Theories and theoretical frameworks used in general practice career choice research: a scoping review

Abstract

Background

The number of medical graduates choosing to specialise as a general practitioner (GP) has not kept pace with need. Understanding how GP specialisation decisions are made has never been more critical. To date, much of the literature in GP specialisation choice is pragmatic and practice-focused, less strongly underpinned by existing theories or theoretical frameworks relating to the decision-making process of medical students or junior doctors. Our aim was to scope the evidence on the use of theories and theoretical frameworks in general practice career choice selection research.

Methods

The review was conducted and reported in accordance with JBI guidelines and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR). PubMed, Web of Science, Embase, and Scopus databases were searched. Only contexts where specialty decisions mostly occur after medical school graduation (e.g. Australia, United Kingdom) were considered for inclusion. Titles and abstracts of all articles were dual-screened against inclusion criteria, followed by full text review of articles.

Results

After removing duplicates, 2,396 titles and abstracts were screened followed by full-text review which identified 30 studies meeting the inclusion criteria. Of these, eight were from Australia, 20 from the UK, one from New Zealand, and one was multi-national. Most studies utilised qualitative methodology (N = 21) or mixed methods (N = 6). These draw from 15 distinct theories/frameworks, while 15 studies explained implicit theories. Twenty-two studies produced some form of theory or framework as an outcome within the study. Subsequent citation of these frameworks has generally been low. There was clear agreement that GP career decision-making was complex.

Conclusions

Our review identified the application and production of many different frameworks relating to GP specialty decision making, both from adapting explicit theories from a range of disciplines and professions, and identification of implicit theories. Researchers undertaking future studies around general practice career decisions are encouraged to utilise and build upon existing theories and frameworks, rather than creating new ones, to avoid duplication of results.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12875-026-03475-3.

Background

Ensuring that sufficient medical graduates choose to specialise in general practice (or family medicine) is important, but has generally not kept pace with need [1–3]. In many countries the general practice medical workforce underpins not only primary care but most other parts of healthcare systems, acting as the ‘gatekeeper’ for most initial triage, testing, diagnosis and treatment (including prescribing) as well as referrals to allied health services, specialist medical and/or tertiary hospital care. For example, hospital discharge documentation is sent to the patient’s regular general practitioner (GP) for ongoing management; similarly, chronic management plans are shared between specialist outpatient clinics and GPs to coordinate and monitor ongoing treatment. Barbara Starfield’s seminal work has demonstrated the effectiveness and efficiency of this primary care-led structure [4]. Recent evidence confirms the connection between a health system strongly based on primary care and improved population health [5].

Historically, most countries did not require GPs to complete additional specialty training. In Australia, it took until 1995 for completion of fellowship with the GP college to become the compulsory end point of training for those choosing general practice [6, 7]. A concurrent trend has seen most other medical disciplines become increasingly narrow in their scope of practice [8]. While these changes to GP specialty designation were made to ensure quality, regulation and training standards of the GP workforce, associated side effects included a slowing in growth of the general practice workforce. Globally, fewer medical school graduates choosing general practice continues to be a major health system concern. Moreover, this trend is most problematic for rural communities, where access to GPs is arguably more critical in the frequent absence of other medical specialties [9, 10].

The need for GPs continues to grow, however. This is partly driven by improved prognoses and treatment options for many diseases, leading to increased management of chronic diseases in the community. More people are living longer, with increased management of multi-comorbidities facilitated by GPs [11, 12]. Additionally, general practice has seen a large growth of provision of mental health services. Yet, junior doctors continue to disproportionately choose sub-specialty careers, reporting the health system and general population better valued, rewarded and recognised narrower specialisations compared with generalist skills or broad expertise [13, 14]. New GP headcounts may not adequately replace those retiring [15, 16]. Newer doctors choosing general practice are also more likely to work fewer or part-time hours, partly related to the increased feminisation of the workforce, younger doctors seeking a better work-life balance and more work variety, rather than traditional consultations alone [17–19].

Understanding how GP specialisation decisions are made has never been more critical. Our recent research within the Australian health system showed that GP career choice is shaped by a complex mix of individual motivations, educational exposures, system‑level influences, and professional culture [20]. However, to date, much of the literature in GP specialisation choice is pragmatic and practice-focused. Importantly, research in this area generally lacks a strong underpinning of existing theories or frameworks relating to the decision-making process of medical students or junior doctors. The identification and use of theories is important to better understand how medical career decisions are made. Absence of theory means research remains descriptive rather than explanatory, thus limiting understanding of why factors matter or how they influence decisions. This scoping review sought to understand the theories and theoretical frameworks used to explain and inform general practice career choice research. Our aim was to scope the evidence on the use of theories and theoretical frameworks in general practice career choice selection research.

Methods

This review was conducted and reported in accordance with the JBI Manual for Evidence Synthesis [21] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Extension for Scoping Reviews (PRISMA-ScR; Supplementary Table 1) [22], respectively. Scoping review methodology was deemed most appropriate for this review as we were interested in broadly exploring factors relating to specialty decisions and aimed to map theories and frameworks reported in the literature rather than to evaluate their quality. The review protocol was not registered due to the limited timeframe to complete this externally funded study, of which this review was one part of. This review utilised a constructivist approach to theory use.

