This systematic review is situated within a larger project that aims to identify possible mechanisms that perpetuate the current imbalanced gender profile in academic medicine and dentistry and also aims to identify and appraise existing interventions that address these issues, as outlined within the introduction to this review.
The present review protocol has been registered within the Open Science Framework (registration number: osf.io/mfy7a) and is being reported in accordance with the reporting guidance provided in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) statement [17] (Additional file 1).
Search and information sources
The following relevant database will be searched for studies: MEDLINE (including MEDLINE Epub Ahead of Print, MEDLINE In-Process & Other Non-Indexed Citations, and MEDLINE Daily), Cochrane Controlled Register of Trials (CENTRAL), PsycINFO, and Education Resource Information Center (ERIC) database. The search strategy will include subject headings and free-text terms for clinical academics. A date limit will be applied to the strategy to restrict retrieval to studies published from 2004 onwards, reflecting the era of the Athena SWAN initiative. The full search strategy for Ovid MEDLINE can be found in Additional file 2. This strategy will be translated to run appropriately on the other databases.
As preliminary searches using the strategy in Additional file 2 retrieved a very high number of potentially eligible records, we developed a narrower search strategy on the same databases in an attempt to limit the number of hits. This strategy is included in Additional file 3. The results from this narrower search (which our pilot work found contains a higher proportion of potentially eligible records) as well as 1000 random records from the main search will be used to ‘train’ the text mining algorithm in the reference management software, Rayyan (further details provided in the ‘Data management and selection process’ section).
Reference lists of relevant systematic reviews and included articles will be reviewed, and forward citation searches of key papers will be undertaken. Authors of relevant studies may be contacted as time allows to seek further studies. We will contact the project funders to request any relevant reports or other work within their portfolio. Published and unpublished studies will be sought and no study design restrictions applied. A time limit for eliciting further studies of 3 months will be applied to ensure that the results of the review are available to inform further aspects of the overarching multi-methods research.
Inclusion and exclusion criteria
Studies will be included in the review if they meet the following criteria:
Population
The study population will include doctors, dentists, and/or those with a supervisory role over their careers (e.g. programme directors, deans). Studies which include mixed groups of professionals will only be included if the doctor/dentist group is reported separately, or if they comprise more than 50% of the participants. Studies of qualified doctors and dentists of all specialties and at all levels of career are eligible for inclusion. Those with academic careers can be at any level from pre-doctoral to professor. The review expressly does not include medical and dental students, though future work may wish to explore the various influences on those at such an early career phase. Studies which explore why doctors and dentists have chosen not to undertake a clinical academic career or why they no longer have a clinical academic career (when they previously were following one) will be eligible for inclusion. For the purpose of this review, an academic career refers to those engaged in research, not purely teaching or educational roles.
While there are pathways that offer a clinical academic career to nurses, midwives, and other allied health professionals, this review will focus exclusively on doctors and dentists, consistent with the needs of the funders of the research. This also reflects that the main pathways for clinical academic careers in the UK (funded by the NIHR) separate doctors and dentists from other healthcare professionals.
Given that the funders of the review and the main dissemination targets for the findings are based within the UK, we have mainly searched for the British terms for clinical and academic career pathways. We have not expressly searched for American terminology, or those from other countries; however, if identified by the search, these studies would be eligible for inclusion.
Topics of interest
-
Factors influencing recruitment and retention to clinical academic careers, including barriers and facilitators. This may include but is not limited to funding, training opportunities, cultural aspects, barriers experienced by underrepresented minorities, issues related to academics with young families, and experiences surrounding roles models.
-
Interventions to increase recruitment to clinical academic careers and to improve retention in clinical academic careers. These may include, but are not limited to, specific funding opportunities, training opportunities, development programmes, mentorship programmes, and strategies which specifically aim to increase academic engagement of specific groups, e.g. family-friendly strategies aiming to increase the involvement of women in clinical academia.
-
Where multiple barriers, facilitators, and interventions are described within and across studies, each will be extracted and included for analysis within the review.
Outcomes
Outcomes will be study defined, but related to success rates of joining or continuing within a clinical academic career, including but not limited to success in gaining funding support, proportion of time spent in academic work, and numbers of awards/higher education qualifications, as well as experiences of professionals within the clinical academic pathway.
Study design
Studies will be included from all forms of quantitative and qualitative research provided they inform the research objectives. This will include but not be limited to:
Quantitative research: randomised controlled trials (RCTs); including quasi-RCTs and cluster-RCTs; observational cohort studies (prospective and retrospective); and studies reporting survey data will be eligible for inclusion within the review.
Qualitative research: methodologies including ethnography, phenomenology, and grounded theory. Studies that use qualitative methods but which do not state an explicit methodology are also eligible to be included, provided they present qualitative data. This includes, but is not limited to, studies using focus group discussions, interview studies, and observational studies. Similarly, mixed-methods studies are eligible for inclusion if they provided sufficient data.
Studies will be limited to those written in the English language for two reasons. Firstly, these are most likely to reflect the cultural experiences of the group in which we plan to apply the results, that is clinical academics in the UK. Secondly, the benefit of qualitative research is to allow participants to express their experiences and perceptions, the clarity of which could be lost through translation and thus the results of the synthesis could become inaccurate. Furthermore, studies will be limited to those performed in high-income countries, in recognition of the cultural and organisational setting in which the research findings are to be applied.
Studies will be included where they are available in full-text format. Conference abstracts will not be eligible for inclusion. Editorials, letters, and opinion pieces will not be eligible for inclusion.
