We conducted a systematic review of research priority-setting exercises in HPE. While there is no specific guidance on how to conduct reviews on research priority-setting studies, this systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Page et al., 2021), and the STructured apprOach to the Reporting In healthcare education of Evidence Synthesis (STORIES) statement (Gordon and Gibbs, 2014) to enhance the transparency and quality of reporting. The following research questions guided this review:

The protocol for this review was registered in the PROSPERO database (ID: CRD42023400999).

This review aimed to include any peer-reviewed study that reported a research priority-setting exercise on HPE. Studies were included if they fulfilled the following criteria: (1) the main objective was to develop research priorities; (2) were conducted in the field of health professions education; (3) were conducted in any setting including local, national, regional, or international studies; (4) outline their process for research prioritization; (5) were published after January 2000; and (6) were written in English or Spanish. The year 2000 was selected as the starting date because it coincides with the emergence of key initiatives that formalized HPE research, such as the Best Evidence Medical Education (BEME) collaboration (Harden et al., 2000), and because research priority setting gained traction due to several reports (Global Forum for Health Research, 2000).

Studies were excluded if: (1) they proposed a research agenda or research priority as a secondary objective; we also excluded systematic reviews, guidelines, or opinions; (2) were published in non-peer-reviewed sources; and (3) failed to include the prioritization process.

While the protocol was pre-registered, five modifications were made. These were as follows:

The initial search was conducted on 30th April 2024 in the following databases: PubMed, Scopus, and Web of Science. An updated search was conducted on February 5th, 2026 using the same databases and an expanded search strategy to capture studies published after the original search. The search strategy was initially developed in PubMed and then adapted to other databases using Polyglot.¹

The search strategy for the PubMed database was: (“education, medical”[MeSH Terms] OR “education, dental”[MeSH Terms] OR “education, nursing”[MeSH Terms] OR “education, pharmacy”[MeSH Terms] OR “Health professions education”[Title/Abstract] OR “Healthcare professions education”[Title/Abstract]) AND (“research priorit*”[Title/Abstract] OR “research agenda”[Title/Abstract] OR “research opportunit*”[Title/Abstract] OR “priority setting”[Title/Abstract] OR “priorities setting”[Title/Abstract]). A professional librarian was not involved in the design of the search strategy, which is acknowledged as a limitation of this review.

Retrieved studies were imported into Deduplicator.² Prior to formal appraisal, both reviewers independently appraised a subset of three studies as a calibration exercise to ensure consistent application of the checklist criteria. After duplicate deletion, studies were imported into RayyanAI (Ouzzani et al., 2016), where two of us (FV and JF) independently screened the title and abstract. Then, potentially eligible studies were examined in full text by the main author (JF) and corroborated by one of us (FV). In case of discrepancies in these phases, these were solved through discussion until a consensus was reached.

Three data extraction forms in Google Spreadsheets were developed to extract relevant information. These were designed to fulfill each research question.

The first was for overall data extraction, based on the REPRISE guidelines for research priority-setting studies (Tong et al., 2019). Here, the main author (JF) extracted relevant information such as: author, year, country, objective, context and scope (i.e., geographic scope, area of focus), governance and leadership team, framework employed, characteristics and methods to involve stakeholders, the process to identify and collect research priorities, methods for prioritization of research priorities, the themes and each research priority as well as the numbers, evaluation and feedback, implementation, funding, and conflict of interest. Two of us (BMC and CCC) audited the entirety of the extracted data to ensure accuracy and completeness.

The second data extraction form was for critical appraisal, based on the checklist of nine good common themes for research priority-setting exercises (Viergever et al., 2010). This checklist has been used in previous reviews (Iqbal et al., 2021). This checklist comprises three major domains (preparatory work, deciding on priorities, and after priorities have been set), with nine subdomains (context, use of a comprehensive approach, inclusiveness, information gathering, planning for implementation, criteria, methods for deciding on priorities, evaluation, and transparency), and a total of 20 items. Two of us (CC and JF) independently appraised all studies using this checklist. Prior to formal appraisal, both reviewers independently appraised a subset of three studies as a calibration exercise to ensure consistent application of the checklist criteria. Then, we met to reach a final consensus on the quality of the included studies.