A key inclusion criteria was to limit studies to those from Australia and countries with similar training structures (primarily focused on United Kingdom and New Zealand), whereby entry to specialty training occurs after graduating from medical school. In contrast, literature from contexts where specialty career decisions are mostly made while still in medical school were excluded, such as the USA, Canada and much of Europe.

Search strategy

For this review, four electronic databases (PubMed, Web of Science, Embase, and Scopus) were searched. These databases were chosen as they are extensive and have a focus on medical and health-related literature. The search strategy was designed to only capture peer-reviewed studies which addressed all four topic areas: (1) related to medical students or doctors, (2) addressed general practice career decision making, (3) addressed specialty decisions, and (4) referenced or developed a theory, theoretical framework or conceptual framework. Henceforth, the term framework is predominantly used when referring to the aggregate of theories and frameworks [23, 24], to limit confusion for the reader. The authorship team take the view that a theory is an interacting system of ‘useful and empirically relevant ideas’ usually tested through peer review and studies as a method of understanding real world situations, often for explanation, prediction and control [25]. However, a framework contains named concepts and assumptions that are possibly interrelated and sometimes ‘interdefined’ [26], normally as the first step for building a theory. Thus, several conceptual frameworks may be included in a wider ranging theory.

Keywords for each topic area were originally searched separately, with search terms joined using the Boolean operator ‘OR’, with the exception of ‘dentists’ which was searched with ‘NOT’. The four search strings (topic areas) were then combined using the Boolean operator ‘AND’. Filters and limits were applied during database searching to narrow the search results based on article type, year of publication, and subject area (medicine only; if available). Keywords used to develop the search are provided in Table 1, and the complete search strategy for each database is provided in Supplementary Table 2.

Table 1.

Keywords used to develop the search strategy

Topic	Keywords used
Medical students or doctors	• Medical student* [MeSH] • Medical education [MeSH] • Pre-registrar doctor* [Title/Abstract/Keywords] • Pre-registration doctor* [Title/Abstract/Keywords] • Medical schools [MeSH] • Internship and Residency [MeSH] • Medical graduate* [Title/Abstract/Keywords] • Junior doctor* [Title/Abstract/Keywords] • Junior physician* [MeSH] • Prevocational doctor* [Title/Abstract/Keywords] • Doctor* [Title/Abstract/Keywords] • Physician* [MeSH] • Dentists [MeSH] – searched using ‘NOT’
Career decisions	• Attract* [Title/Abstract/Keywords] • Career pathway* [Title/Abstract/Keywords] • Career aspiration* [Title/Abstract/Keywords] • Career decision* [Title/Abstract/Keywords] • Career choice* [MeSH] • Reason* [Title/Abstract/Keywords] • Interest* [Title/Abstract/Keywords] • Prefer* [Title/Abstract/Keywords] • Recruit* [Title/Abstract/Keywords] • Choos* [Title/Abstract/Keywords] • Choice* [Title/Abstract/Keywords]
Specialty	• Specialt* AND (general* OR ‘primary care) [Title/Abstract/Keywords] • General practice* [MeSH] • General practitioner* [MeSH] • Primary healthcare [MeSH] • Primary health care [Title/Abstract/Keywords] • Primary care [MeSH] • Family physician* [MeSH] • Primary care physician* [MeSH] • Family medicine [Title/Abstract/Keywords] • Family practice [MeSH] • Generalist* [Title/Abstract/Keywords] • GP [Title/Abstract/Keywords]
Theory or framework	• Theor* [Title/Abstract/Keywords] • Theoretical framework [Title/Abstract/Keywords] • Conceptual framework [Title/Abstract/Keywords]

Screening and data extraction

Only original research or literature reviews from included countries, published in English between January 2000 and October 2024, were included for relevance. Following completion of reviewer assessments, the search was updated in May 2026, for recency. The complete lists of inclusion and exclusion criteria are provided in Table 2.

Table 2.

Inclusion and exclusion criteria

Include	Exclude
Medical students, junior doctors, registrars and specialists	Non-medical staff or trainees
Reasons for choosing or not choosing general practice	No reference to career decision-making
Intention and actual specialisation
Australia or countries with a similar medical system (i.e. with pre-registration period between internship and specialty choice; United Kingdom and New Zealand)	All other countries
Original research or literature review/meta-analyses	Conference proceedings, editorials, letters, and commentaries
Published on or after 01/01/2000	Published prior to 2000
Published in English

All search results were exported into Covidence software (https://www.covidence.org/) which facilitated de-duplication, screening, and extraction. First, all titles and abstracts were screened by two reviewers (from FY, JF, MM). Any disagreement regarding article inclusion was discussed by all reviewers; however, the third reviewer (who did not screen the article) was responsible for the final decision if agreement could not be reached. Following title and abstract screening, full-texts were reviewed via the same process. During full-text screening, literature reviews which were not specific to the countries of interest were excluded. Reference lists were also checked and any relevant studies were added to title and abstract screening to be reviewed for relevance.