Data management and selection process
References will be managed in EndNote X9 [18] and exported into Rayyan [19] for study selection. Initially, we will import records identified from the narrower secondary search strategy which we expect to contain a higher proportion of potentially eligible records. Following this initial training of the software, we will import 1000 random unique records from the wider search (see Additional file 2) into Rayyan. This batch of records will also be screened in its entirety. These titles and abstracts will be screened independently and in duplicate by a core team of reviewers who will liaise closely to ensure consistency in eligibility decisions. This process will ‘train’ the text mining algorithm within Rayyan to recognise and prioritise the most relevant records.
As a final step, we will import all further unique records from the broader search strategy (see Additional file 2). The text-mining algorithm will then automatically prioritise the most relevant records and bring them to the top of the list. To manage workload, at this point, we will bring in reviewers from the wider project team to support the screening process. Once there is good agreement between all reviewers, records will be single screened. We will keep track of the rate of records marked for inclusion for each set of 1000 search results. At least 25% of identified titles and abstracts in the broader search strategy will be assessed. However, if the rate of screening includes has not fallen dramatically from baseline at this point, then we will continue until there is team agreement that the rate of includes has fallen sufficiently. We will further explore the similarity index in Rayyan to ensure that no relevant titles and abstracts have been missed.
Disagreements regarding which studies to include will be resolved by consensus or, if this proves impossible, by recourse to another team member. A similar approach will be taken to screening full texts.
Data extraction and study quality
Data will be extracted by one researcher using a standardised data extraction form and will be independently checked by a second researcher. The information to be extracted is given in detail in Additional file 4. Broadly, this will include general information, detailed study information, participant details, and outcomes. Qualitative data from research reports will be coded by one researcher and reviewed by other members of the research group. The quality of studies will be assessed using the Cochrane risk of bias tool for RCTs [20], the Newcastle-Ottawa tool for non-randomised studies [21], and the QARI tool for qualitative studies [22]. The components of this quality assessment will be presented in both narrative and tabular form.
Analysis
Key study characteristics and outcome data will be summarised in narrative and tabular form. An overview of the literature base, including any significant gaps in the current understanding of the issues, will be provided. In the first instance, we will analyse quantitative evidence, i.e. from RCTs and any non-randomised studies, and qualitative evidence separately.
Quantitative analyses
Where appropriate, we will combine quantitative data in meta-analyses but we anticipate that we will not have sufficient data to conduct meaningful statistical analyses. Therefore, we will narratively synthesise quantitative evidence on interventions that address facilitators and barriers to clinical academic careers, following suitable techniques outlined in the CRD guidance [23]. We will synthesise data at individual, departmental, and organisational levels, paying particular attention to gender, ethnicity, clinical specialty, primary vs secondary care setting, and academic field (e.g. laboratory-based research, clinical trials, systematic reviews, other research methodologies). If appropriate, we will address these factors in formal subgroup analyses.
Similarly, sensitivity analyses will be performed where appropriate, including but not limited to, location of study, risk of bias, conference abstract vs full-text articles, and era of publication. Heterogeneity in any quantitative analyses will be explored both narratively and statistically (using χ2 tests, the I2, and tau2 statistics and by visual inspection of the forest plots). The risk of publication bias will be explored if there are ≥ 5 comparative studies reporting the same outcome using contour-enhanced funnel plots and Harbord and Peters tests [24].
Qualitative (narrative) analyses
A framework analysis will be performed which will allow for the integration of findings across the different components of the project, providing triangulation and further understanding of the research project [25]. The qualitative synthesis will be led by one researcher and reviewed with other researchers. Again, analyses will focus on the influence and impact of factors such as gender, ethnicity, clinical specialty, and academic field (e.g. laboratory-based research, clinical trials, systematic reviews, other research methodologies). The conceptual contribution of each study will be explored in relation to the final synthesis. We will also examine the literature base to establish how it is conceptually organised and to investigate whether there is any dominance regarding geography, professional interest, and theoretical standpoints.
Combined synthesis
Following individual analyses of quantitative and qualitative evidence, we will draw the two components of the review together to allow comparisons between the different findings and informing further exploration to provide depth to the review. This is a key stage in the overarching project as these findings will inform the primary qualitative research and will be conducted and reviewed in close collaboration of the entire project team. The report will detail the various aspects of the review and the literature development of constructs within this process. The strength of the whole body of evidence will be assessed narratively, taking into account the various aspects of the review, alongside the risk of bias findings.
Dissemination plan
Any substantial amendments to this protocol will be documented on the Open Science Framework page for this project (https://osf.io/mfy7a).
The results of this systematic review will crucially inform the development of qualitative research as part of the overarching project. As such, findings will be disseminated to qualitative research participants as appropriate to inform their taking part in the interviews and audio-diaries.
Reports will be provided to the funders at the half-way point and upon completion of the work. The final report will also be published. The study team will be working closely with the UK Clinical Academic Training Forum (CATF) and the study funders to ensure that the findings are communicated to those involved in the clinical academic career pathway and its development.
In addition, the systematic review will be submitted for publication in a scientific journal reported according to PRISMA guidelines [26]. We will also submit our findings to relevant conferences for oral or poster presentations (e.g. the Association for the Study of Medical Education (ASME), the Association for Medical Education Europe (AMEE) Annual Meeting). The findings of the systematic review will inform any outputs from the overarching project, including oral presentations, workshops, and seminars.
To increase the accessibility of our work to a wider audience, we will produce blogs/podcasts and maintain an active social media profile ‘Gender Inequalities in Clinical Academic Careers’ (@GenderClinical), sharing findings of the review and their relevance within the dissemination strategy of the overarching project. We will liaise with established initiatives such as Women Speakers in Healthcare (WSH) and the Medical Women’s Federation (MWF), seek collaboration with existing Athena SWAN/Equality & Diversity activity locally, and aim to increase exposure via contribution to high impact (social) media outputs (e.g. Guilty Feminist podcast, The Conversation, BMJ Opinion).