‘Being in a place where it matters’: GPs who do and do not work in Aboriginal Health Services

Recruitment and data collection

We conducted both studies within Australian General Practice training, where registrars complete training across multiple practices. Study-Remained and Study-Left were completed in 2017 and 2019, respectively. Both studies were overseen by a steering committee. The committee included Aboriginal representation, providing cultural guidance regarding the design of the study, interpretation of the findings and consideration of translating the findings.

The sampling frame comprised GPs who had completed fellowship training via a South Australian training organisation between 2007 and 2017 (Study-Remained) or 2012 and 2019 (Study-Left), had completed an AMS training placement (including part-time placements), and were (Study-Remained) or were not (Study-Left) working in an AMS at the time of the respective study. All individuals meeting these criteria were invited to participate via email and followed up through phone calls. Given the small sampling frames for both studies, recruitment ceased once no further responses were received from prospective participants. Snowball sampling was not used, as the researchers had access to the contact details for the entire sampling frame.

Participants completed a brief demographic survey (available in Appendix 2) and a 30–60-min semi-structured interview (question schedules available in Appendix 3). Participants were reimbursed for their time in line with contemporaneous RACGP guidelines (A$125 per hour). TE or JB (Study-Remained) and AB (Study-Left) conducted interviews in-person or via phone. With participant permission, we audio-recorded interviews. Recordings were transcribed verbatim by SP (Study-Remained) and either AB or a professional transcriber (Study-Left). After we de-identified transcripts (e.g. removing names, locations), we invited participants to review their transcript for accuracy.

Analysis

We analysed data for each study separately, following the same procedure. SP (Study-Remained) and AB (Study-Left) used NVivo for Windows (QSR International) to line-by-line thematically analyse the data, informed by grounded theory (Glaser and Strauss 1967; Kennedy and Lingard 2006; Hall et al. 2013). We followed the six stages of thematic analysis defined by Braun and Clarke (2006). Although we identified themes inductively, we used the research aims to categorise themes. Analyses finished once thematic saturation was reached. We employed an iterative process of coding verification with other members of the research team, involving independent re-coding of sample interviews, discussions of disagreements and revisions to the coding structure. This provided forums to identify and consider the effect of each researcher’s experiences. During involvement in the respective studies, AB and JB were working in Aboriginal health, whereas JB, TE and SP were affiliated with the South Australian GP training organisation. To enhance the trustworthiness of the coding scheme, TE independently coded a transcript selected by the primary analyst in each study. Given the few codes in the structure minimised the likelihood of chance agreement, we evaluated inter-rater reliability by dividing the number of agreements by the number of agreements and disagreements (Miles and Huberman 1994). Final inter-rater reliability reached 71% in Study-Remained and 84% in Study-Left.

Ethics approval

The present manuscript reports on two studies that were each granted ethical approval by a formal Human Research Ethics Committee (University of Adelaide Human Research Ethics Committee HREC H-2019-070, and Flinders University Social and Behavioural Research Ethics Committee Project number 7684). The research was undertaken with appropriate informed consent of participants or guardians.

Participants

For Study-Remained, 16 GPs were identified as eligible to participate and contacted, with 11 participating (69% response rate). Of the 75 GPs initially contacted for Study-Left and thought to be eligible, 21 responded, of which 12 confirmed as eligible to participate. Nine GPs ultimately participated in Study 2 (75% response rate). There was a sex balance in the Study-Remained participants (n_male = 6), but only one male (11% of sample) participated in Study-Left. Given the thematic overlap between the studies, we present the results from both studies for each question simultaneously, comparing the studies throughout. Themes are summarised in Table 1. Note that themes are bolded in the text and quotes attributed to participants in the format Study-Participant number (e.g. ‘R5’ refers to Study-Remained, Participant 5; Table 1).

What motivated choosing an Aboriginal health placement?

A common motivation for both groups requesting an Aboriginal health placement was an interest in Aboriginal culture or working in this sector. Some participants reported having an ‘…intuitive respect for [Aboriginal peoples]…’ (R5) and so used these placements as opportunities to ‘…meet Aboriginal people and to work in Aboriginal Health…’ (R6). Similarly, participants described a ‘…duty to serve them [Aboriginal peoples] for some time at least…’ (R3), as it was ‘…an area where you can do a little bit and make a lot of difference’ (L7). Many participants also cited pre-placement exposure to the Aboriginal health sector, which had helped to put ‘…Aboriginal health on the map…’ (R9).