The third data extraction form was used to identify major themes among research priorities (Eady and Moreau, 2024). Here, each research priority for all studies was extracted verbatim. If the study did not report research priorities, the minimum reporting unit was extracted (i.e., if the study did not report research priorities but reported themes, then themes were extracted). The extraction was conducted by the main author (JF) and the entirety of the extraction was audited by one of us (MPS) for accuracy.

As meta-analysis was not feasible for this kind of study, the data were summarized using a narrative synthesis. Key characteristics and quality appraisals of included studies were presented in tabular and narrative formats as described in the protocol.

To summarize research priorities, a thematic analysis was conducted following the methodology recommended by Popay et al. (2006), using a three-step process. First, each research priority was extracted verbatim from the included studies. If a study did not report individual research priorities, the minimum reporting unit was extracted (e.g., themes or topics). Second, each extracted priority was deductively coded against the thirteen core topics of the Medical Education Research Library (Eady and Moreau, 2024), which served as the analytical framework for categorization. This is a list of thirteen core topics of medical education research proposed by editorial board members from the top 10 medical education journals. Examples of topics include assessment, clinical reasoning, curriculum, and continuous professional development. When a priority did not fit an existing topic, it was assigned to an ‘Others’ category. Third, within each topic, subthemes were inductively developed when multiple priorities addressed distinct aspects of the same topic. This process was iterative: the main author (JF) conducted the initial coding, which was reviewed and discussed with the rest of us. Discrepancies in the categorization were resolved through discussion. Subthemes were then quantified by counting the number of studies contributing to each and were presented in tabular format with their number and frequency of appearance.

Table 1 describes the 24 included studies and their key characteristics. Studies were published between 2001 and 2025. Only one study was published before 2010 (Wolf et al., 2001), and the majority (n = 18, 75.0%) were published in the last decade (2016–2025). The years with the most publications were 2020 and 2025, each contributing three studies (12.5%). The country contributing the most studies was the United States (n = 5, Anton et al., 2020; Harper et al., 2012; Huang et al., 2010; Stefanidis et al., 2015, 2022), followed by Iran (n = 2, 8.3%; Nabeiei et al., 2016; Tootoonchi et al., 2012). Notably, six studies (25.0%) were conducted in international or multi-country settings (Ball et al., 2016; Bollela et al., 2025; Chan et al., 2024; Keller et al., 2024; Mullen et al., 2025; Wolf et al., 2001).

All included studies were conducted within the broader field of health professions education. The most common subdiscipline was medical education (n = 12, 50.0%), which encompassed general medical education (n = 8), medical nutrition education (Ball et al., 2016), general practice education (de Groot et al., 2024), precision medicine education (Chan et al., 2024), and medical education regulation (Bollela et al., 2025). Simulation-based education was the focus in five studies (20.8%), spanning surgical (Anton et al., 2020; Stefanidis et al., 2022), emergency medicine (Chaplin et al., 2020), and nursing simulation (Wong et al., 2021). The remaining studies addressed health professions education broadly (n = 2, 8.3%; Palermo et al., 2019; Van Schalkwyk et al., 2020), clinical education (Burford et al., 2025), behavioral sciences in acute care (Keller et al., 2024), prosthetics and orthotics education (Mullen et al., 2025), surgical education (Stefanidis et al., 2015), dental education (Ajjawi et al., 2017), and nursing professional development (Harper et al., 2012).

The Delphi technique was the most common method used (n = 10, 41.7%), either alone or combined with the nominal group technique. Other methods included mixed-methods approaches (n = 3), workshops (n = 2), online surveys (n = 2), the nominal group technique alone (Bollela et al., 2025), Q-methodology (de Groot et al., 2024), the Foresight and Strategy Technique (Chan et al., 2024), systematic prioritization with expert focus groups (Keller et al., 2024), and participatory action research (Mullen et al., 2025). Two studies did not report their method in detail (Burhan, 2022; Chaplin et al., 2020).