The data extraction template within Covidence was jointly developed and piloted by all authors, after which two reviewers (FY, JF) each extracted and verified the data. Data extracted from each study included publication details and study information, details relating to frameworks cited in the methods section, and details of any frameworks developed as part of the studies. Epistemology and ontology for each study was classified according to Brown and Duenas [27] by both reviewers to provide an understanding of the research perspectives being used. For studies with no explicitly referenced theory, a description of the theory implicit to the study rationale was extracted.

Although Brown’s theory character roles were originally assigned to the included studies with explicit theories [28], repeated classification by the same researchers resulted in differing results due to the breadth of the character classifications. Since the original rationale for this step was to understand how theories were being utilised within medical education research, theories were instead classified for their assigned tasks and audience stance as per Brown and Duenas [27].

A further search was completed to understand the usefulness and impact of frameworks produced by included studies. Each included article which developed a framework was searched in PubMed using the ‘cited by’ function to determine whether any of the developed frameworks had been utilised in subsequent studies. Citations were counted and analysed for citation use.

Results

The electronic search yielded 4,328 results (PubMed: 1145; Web of Science: 1,564; Embase: 245; Scopus 1,337). An additional 29 articles were identified via excluded literature reviews, and eight from the reference lists of included articles. 1,895 were duplicates, leaving a total of 2,396 for title and abstract screening. After excluding 2,323 articles based on titles and abstracts, 73 full texts were reviewed. After screening full-texts, 43 were excluded, leaving 30 articles for inclusion in the review. The PRISMA diagram showing the screening results and reasons for exclusion is provided in Fig. 1.

Fig. 1.

PRISMA diagram

Details of each of the included studies are provided in Table 3. Of the 30 studies, eight were from Australia, 20 from the UK, one from New Zealand, and one was an international research team. The majority of studies focussed on career decisions during medical school (N = 15), with some investigating training stages from medical school through to fellowship (N = 9). Most studies utilised qualitative methodology (N = 21) or were mixed methods (N = 6). One review [50] was included despite some reference to international literature as they developed a detailed framework for GP specialty decisions which was deemed a high priority for this review.

Table 3.

Description of included studies

Study ID	Country	Training stages	Methodology	Intervention(s)	Data collection instrument(s)	Study outcomes
Alberti et al., 2017 [29]	UK	Prevocational to registrar	Mixed-methods	NA	• Survey with yes/no and free-text questions • Focus groups	• Positive/negative comments about general practice as a career choice
Ashley et al., 2024 [30]	UK	Medical school to registrar	Qualitative	NA	• Telephone and online interviews	• Interviewees’ childhood and education • Why they had chosen to study medicine • Their experiences and aspirations when starting out at medical school • Experiences getting through medical school and moving into foundation year and specialty training, and the sorts of resources they felt are accorded value in medical careers • How their relevant understanding changed over time and whether, how, and why their aspirations changed
Beattie et al., 2024 [31]	Australia	Prevocational to fellowship	Qualitative	Year long Longitudinal Integrade Clerkship (LIC)	• Interviews	• How the rural LIC program influenced graduates’ career decisions in terms of both medical specialty and chosen geographic location
Brooker et al., 2018 [32]	Australia	Medical school	Mixed-methods (Qualitative data)	GP placement in prison health clinics	• Pre and post placement questionnaire • Face-to-face or telephone semi-structured interview	• Knowledge and experience of health issues in the criminal justice system • Views of working with people in prison or a history of incarceration
Chellappah & Garnham, 2014 [33]	UK	Medical school	Mixed-methods	NA	• Survey with Likert scale and free-text questions	• Characteristics of general practice, GPs, and hospital specialties • Influences on specialty decisions
Cole et al., 2020 [34]	UK	Medical school	Mixed-methods	NA	• Online questionnaire	• Views regarding the online curriculum