As shown in Table 1, participants in Study-Left raised a further two motivations. First, they identified broader educational opportunities of the placement than Study-Remained participants. Such opportunities included building their communication skills and understanding of the social determinants of health, or gaining exposure to subspecialties. Some participants also noted pragmatic considerations, such as choosing their placement to experience Aboriginal health without relocating to a rural area.

What were the positive aspects?

Participants described a variety of positive factors, which encouraged Study-Remained participants to remain in Aboriginal health. The dominant theme within Study-Remained was the rewards of the work. Many drew satisfaction from improving their patients’ wellbeing, whereas others relished the challenge of the clinical complexity. Most also found their job satisfied their social justice drive to ‘…[do] something really worthwhile…’ (R9). The rewards of the job were less frequently raised by Study-Left participants.

Participants across both studies valued organisational structures. They contrasted the greater freedoms and flexibility in AMSs with the pressure to ‘…quickly see and get them [patients] through the door…’ in mainstream practice (R3). This flexibility enabled participants to ‘…not just [treat] the medical problem, but [look] at the social issues’ (L2). Another consequence of organisational structures was an emphasis on teamwork, which provided a ‘sense of community’ (L2) and ‘…a good way of approaching the delivery of health care to that group of people…’ (R10). For one participant in Study-Remained, this contrasted with her experiences of loneliness in mainstream practice. However, within this theme, only participants in Study-Remained commented on the supportiveness afforded by organisational structures, particularly the insights offered by Aboriginal Health Workers (AHWs) to ‘…target my … management plan…’ (R8).

Closely related was the theme of building relationships. Both participant groups acknowledged ‘…develop[ing] a connection and a rapport…’ with their patients and communities (R7). These bonds encouraged Study-Remained participants to continue working in AMSs. Participants also appreciated the connections with, and support from, their colleagues, although Study-Remained participants held more favourable views of their supervisors than Study-Left participants.

Across both studies, participants noted they learnt much from their placements. Topics included Aboriginal culture, the social determinants of health and specific diseases.

What barriers did registrars face?

Both participant groups acknowledged difficulties from a lack of resources and support within the Aboriginal health sector, such as lacking an on-site supervisor. Study-Remained participants viewed this as challenging, but an opportunity to build self-reliance. Conversely, Study-Left participants described this situation as isolating. Both groups noted needing to adopt responsibilities typically beyond those of a GP, such as organising patient transport and monitoring recall. Again, this appeared to be more problematic for Study-Left participants. Where Study-Remained participants largely discussed the implications of workforce instability as impeding patient care and trust in the medical system, one participant in Study-Left reflected on the chaos this brought to their clinic and how they needed to adopt a support role for their colleagues.

Politics and bureaucracy associated with Aboriginal health were raised in both studies. Again, this was a greater barrier for those in Study-Left – ‘It was those sort of management interactions … which probably left me with a slightly negative feeling about working in Aboriginal health, because I had really seen how these organisations can run’ (L7).

Challenges from navigating cultural differences were raised by both groups. Differences between registrars’ and patients’ ‘priorities in life… [made it] challenging to build rapport’ (R3). Social inequalities needed to be handled sensitively; one participant reported having to reduce his enthusiasm to ensure his ‘…desire to be ‘the good guy’ [didn’t] drag up some of that patriarchal … anger at the past’ (R6). Likewise, a female participant in Study-Left commented on challenges of being the only available GP and treating traditional male patients who preferred a male doctor.

Both groups also struggled with patients’ medical and social complexities. Patients would often present with severe, complex and unmanaged medical conditions. Beyond testing registrars’ medical competence, accumulating these experiences could leave a feeling of ‘…inadequacy at the vastness of some of the health problems…’ (R6), particularly ‘…when you see somebody that’s so much a … product of the social determinants’ (L7). The high mental health caseload could be ‘taxing’ (R2), and ‘…the vulnerability of the clients…’ left one participant feeling pressure to not ‘[disadvantage] clients who are already disadvantaged by making a wrong call [medically]’ (R4).

A further frustration, particularly for Study-Left participants, was the unpredictability of working in an AMS. Patients’ sporadic attendance could prompt discomfort. Likewise, some new patients would visit the clinic and request medications, which meant ‘…chas[ing] down people’s history, … what medications they were on, … it was just a lot of work … and can be quite exhausting’ (L2).