Most studies were led by academics (n = 19, 79.2%), while the remaining five were led by professional associations or institutional groups. The number of stakeholders varied considerably: some studies involved small expert panels of 10 to 20 participants (e.g., Stefanidis et al., 2022, n = 10; Wong et al., 2021, n = 11), while others involved larger samples (e.g., Palermo et al., 2019, n = 796; Ajjawi et al., 2017, n = 734; Tootoonchi et al., 2012, n = 451). Two studies did not report the specific number of different stakeholders involved (Burhan, 2022; Huang et al., 2010). The most common participants were health professions educators and researchers. Trainees were involved in seven studies (29.2%; Ali and Sethi, 2021; Ajjawi et al., 2017; Ball et al., 2016; de Groot et al., 2024; Mullen et al., 2025; Palermo et al., 2019; Tootoonchi et al., 2012), and only four studies (16.7%) involved patients or public representatives (Ajjawi et al., 2017; Ball et al., 2016; Burford et al., 2025; Palermo et al., 2019). Only four studies (16.7%) reported the timeframe for their research priorities. Nine studies (37.5%) were funded, eight (33.3%) were unfunded, and seven (29.2%) did not report funding sources.

Table 2 presents the quality appraisal of the 24 included studies according to the nine common themes of good practice for research priority-setting exercises (Viergever et al., 2010), organized under three domains: preparatory work, deciding on priorities, and after priorities have been set.

Table 3, shows how the research priorities were distributed across 14 themes and 46 subthemes. The four most prevalent themes were curriculum (13 subthemes), simulation and technology-enhanced learning (8 subthemes), continuous professional development (4 subthemes), and assessment (4 subthemes).

Curriculum was the most frequently represented theme. The three most common subthemes were curriculum and program evaluation (n = 11, 45.8%), effectiveness of teaching and learning strategies (n = 8, 33.3%), and curriculum to improve communication skills (n = 8, 33.3%). The second most common theme was simulation and technology-enhanced learning comprising eight subthemes. The three most frequent were development and evaluation of technology in health professions education (n = 11, 45.8%), including studies addressing emerging technologies such as precision medicine (Chan et al., 2024) and behavioral sciences applications (Keller et al., 2024); role and frameworks of simulation education (n = 7, 29.2%); and cost-effectiveness of simulation (n = 5, 20.8%). Continuous professional development comprised four subthemes centered on the faculty workforce. The three most frequent were faculty recruitment, motivation, and retention (n = 9, 37.5%), faculty training (n = 8, 33.3%), and faculty evaluation and promotion (n = 7, 29.2%). Assessment comprised four subthemes. The three most common were assessment of competencies and performance assessment (n = 7, 29.2%), effective and useful feedback (n = 6, 25.0%), and assessment validity (n = 4, 16.7%).

Beyond these four major themes, interprofessional and teamwork education was the most frequently reported single subtheme across the review (n = 10, 41.7%). Student admission in undergraduate and graduate education was reported in eight studies (33.3%). Equity, inclusion, and diversity was identified in six studies (25.0%), encompassing perspectives from medical education regulation (Bollela et al., 2025), clinical education (Burford et al., 2025), and precision medicine education (Chan et al., 2024). Workplace-based learning (n = 6, 25.0%), social accountability (n = 5, 20.8%), accreditation, evaluation, and quality education (n = 5, 20.8%), and management and education policy (n = 5, 20.8%) were also represented.

Lastly, we found distinct temporal trends among research priorities. Certain subthemes remained persistent across all periods, including curriculum and program evaluation (n = 11, 45.8%), development and evaluation of technology in health professions education (n = 11, 45.8%), faculty recruitment, motivation, and retention (n = 9, 37.5%), and assessment of competencies and performance assessment (n = 7, 29.2%). Simulation-related subthemes such as cost-effectiveness, translation to workplace settings, and certification were largely absent before 2015 and concentrated in the most recent decade, driven by the surgical and emergency medicine simulation literature (Anton et al., 2020; Chaplin et al., 2020; Stefanidis et al., 2015, 2022; Wong et al., 2021). The most notable trend was the emergence of priorities reflecting broader societal concerns in the 2020–2025 period: equity, inclusion, and diversity grew from a single study before 2020 (Amini et al., 2012) to six studies overall (25.0%); wellbeing and safety culture appeared as an entirely new subtheme (Burford et al., 2025; Keller et al., 2024); and social accountability re-emerged after being absent in the 2015–2019 period (Bollela et al., 2025; de Groot et al., 2024).