Cuesta-Briand, 2020 [35]	Australia	Prevocational	Qualitative	NA	• Semi-structured phone interviews	• Work since graduation • Career intentions • Training and support
Edgcumbe et al., 2008 [36]	UK	Medical school	Qualitative	Expanded general practice component during medical school	• Individual, semi-structured interview • Nominal group study	• Views of general practice as a potential career
Egan et al. 2025 [37]	Ireland (UK)	Registrar	Qualitative	NA	• Semi-structured interviews	• Motivators and career priorities of GP registrars
Firth & Wass, 2007 [38]	UK	Medical school	Qualitative	Problem-based curriculum (20% time in GP attachment)	• Semi-structured interviews	• Views and influence of peers, supervisors, healthcare professionals, media, and family regarding general practice • Career aspirations • Views of future placements • Effects of policy and increased focus on primary care
Firth & Wass, 2011 [39]	UK	Prevocational	Qualitative	2-year Foundation Program where 97% of trainees undertake a four-month general practice rotation	• Focus groups • Individual interviews	• Prior general practice experience and attitudes towards the rotation • Benefits and issues relating to the rotation
Harris et al., 2020 [40]	UK	Prevocational	Qualitative	NA	• Semi-structured phone interviews	• Factors influencing career choices
Hogg et al., 2008 [41]	UK	Medical school	Mixed-methods	Final Year General Practice Module	• Two fixed choice questionnaires • Focus groups	• Factors influencing student’s towards or away from general practice • Current career choice and reasons • Influence of the community-based curriculum • Importance of role models
Howe & Ives, 2001 [42]	UK	Medical school	Quantitative	New model for fourth year students comprising eight weeks in a general practice	• Questionnaire	• Career hopes • Causation of illness beliefs • Expectations of community-based teaching • Roles of behavioural science and primary care • Doctor-patient relationship and teamwork
Jones et al., 2005 [43]	UK	Medical school	Qualitative	Intercalated BSc in primary health care	• Semi-structured interviews	• Motivations and expectations of the course • Perceptions of peer’s opinions about the course • Career intentions • Positive/negative aspects of undertaking the course
MacIver et al. 2025 [44]	UK	Fellowship	Qualitative	NA	• Interviews	• Early to mid-career female academic GP career narratives
Merrett et al., 2017 [45]	UK	Prevocational	Qualitative	NA	• Semi-structured focus groups	• Attitudes towards a career in general practice
Neve, 2019 [46]	UK	Medical school to fellow	Other: Narrative review	NA	• NA	• NA
Nicholson et al., 2016 [47]	UK	Medical school	Qualitative	NA	• Focus groups	• Specialty preferences
O’Sullivan et al., 2020 [48]	Australia	Prevocational to fellow	Qualitative	NA	• Semi-structured interviews	• Becoming a generalist or specialist doctor
Parker et al., 2014 [49]	New Zealand	Medical school	Qualitative	NA	• Focus groups	• Student’s and peer’s views of general practice as a career • Experiences which would change opinions of general practice
Pfarrwaller et al., 2017 [50]	International	Medical school	Qualitative Review	NA	• Review • Framework development	• Choosing a primary care or non-primary care specialty at graduation
Reid & Alberti, 2018 [51]	UK	Medical school	Qualitative	NA	• Focus groups	• NA
Robinson & Lefroy, 2022 [52]	UK	Medical school	Realist Review	NA	• Metasynthesis	• Medical students' opinions about general practice careers
Rogers et al., 2009 [53]	Australia	Medical school to fellow	Mixed-methods	NA	• Survey • Semi-structured focus groups	• Factors which provide confidence to choose a specialty and practice location • Perceived advantages/disadvantages of the chosen specialty/practice location • Steps taken towards choosing a specialty/practice location
Sutcliffe et al. 2025 [54]	UK	Medical school	Qualitative	NA	• Interviews	• Student views on GP role modelling and careers after GP placements
Thistlewaite et al., 2008 [55]	Australia	Medical school to fellow	Qualitative	NA	• Interviews	• Factors influencing career choice • Positive and negative influences on general practice career choice
Tolhurst & Stewart, 2005 [56]	Australia	Medical school	Qualitative	NA	• Focus groups	• Attitudes/intent towards future medical disciplines, geographical locations, and practice environments
Vohra et al., 2017 [57]	Australia	Medical school to fellow	Qualitative	NA	• Interviews	• Choice of specialty
Wainwright et al., 2019 [58]	UK	Prevocational	Qualitative	NA	• Semi-structured telephone interviews	• Whether/how banter effected intended specialty/choice of career

