned about the onerous assessment (and associated administrative) load, particularly when a registrar was not meeting expectations and required remediation.

To manage this ‘burden’, supervisors clearly defined the scope of their supervision to avoid becoming overwhelmed. Debriefing with others was also recommended, either at a practice level or within a broader supervisor community of practice.

However, supervision also provided meaning. Participants spoke about the sense of appreciation they drew from supervising; supervision could represent ‘… one of the times where you do get some thanks or acknowledgement for the work that you do …’ (Supervisor). Some also found joy when teaching registrars, noting supervision as a learning opportunity, a chance to ‘give back’ to the community, and greatly impacted a registrar’s career.

Learning/practice environment and organisational factors

An extensively discussed theme was the supportive culture of the practice for supervising. This included the administrative team understanding supervisors’ pressures and making accommodations (eg not booking over teaching time, allowing supervisors to hold a reduced clinical load). Motivational misalignment between supervisors and practice management, particularly if management’s motivation for supervision was financial, could mean ‘… (the workload is) going to fall to the supervisor’ (Supervisor). Practice culture also impacted a supervisor’s actual or perceived level of control. Having autonomy over appointment scheduling, supervision load and practising style was important. Conversely, feeling ‘trapped’ could prompt hopelessness.

Belonging to a supervisor community of practice could facilitate debriefing and sharing of teaching ideas. Co-located supervisors could share the supervisory load, although this required compatibility in supervision style to prevent registrars from ‘… call(ing) the doctor who … was more engaging with them’ (Supervisor) and overloading one supervisor. Participants encouraged a team approach to supervision to reduce the burden on any one supervisor. Participants recommended establishing regional or national supervisor communities of practice. This could facilitate delegation of supervision for specific areas of practice to those with greater expertise. Participants also suggested other practice staff could undertake administrative tasks, reducing supervisors’ workload. They also called for rationalisation of the assessment and administrative load.

Regulatory, business and payer environment

Various challenges facing Australian general practice were highlighted, including workforce shortages, low remuneration and feeling undervalued. These elements increased workloads and exacerbated existing demands by instilling hopelessness and futility – ‘the future of general practice seems somewhat bleak to me …’ (Supervisor). To offset these threats, participants called for growth in the GP workforce to reduce the workload facing the profession. Likewise, supervisors sought greater remuneration for supervision as acknowledgement of their role’s value.

Sociocultural factors

Participants discussed medical culture’s discouragement of doctors prioritising their wellbeing, describing strong stigma against being ‘… seen as the one with the mental health issue’ (Supervisor), exacerbated by a default attitude of thinking ‘… it is their fault and they should be able to fix it …’ (Supervisor). The combination of this stigma and an emphasis on self-sacrifice encouraged supervisors to prioritise patients and persist in the face of adversity rather than pause and reflect. Consequently, ‘Any suggestion that you’re not coping, even if it’s with the best of intentions, might actually push people away’ (Supervisor).

Discussion

Key findings

This study explored factors contributing to, or protecting against, burnout among Australian general practice supervisors. Among the areas highlighted by Brigham et al,³² participants emphasised personal factors (eg ability to disconnect from work, financial stress, caring responsibilities), healthcare role (eg demands from clinical work, the ‘burden of supervision’, professional meaningfulness) and the practice/organisational environment (eg practice culture, autonomy, having a supervisor community of practice). Many of these themes align with previous findings. For example, patient demands, workforce shortages and administrative duties affect the wellbeing of clinicians more broadly.^32,33 Similarly, connection with peers, meaningfulness from work and autonomy have each been raised within medical literature.^21,34 This study’s emphasis on meaningfulness is supported by a rich body of literature highlighting its protective effects for doctors’ wellbeing.^34,35 For example, Woodward et al found physicians with a low risk of burnout also discussed concepts of community and connection, calling and spirituality, and empowerment.³⁶

A novel theme was the ‘burden of supervision’. This primarily arose from interactions with trainees who were disengaged with learning or who required intensive support. This contrasted with the energising effects of supervising a trainee who was engaged with learning.³⁷ Further demands placed on supervisors entailed the educational and administrative load, which have previously been documented as problematic,^38,39 suggesting a need for educational and administrative rationalisation. The present study introduced additional factors influencing supervisors’ wellbeing, in particular the supportiveness of practice culture, medical culture, and concerns with the structure of general practice. Recognition of these organisational and systemic contributors addresses important literature gaps, enabling more holistic and nuanced recognition of the complex interplay of burnout causes facing this group.²¹

Implications

The present study findings offer pragmatic guidance at several levels. Box 1 lists practical strategies to prevent and manage burnout raised by participants in this study; the validity of many is bolstered by previous literature. Readers should note that the strategies highlighted for practices and training/health systems might overlap. An overarching message is the importance of supporting individuals to build professional meaningfulness. Values offer a useful framework for defining meaningfulness and could span aspects including professional relationships or ‘giving back’ to the medical community.⁴⁰