This systematic review identified 24 research priority-setting exercises in health professions education conducted between 2001 and 2025. Our major findings were as follows: (1) most studies were conducted in the last decade within the field of medical education and in regional to national settings, with an underrepresentation of patients and trainees as stakeholders; (2) studies published before 2024 showed limited adherence to good practices in research priority setting, whereas more recent studies demonstrated consistently higher methodological quality; and (3) common threads among research priorities were identified across studies, with curriculum, simulation and technology-enhanced learning, continuous professional development, and assessment emerging as the most prevalent themes, while recent studies reflected a growing emphasis on equity, wellbeing, and social accountability.

Our findings show a growing research output in priority-setting exercises in HPE, with three-quarters of included studies published in the last decade. This growth, however, differs between regions. Studies from the Global North primarily focused on healthcare simulation (Anton et al., 2020; Chaplin et al., 2020; Stefanidis et al., 2022; Wong et al., 2021), while those from the Global South addressed general medical education (Ali and Sethi, 2021; Burhan, 2022; Van Schalkwyk et al., 2020). Notably, the updated search identified a higher proportion of international or multi-country studies (Bollela et al., 2025; Chan et al., 2024; Keller et al., 2024; Mullen et al., 2025), which may suggest a shift toward broader collaborative efforts in the field. Despite these regional differences in focus, the thematic analysis revealed shared values across studies, as discussed below.

The identified research priorities aligned with the topics proposed in the Medical Education Research Library (Eady and Moreau, 2024). Curriculum, assessment, faculty development, and simulation and technology-enhanced learning were the most prevalent themes. However, other topics such as clinical reasoning, interprofessional education, and workplace-based learning were less frequently proposed as priorities. A plausible explanation is that these topics may function as cross-cutting themes rather than discrete priority areas. Their lower frequency as standalone priorities may not reflect lesser importance but rather their trans-topical nature, they are implicitly embedded across multiple priority domains simultaneously, and may suggest that fixed taxonomies may underestimate the prominence of inherently cross-cutting themes when used to categorize interconnected research priorities.

Clinical reasoning, for instance, has a well-established and specialized research community that may pursue its agenda outside of formal priority-setting exercises (Yazdani and Hoseini Abardeh, 2019). Interprofessional education, while reported in over 40% of the included studies, may be underrepresented as a standalone priority because it is frequently embedded within curriculum or simulation subthemes. Workplace-based learning may be underreported as a distinct priority because it intersects with clinical education and assessment subthemes, making its boundaries as a category less distinct. Additionally, we identified research topics not included in the Medical Education Research Library, such as quality assurance, management, education policy, and accreditation. Some of these relate to the “wicked problems” proposed in by Mennin (2021), and which reflects current priorities for the scholarly community.

The thematic analysis revealed common threads and shared values across studies conducted in different countries, disciplines, and settings. Despite the diversity of contexts, the recurring identification of priorities such as curriculum evaluation, faculty development, and assessment of competencies suggests that the HPE community shares fundamental concerns about the quality of education and its workforce. At the same time, the emphasis on specific subthemes varied across contexts: simulation-related priorities were concentrated in studies from higher-resource settings, while studies from sub-Saharan Africa and the Eastern Mediterranean Region emphasized community-based curricula and social accountability. These findings suggest that while a core set of shared values exists, their expression and prioritization are shaped by disciplinary, cultural, and resource contexts. Therefore, efforts to develop global research priorities should build on these common threads while remaining responsive to local needs and specificities.

The temporal analysis further revealed how priorities evolve. Persistent themes such as curriculum evaluation and faculty development reflect enduring structural challenges in HPE. The emergence of equity, inclusion, and diversity; wellbeing and safety culture; and social accountability in the most recent period (2020–2025) mirrors broader societal shifts, including the impact of the COVID-19 pandemic and growing attention to health workforce sustainability (Burford et al., 2025; Keller et al., 2024). Conversely, the decline in subthemes such as the effectiveness of teaching and learning strategies and community-based curriculum may reflect either a perceived saturation of evidence in these areas or a shift in research attention toward newer concerns.