Note: UK = United Kingdom; NA = Not applicable; GP = General Practitioner

The frameworks used and produced by included study authors are summarised in Table 4. These draw from 15 distinct frameworks, which are described verbatim by the original framework authors in Supplementary Table 3. Research perspectives were generally not declared. Ontology and epistemology classified by the research team was deemed to be largely critical theory (N = 15). Twenty-two studies produced some form of framework as an outcome within the study (eight from explicit frameworks and 14 from implicit frameworks).

Table 4.

Explicit theories and frameworks within included articles

Theory or framework	Study ID	Ontology	Epistemology	Tasks assigned to theory
Bland-Meurer model for medical student specialty choice	Harris et al., 2020 [40] &	Constructivist	Constructivist	Alignment: study framing was aligned with the Bland-Meurer model of specialty decision-making for educator-practitioners
Bordieu’s formula for theory of practice	Ashley et al., 2024 [30]	Critical theory	Critical theory	Mentor: Bordieu’s formula for theory of practice constructs (habitus, captila, field) were used to interpret interview transcripts, using these as methodological tools to understand how childhood and educational influences affected medical school experiences and goals, and ultimately their specialty decision-making. For policymakers, faculty members and researchers Alignment: Used to focus on individual and societal influences on female GP academics’ careers, including expected norms, power, unseen rules, negotiations and interactions Interpretation: Habitus was defined as how women’s self-perceptions were built, particularly in entering GP academic careers, and the influences of social and cultural capital (e.g. networks, support structures, family, qualifications, education), providing context and rich meaning to career narratives collected Identification of research problem: Used to elicit lived experiences and perceptions of female GP academics For educator-practitioners, faculty members and researchers
	MacIver et al. 2025 [44]	Constructivist	Constructivist
(Adaptation of) Chang’s transactional process of decision making	Vohra et al., 2017 [57]	Critical theory	Critical theory	Alignment: model adapted by authors after literature review, and before data analysis, to explain how different factors are used in GP decision-making; not mentioned in findings or discussion. Model for theorists, paper for policymakers
Constructivist grounded theory	Beattie et al. 2024 [31]	Constructivist	Constructivist	Methodological tool to ensure research question was answered and to develop a conceptual framework around longitudinal integrated clerkship. For researchers Alignment of study reporting with epistemological stance on reflexivity and biases in research, and used within interpretation for creation of meaning during social interactions and settings, including self-presentation and construction of selves
	Sutcliffe et al. 2025 [54]	Critical theory	Constructivist
Hermeneutic phenomenology	Reid & Alberti, 2018 [51]	Critical theory	Critical theory	Interpretation: theoretical approach used to make sense of student focus group data around general practice career perceptions. For educator-practitioners and family members
Kotter’s eight steps for leading change	Cole et al., 2020 [34]	Critical theory	Critical theory	Methodological tool: Used to guide research team planning and design of mixed-methods questionnaire for faculty member audience
Pfarrwaller et al.'s conceptual framework of medical students' primary care career choice	Cuesta-Briand, 2020 [35]	Constructivist	Constructivist	Object of examination: testing the framework for decision-making specialty and rural location choices, through comparison for relevance between findings and Pfarwaller’s framework, for educator-practitioners, faculty members and researchers
The 'propaganda' model of banter and the theory of 'person-environment fit'	Wainwright et al., 2019 [58]	Critical theory	Critical theory	Object of examination: determining whether the theory applies to trainee perceptions of person-environment fit Identification of research problem: framing ‘banter’ as a problem that influences medical trainees’ perception of person-environment fit with primary care Interpretation: to focus the effects of banter in a propaganda lens upon perceptions of person-specialty fit. For educator-practitioners and faculty members
Social cognitive career theory (SCCT)	Brooker et al., 2018 [32]	Constructivist	Constructivist	Interpretation: Used in Brooker et al. 2018 to interpret interview data as to whether placements in prisons influenced medical students’ career choices (p2), reflected in Results headings, for an audience of educator-practitioner, faculty member, and researcher Used in Egan et al. 2025 to interpret how trainees sought to reach career ambitions with self-efficacy, capabilities, environmental and personal influences in mind, for an audience of educator-practitioners Methodological tool: Used in Rogers et al. 2009 to create survey instrument for predicting medical student specialty choice and practice location; as alignment: self-efficacy for career goals, including expectations for their ability to produce desired professional and lifestyle outcomes, are related to specialty and practice location intention (and thus choice). For researcher audience Also used as methodological tool in Egan et al. 2025 to create the interview guide
	Egan et al. 2025 [37]	Post-positivist	Constructivist
	Rogers et al., 2009 [53]	Constructivist	Critical theory
Socialisation theory	Reid & Alberti, 2018 [51]	Critical theory	Critical theory	Alignment with the findings that medical training shapes students’ decision-making processes and perceptions of specialty choice. For educator-practitioners and family members
Threshold concept framework	Neve, 2019 [46]	Critical theory	Critical theory	Object of examination: Furthering application of threshold concepts to primary care education Vehicle: for the idea that primary care education requires different competencies and learnings for medical learners compared to hospital specialties For educator-practitioners and faculty members
Realist evaluation	O’Sullivan et al., 2020 [48]	Critical theory	Critical theory	Methodological tool: Used to guide research design and focus (i.e. refining initial program theory and constructs through data analysis process) for policymakers, faculty members, researchers and theorists, with implications for educator-practitioners
	Robinson & Lefroy, 2022 [52]	Critical theory	Critical theory
Reciprocal determinism in social cognitive theory	O’Sullivan et al., 2020 [48]	Critical theory	Critical theory	Alignment: to support idea of personal agency and environment in shaping specialty decision-making. For educator-practitioner and theorist audiences
Vygotsky’s Sociocultural Theory	Egan et al. 2025 [37]	Post-positivist	Constructivist	Alignment: Used to understand how people learn, collaborate and interact with mentors and leads to professional outcomes and decisions Methodological tool to create interview guide. For educator-practitioners
(Adapted) Wengraf’s biographical narrative interview method	MacIver et al. 2025 [44]	Constructivist	Constructivist	Methodological tool for conducting and interpreting interviews for cohesive career narratives, also providing alignment and identification of the research problem within the context of complex GP academic career experiences for faculty members, researchers, and educator-practitioners

A summary of the main outcomes from included quantitative and mixed methods studies is provided in Supplementary Table 4. Regarding overall qualitative study themes, there was agreement that GP career decision-making was complex, taking into account time before and during medical training [29, 33, 36, 39, 41, 45, 47, 49, 50, 52, 55–57, 59]. This involved incorporating perceptions from multiple stakeholders, including medical educators and role models. It seemed that what individuals were taught about the medical profession and general practice throughout the medical curriculum varied between medical schools, educators and hospital clinicians and shaped how medical trainees thought of different specialties, including GPs [29, 38, 39, 41, 49–52, 55–58, 60]. In keeping with these learnings, individuals seemed to utilise concepts of person-environment fit and effort-reward imbalance to decide on their specialty choices, taking into account their desired personal lifestyles/goals, individual medical interests, GP role models, specialty training programs and associated status/income to decide on the best person-environment fit with the best effort-reward balance for themselves [36, 45, 48–50, 54, 56–58].