Study-Remained participants cited limited opportunities to train in an AMS. This included restrictions on registrars being able to complete an AMS placement early in training or completing an extended (12-month) AMS placement. Study-Left participants sometimes found their placement’s location as problematic, particularly when working across sites produced fragmentation. Some Study-Left participants reported negative interactions with clinic staff, including limited support from clinicians, and staff seeking personal medical advice.

Study-Left participants’ decision not to return to an AMS was largely driven by ties to their current practice. Location re-emerged, with their current clinic being closer to home. Some also enjoyed the culture and relationships in their practice. Like Study-Remained participants, Study-Left participants noted rapport with patients in their clinic created a ‘tendency … to want to keep doing it[;] … it becomes … harder to leave when … you know whole families’ (L7). Participants noted the lower income and autonomy associated with Aboriginal health. However, eight of the nine Study-Left participants wanted to return to Aboriginal health, highlighting a further barrier of difficulty identifying job vacancies.

How did registrars overcome these barriers?

Participants in both studies described two categories of strategies to overcome barriers – drawing on others for support, and adaptation. As highlighted in Table 1, though, the two samples differed in their utilisation of these strategies. Regarding supports, all Study-Remained participants reported drawing on colleagues to learn how to overcome barriers and debrief, whereas this was infrequent among Study-Left participants. Specifically, although relying on the training organisation for support was common to both samples, seeking supports within the clinic was more common among Study-Remained participants. AHWs supported patient follow-up, provided cultural guidance and helped guide case management. Speaking with Elders helped one Study-Remained participant to understand the cultural and historical background of their location, which facilitated greater empathy. Two Study-Remained participants’ families helped support them through challenges. Underlying this, Study-Remained participants reported being more proactive in their help-seeking than Study-Left participants; for example, actively building rapport with AHWs to build their standing in the community.

In adapting to their situation, a strategy common across samples was accepting one’s limitations – ‘…health professionals have one part to play, but it’s also a lot more social issues and the wider issues to do with the community themselves and … as a doctor you can’t solve that…’ (R10). This acceptance was not absolute; one Study-Remained participant reported building his assertiveness to ensure patients received appropriate treatment. Both groups adapted their communication styles to be more culturally appropriate. Beyond this, each group’s strategies differed. Study-Remained participants took an external focus. Many emphasised flexibility; for example, practising opportunistic medicine or not fixating on appointment times. Participants emphasised empathy, and how acknowledging structural factors could help them manage work frustrations. For instance, two participants reported considering their patients’ perspectives and the uncontrollable barriers they face when forming their opinions and approach to consults.

Study-Left’s participants instead focused on reframing challenges with an internal focus. For example, one reframed lacking a regular supervisor into an opportunity to learn different consulting techniques. Likewise, patients’ confronting psychosocial circumstances could enthuse them about their work.

I just remember coming home, thinking … I was coming home to a safe house with food and a safe partner. And then there were people … a few kilometres away from you [for whom] that… wasn’t the case. … I think that was [inhale] probably challenging, not being able to stick your head in the sand about things … but that’s also I guess what makes you passionate and engaged. (L1)

Similarly, some reframed patients’ minimal clinical progress to the ‘…hope that you’re actually making some kind of difference, and a lot of the time you don’t, but at least you’re there trying and [that’s] better than not helping’ (L2).

How can training organisations improve?

Generally, participants were pleased with their training. Broadly, participants recommended better preparing registrars. For instance, some suggested developing an orientation session or handbook for each AMS detailing the community’s historical and demographic background, and the setup of the AMS (e.g. the role of AHWs, patient screening processes). Study-Left participants supported the training organisation’s mandatory cultural awareness training, although recommended it focus more on day-to-day practice. Several Study-Left participants believed registrars completing an AMS placement would benefit from exposure to rural communities to understand ‘…when people come down from the lands, what barriers [they will face] when you send them back’ (L1). Finally, Study-Remained participants suggested providing registrars more advanced training about mental health (particularly regarding trauma, and drug and alcohol issues) and public health theories.

Another overarching theme was better supporting registrars. Participants advocated for mentorship separate from their supervisor, particularly a same-sex mentor with a background in Aboriginal health. Similarly, Study-Remained participants suggested developing specific support services for AMS registrars, as generic support services’ lack of understanding about Aboriginal health limited their helpfulness. Some recommended providing debriefing sessions, self-care workshops and check-in calls. One suggested creating networks of registrars in Aboriginal health placements to build connections. Study-Left participants emphasised clinics supporting registrars’ immersion in the local community to improve registrars’ local understanding and relationships, and incentivise returning to Aboriginal health. Although having on-site and consistent supervision was desirable, participants acknowledged feasibility constraints. Some Study-Left participants urged training organisations to be more selective in the clinics that were training posts, emphasising supportiveness and stability.