Use of frameworks

Findings relating to frameworks used in included studies have been summarised as 1) explicit framework studies where the framework used was explicitly noted by authors, 2) implicit framework studies where the framework used within studies were undeclared, and 3) citation of produced frameworks, pertaining to citations and usage of the frameworks produced as part of the included studies.

Explicit framework studies

The 16 included studies with theories or frameworks that were explicitly acknowledged with relevant citations (i.e. ‘explicit framework’ studies) were identified as being within constructivist (N = 8) [31, 32, 35, 37, 40, 44, 53, 54] or critical research paradigms (N = 8) [30, 34, 46, 48, 51, 52, 57, 58]. In general, ontology and epistemology paradigms were consistent within all these studies, except three studies where the ontology was deemed different to the epistemology [37, 53, 54]. Some studies used the same framework: three used social cognitive career theory [32, 37, 53] and two used Bourdieu’s theory of practice [30, 44]. The remaining studies did not utilise overlapping frameworks. Excluding interview guide or survey creation, methodological approaches used were as follows: two studies each used realist [48, 52] and phenomenonology approaches [44, 51], one study used Wengraf’s biographical narrative interview method [44], and two studies used constructivist grounded theory methodology [31, 54].

Cited theories were generally drawn from vocational psychology [32, 37, 53], organisational theory and educational disciplines; one was derived from successive medical student career decision-making literature [50]. Discussion sections were generally pragmatic in nature, with comparisons within the medical profession between GP and non-GP ‘hospital’ specialties and other healthcare professions within the same country’s healthcare system, or other countries’ systems. Studies identified as being from the critical theory research paradigm, differing from constructivist research paradigm studies by acknowledging differing levels of power and authority within medical, GP and hospital hierarchies. However, they mostly did not collect data upon, acknowledge nor discuss differences between genders and ethnic or socio-economic backgrounds. The exception was Ashley et al. [30], who wrote to a different audience of medical sociologists.

Nine studies used frameworks singularly and three used them in multiple ways. Seven studies used frameworks for alignment to frame study rationale and context. Six studies used frameworks as methodological tools to guide study design and processes. Five studies used them for interpretation or data analysis, three as the object of examination, and two each as vehicle or identification of research problem. The majority of these studies did not declare an audience stance (i.e. target audience) but were more commonly targeted towards educator-practitioners and university faculty members, although some were also written for policymakers, researchers and theorists. Eight studies built upon these explicit frameworks to produce their own frameworks (outcomes), which are further summarised in Table 5.

Table 5.

Citations of frameworks produced as outcomes by included studies with explicit theories

Study ID	Description of frameworks produced as outcome	Subsequent studies using outcome
Cole et al., 2020 [34]	‘As identification with general practice as an independent discipline increases, so does support for a national curriculum of general practice. Conversely, as identification with general practice as an independent discipline diminishes, support for a national curriculum reduces.’	NA
Neve, 2019 [46]	‘The TCF offers us a new way of identifying and exploring those concepts which may be fundamental to thinking and practising effectively in primary care. It reminds us that becoming a primary care professional takes time, involves struggle and significant transformative shifts. The TCF can also help us understand when and how to support, guide or challenge learners.’	4 citations; Informed the study (N = 2) Expanded on the study/findings (N = 1)
O’Sullivan et al., 2020 [48]	Mechanisms to produce a generalist or specialist doctor. For the mechanism ruling in or out	8 citations; Informed the study (N = 2) Expanded on the study/findings (N = 1)
Reid & Alberti, 2018 [51]	Conceptualisation of findings in relation to socialisation theory	12 citations; Informed the study (N = 1)
Robinson & Lefroy, 2022 [52]	Context-Mechanism-Outcome Configurations (CMOCs) identified from meeting with expert informers	2 citations
Rogers et al., 2009 [53]	Social Cognitive Career Theory scale measures	NA
Vohra et al., 2017 [57]	Adaptation of Chang’s decision-making process, for a specialty decision-making framework Key issues that affected the choice of speciality within the context of the present study and their relative strength based on the frequency of utterances	7 citations; Informed the study (N = 1)
Wainwright et al., 2019 [58]	The 'person-fit model' provided a better understanding of how banter informs career choice	7 citations; Informed the study (N = 1)

Note: citation details (subsequent studies using outcome) only relate to subsequent studies having used the framework within their research

Implicit framework studies

Fifteen included studies did not explicitly acknowledge pre-existing theories or frameworks, but provided specific theses (i.e. implicit frameworks) within the medical literature about GP career decision-making, and were thus included in this review. The research paradigms for these studies varied between positivist, post-positivist, constructivist and critical theory paradigms. Many of the studies were deemed to be critical theory due to their acknowledgment of power, authority and hierarchical structures in medical, GP and hospital settings, but generally did not provide illumination around factors like gender, ethnic or differing socio-economic backgrounds. Instead, many of these studies were descriptive, providing in-depth context and insight into medical understanding and experience around GP career decision-making. A summary of these studies and their implicit frameworks are included within Supplementary Table 5.

Citation of produced frameworks

Of the studies which subsequently produced their own explicit theory or framework, the number of citations of these to date (based on PubMed’s ‘cited by’ function) ranged from zero to 35. Studies utilising implicit frameworks were more commonly cited than those utilising explicit frameworks (mean of 10.6 vs 5.0 citations); however, the proportion of studies which actually used the developed framework within their study methods relative to the total number of citations was higher for studies utilising explicit frameworks (22.5% vs 9.5%). Overall, the number of studies using these frameworks in conducting their research was low (range: zero to eight), with most simply citing the studies as a demonstration of past research completed on similar topics.

Discussion

This scoping review highlights some of the complexities relating to medical specialty career choices, particularly for general practice, despite it being the most common specialty within most health systems. Twelve studies built their own adaptation upon explicit frameworks drawn from a range of disciplines, whilst a further 15 studies did not acknowledge a pre-existing framework but continued to produce their own implicit framework. The resulting 27 different frameworks from these studies provide a range of differences and similarities to the key factors relating to uptake of general practice as a career choice within those already training to be a doctor. Overall, there appeared to be no clear standout amongst these frameworks, either in terms of ‘best’ alignment with GP career choice decisions, or subsequent contribution to future research of medical career decision-making.

Instead, understanding how the frameworks were used in GP research was most helpful, whether as a tool for ‘interpretation’ or ‘providing a methodological tool’, ‘object of examination’, ‘identification’ of research problems, or ‘a vehicle for an idea’ or ‘alignment’. We found that classification of frameworks according to ‘character roles’ [28] were less useful in our analysis, as it was not easy to translate their use into implications for our research, nor reconcile them with a face ‘impact’ or usefulness analysis. Instead, classifications according to ‘tasks assigned to theory’ seemed better suited as an analysis method to understand how different theories of GP career decision-making may be useful to the discipline. Mostly, the related audience stance classifications were not explicitly reported upon, but we determined these based on journal titles and Brown’s stated methodology of assigning intended audience based on stated aims or ‘domain of interest’ [28]. Reporting audience stance with these ‘tasks assigned to theory’ classifications (as per Table 4) thus provides a handy overview of relevant theories for future GP research studies on career choice in countries where specialty decision making mostly occurs after medical school.

We classified study ontology, epistemology [27] and theory character roles as per Brown’s references [28]. However, not all studies fit neatly into these. As an example, two of our studies [48, 52] used realist methodology, which are not included within Brown’s medical education classification of ontology and epistemology within general practice research. ‘Realist’ thought is technically one extreme of ontology (i.e. believing reality/truth can be ‘found’ objectively, and subjective matters are secondary to the ‘truth’). Whilst related realist methodologies may be interpretivist/constructivist in nature (i.e. researchers interpret transcripts into context, mechanism, outcomes and consequences), they may treat truth in a positivist way (i.e. reality/truth can be found objectively), forming somewhat of a quandary if the research team is not consistently reflexive or self-aware, or have difficulties identifying and analysing situations with power hierarchies. The realist research approach is closer to pragmatic epistemology, which ‘occupies a middle ground between realist and idealist ontology’, accepting phenomena operate independently of our ideas (i.e. realist ontology) but that understanding such phenomena requires our ideas (i.e. idealist ontology) [61].

In this review, the studies with implicit (i.e. undeclared) frameworks tended to be cited more often: many were singular studies that were descriptive and exploratory in nature, with some producing their own framework outcomes. However, studies utilising an explicit theory could interpret and describe more complex situations in GP career decision-making, providing frameworks that could be built upon by subsequent research teams. It could be that many medical researchers did not have the time or training to apply medical sociology to practice research [62]. As evidenced in Table 4, theories can be used in a variety of ways to enrich the framing and implications of a study, [28] and should be selected during study conceptualisation to ensure their use is not tokenistic. Theories represent the cognitive outcomes of multiple researchers who have focused their attention on related work, ensuring previous research is not repeated – particularly important considering resources for care and research are limited in scope, and system-wide interventions have not often been successful [63]. Instead, we recommend investing time to understand the explanatory power of theories of interest through reading critical evaluations of the theories of interest and the studies, including their methodologies and sample populations, which make up its evidence base. In other words, understanding the theory itself can include understanding what previous work it embodies, what it does not represent, and limitations in use (e.g. population types, methodology, time periods, etc.).

The complexities illustrated by the examples discussed in this section demonstrate that when using theories and frameworks in research, it is invaluable to either have a sociological research mentor, include a qualitative researcher on the team, or upskill members in qualitative research for navigating this complexity. Psychology, sociology, organisational and philosophical disciplines have comprehensive bodies of work to assist in avoiding cognitive pitfalls that practice-based research can find itself within, and provide helpful frameworks for study progress and implications. In resource-limited projects, it may be best practice to include a team member experienced in sociological research for assistance and upskilling other researchers in the team. For another example of a common misunderstanding: although we have included constructivist grounded theory in our list of included theories/frameworks, it is in fact a methodological process used for creating and building theories from collected data, originating within the sociological discipline [64]. Being from a different discipline, this process comes with specific assumptions that correct studies towards a constructivist stance (i.e. there is no such thing as objectivity; all things are subjective) which is appropriate for qualitative studies. However, for many clinicians, such thought comes into conflict with the more dominant positivist or post-postivist nature of clinical evidence-based approaches. This in itself creates pitfalls for the medical or clinical researcher [65], with qualitative mentorship strongly recommended for medical research teams.