Consistent with exposure, Study-Remained participants suggested ‘…to continue these placements and create more…’ to foster interest in the sector (R6). These participants raised alternative arrangements to maximise exposure, such as job-sharing and offering 12-month placements.

Key findings

This article reports on two studies examining how GPs’ training experiences in AMSs impacted their career decisions. The results identified a variety of motivations and identified improvements.

The career motivators for GPs working in Aboriginal health overlapped with those amongst GPs in rural or other under-served areas. These include social justice values, finding the work to be rewarding and exposure to the sector (Worley et al. 2008; Odom Walker et al. 2010; Wayne et al. 2010; Stevenson et al. 2011; Eley et al. 2012; Playford et al. 2014; Kondalsamy-Chennakesavan et al. 2015). Although relationships have not been raised in the literature, rural doctors have been characterised as enjoying relationship building (Eley et al. 2009). Another strategy for career longevity was perspective-taking, which was raised by GPs working with disadvantaged groups (Stevenson et al. 2011).

The present research also aligns with Abbot et al.’s (2014) findings about important attributes for GPs delivering health care for Aboriginal patients. These included the importance of cultural sensitivity and awareness, engaging with the community to establish trust, proactively seeking cultural guidance, and using empathy to acknowledge historical and social factors. Thus, strategies that support quality care delivery also appear to help GPs to continue working in the sector.

Table 1 compares experiences of GPs who remained and left Aboriginal health. Both groups held similar motivations for pursuing a placement and reported common positive experiences. Those who remained found the experience more rewarding and enjoyed more positive collegial interactions. Despite facing similar barriers, those who left Aboriginal health appeared to place greater weight on some barriers and faced the additional barrier of the ties to their current practice. Both groups demonstrated adaptation when responding to barriers. However, those who stayed in Aboriginal health employed a broader array of coping styles, including self-reliance, support seeking, information seeking and accommodation. Indeed, their acceptance of factors beyond their control mirrors the complex locus of control previously identified (Stevenson et al. 2011). Conversely, those who left Aboriginal health tended to focus on reframing their situation (i.e. more of a negotiation coping style; Skinner et al. 2003). These findings suggest that differences in career decision appears to largely – but not entirely – stem from individual differences regarding perceptions of, and responses to, barriers.

Implications

As raised by Study-Remained participants, opportunities for undertaking AMS placements need to be maximised. This aligns with evidence that emphasises exposure (Worley et al. 2008; Playford et al. 2014; Kondalsamy-Chennakesavan et al. 2015). Beyond increasing placement sites and opportunities for engaging in AMS advanced and extended skills placements, participants proposed job-sharing and fly-in-fly-out opportunities. The feasibility of these should be explored.

Greater supports may be provided prior to registrars commencing their placement. Orienting registrars to the local culture and community may support registrars’ confidence and ability to engage with the placement more thoroughly. Opportunities for greater clinical training (i.e. mental health, public health) were recommended.

During placements, mentorship programs could offer registrars independent support and guidance. These programs could also provide contextualised mental health support services. Considering the differences in participants’ responses to barriers across the studies, registrars may benefit from guidance to not only reframe stressors, but use them as growth opportunities and more proactively pursue support. Given the strong overlap between the groups, future research examining personality traits and coping styles of those who remain and leave Aboriginal health may provide further direction (e.g. Eley et al. 2009). This could inform training for medical students and GP registrars to optimise their coping strategies prior to commencing AMS placements.

Strengths and limitations

Most participants in both studies completed placements in rural areas (Study-Remained n_Rural = 7 [64%], Study-Left n_Rural = 5 [56%]), which could explain the overlap in themes between the present article and research with rural doctors. However, the similarity in urban and rural participants’ themes in this research suggests overlap in the experiences between AMS and rural registrars, rather than confounding.

We acknowledge that the views of participants may not represent the broader population of GPs. This is particularly relevant for those who did not continue working in Aboriginal health, as non-participants may have feared judgement for their career decision. Similarly, we only recruited participants from one state. Although participants were sourced from numerous practices, this impedes our understanding of potential changes to training practices and other contextual factors in different jurisdictions.