Amongst our findings, we seek to highlight the strongly reported effect of professional norms upon general practice pathways, including how GPs, other medical specialists and junior doctors perceive GP careers [40, 58, 66]. Further, we acknowledge the effects of a changing society in the face of globalisation, with its ensuing complexity of work [67, 68] and the need to move away from reductionist attempts of understanding practice research [69]. An increased emphasis on personal wellbeing and resilience of younger generations of doctors may be a healthy change for the medical profession: extensive research is being done to improve junior doctor working conditions, and could have relevance in general practice [70].

Strengths and limitations

This review was conducted rigorously, following the JBI guidelines for conduct and PRISMA-ScR guidelines for reporting. Review team members are from different health professions (FY, JF, PM) and health services research (MM) and collectively have recognsied expertise in qualitative research methodology (FY, PM), quantitative research methodology (JF, MM), medical education (all), and evidnece synthesis (all). Therefore, the team utilised several lenses to arrive at and interpret review findings. Our review purposefully did not include literature from the USA, Canada and much of Europe, though we acknowledge similar important evidence [71–73]. The amount of work done in this space reflects the complexity of this area of research, where multiple stakeholders are involved in the career choice of individuals within medical training that often spans nearly a decade, requiring a systems perspective and longitudinal monitoring for the resultant career outcomes. We contend that context relating to the timing of medical career decisions is a critical factor when assessing the relevance of available evidence. Additionally, we acknowledge the large literature base relating to identifying factors associated with general practice career decisions from studies without a clear underpinning framework [74, 75], and note the existence of a substantial contribution of North American literature in this space. However, given the key career decision-making process differences, it is unclear how findings about the different systems of medical training may intersect.

Our scoping review search strategy was targeted towards finding studies with an explicit (rather than implicit) theory. We included many of the latter, but did not overtly target them in our search terms; hence some may have been missed. There was insufficient quantitative data or homogenous outcome data for meta-analysis of included studies, although included qualitative studies have been briefly summarised in the manner of a meta-synthesis within the results. However, our application of Brown and Duena’s respective classifications to explicit theories provides precedents and implications for future similar reviews.

In an attempt to understand usability or impact of the included studies, we used citations to determine studies which used the frameworks. However, this metric only includes academic literature captured via a single database (PubMed), and does not demonstrate whether the frameworks were being used in more practical contexts (e.g. policy). Furthermore, use of self-citation and the varied publication dates of included studies also means that emphasis should be placed on the proportion of citing researchers ‘using’ the frameworks, rather than citation counts. More appropriate measures of understanding theoretical impact in reviews could also be considered.

We acknowledge that authors of included studies may not have identified with our interpretation of their theories, ontology, and epistemology; however, when utilising theories and frameworks, it is important convention within sociology to declare theoretical framing. This study used a constructivist stance in acknowledging the reporting of each study and context was likely to be subjective, and thereby seeking to identify useful patterns in GP career decision-making research for future research. Similarly, reflexivity and self-awareness of the research paradigm in which the research team works within should be declared, since this provides a quick summary of the assumptions about truth or reality (i.e. ontology) and knowledge (i.e. epistemology) made in the studies, which is often not commensurate with the traditional scientific positivist approach taught in most health sciences [27]. Lack of such declaration could reflect isolation of clinical researchers from complementary research disciplines, where subjectivity (and its effects on human emotion or affect) are recognised to be an essential part of sociological and health services research [61]. Further, whilst we deemed nine studies to be within critical theory paradigms, only Ashley’s study [30] appeared to purposively select this framing: the other studies seemed to happen upon power dynamics and hierarchies due to a level of understanding of practice environments. More in-depth research may be required to explore how power dynamics within medical training affect doctors of differing demographics for GP pathways specifically, and how this affects care within healthcare systems.

Conclusion

Within the context of medical career choices being made in the years following completion of medical school training, our review identified 15 explicitly acknowledged theories or frameworks used to explain general practice career decision-making. Overall, there appeared to be no clear ‘best’ option for alignment with GP career choices or subsequent contribution to future research of medical career decision making, although comprehensive maps of GP career-decision making factors already existed based upon international data. However, theoretically-informed studies tended to have more novel and detailed insights into specific aspects of GP career decision-making processes, and were more likely to build upon previous literature. Studies with no declared theories tended to be cited more but were less informed, with fewer implications for practice. Researchers undertaking future studies around general practice career decisions are encouraged to utilise and build upon existing theories and frameworks, rather than creating new ones, and would be best supported in multidisciplinary teams including an experienced qualitative researcher.

Abbreviations

GP

General Practitioner

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

PRISMA-ScR

PRISMA Extension for Scoping Reviews

Authors’ contributions

MM, FY and PM conceptualised the study; FY and JF confirmed the study design and methodology with support from PM; JF created and executed the search strategy; JF, FY and MM performed data collection and screening; FY and JF led data extraction and analysis; all authors contributed to manuscript drafting and editing; all authors reviewed the manuscript; MM had overall project supervision responsibility.

Funding

This study was externally funded by a Royal Australian College of General Practice (RACGP) Education Research Grant (ERG) grant number ERG2023-07, which supported the postdoctoral researcher’s work on this project.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.