Regarding methodological quality, our review found that studies published before 2024 showed limited adherence to good practices. Among the original 18 studies, only one (Ball et al., 2016) reported using a comprehensive approach, and none planned for evaluating or updating their priorities. This pattern is consistent with findings from a scoping review of 416 research priority-setting exercises, in which only 10% of studies reported using a framework (Lund et al., 2022). However, the six studies identified in the updated search (2024–2025) demonstrated markedly higher quality across all domains, particularly in the use of comprehensive approaches and explicit scoring criteria. This improvement may reflect the increasing availability of reporting guidelines such as REPRISE (Tong et al., 2019) and a growing awareness of methodological standards in the field (Cole, 2025; Kitto et al., 2025). Nevertheless, no study, regardless of publication date, reported a plan for evaluating or updating their priorities, which remains a critical gap.

The underrepresentation of trainees and patients in these exercises warrants attention. Only seven studies (29.2%) involved trainees, and four (16.7%) involved patients or public representatives. This is consistent with findings in other areas of HPE, where patients are frequently underrepresented in competency framework development (Lepre et al., 2021) and medical education research (Emery et al., 2022). In that global survey, the primary reason for involving patients was to improve the relevance of research, a rationale that applies directly to priority-setting exercises. The inclusion of patients and trainees could enrich the identification of priorities that are more aligned with the needs of end-users and beneficiaries of health professions education.

Three main implications emerge from this review, each grounded in specific findings. First, the limited adherence to good practices in research priority setting, particularly the absence of comprehensive approaches, explicit criteria, and evaluation plans, highlights the need for specific guidance for conducting these exercises in HPE. This could take the form of an AMEE guide or a consensus statement that adapts existing frameworks such as REPRISE and the nine common themes checklist (Tong et al., 2019; Viergever et al., 2010) to the HPE context, addressing the unique challenges of involving diverse health professions, multiple educational settings, and both educators and learners as stakeholders.

Second, the identification of common threads and shared values across studies conducted in diverse contexts supports the development of a set of global research priorities for HPE. Rather than imposing a single agenda, such an effort would aim to establish a core list of research priorities that can be adapted to regional, national, or institutional contexts. The persistent themes identified in this review, curriculum evaluation, faculty development, assessment of competencies, and technology in education, could serve as a foundation for this endeavor, complemented by the emerging priorities that reflect contemporary societal concerns.

Third, the underrepresentation of patients and trainees as stakeholders calls for deliberate strategies to include these groups in future exercises. Guidance on how to meaningfully involve patients, trainees, and other underrepresented stakeholders in research priority-setting would strengthen both the process and the relevance of the resulting priorities.

This review has several strengths. It follows a registered protocol (PROSPERO: CRD42023400999), uses established reporting guidelines (PRISMA and STORIES), and employs a validated checklist for critical appraisal (Viergever et al., 2010). Data extraction was guided by the REPRISE guidelines (Tong et al., 2019). The updated search extended the coverage to January 2026, incorporating six additional studies. Forward and backward citation searching was used to identify further studies.

However, several limitations should be acknowledged, including deviations from the original protocol. First, the search was limited to three databases (PubMed, Scopus, and Web of Science), and PsycINFO, CINAHL, and EMBASE were not searched. While previous evidence suggests that the contribution of PsycINFO and CINAHL to systematic review yield may be limited (Hartling et al., 2016), PsycINFO has expanded its indexing of education-related content in recent years, and the exclusion of these databases may have resulted in missed studies. Similarly, adding EMBASE to PubMed has been estimated to increase coverage by approximately 6.85% (Frandsen et al., 2021). Second, no professional librarian was involved in the development of the search strategy, and key terms such as “clinical education,” “allied health education,” “Delphi technique,” “consensus statement,” and relevant MeSH terms were not included. This may have reduced the sensitivity of the search. Third, the search was limited to studies published in English or Spanish and to peer-reviewed sources, potentially excluding priority-setting exercises published in other languages or in grey literature. Fourth, while the protocol was pre-registered, deviations were made regarding the databases searched, the timeframe, the research questions, and the approach to thematic analysis, as documented in the Methods section. Lastly, the thematic analysis was conducted primarily by a single author, and while reviewed by a second author, the alignment of priorities with the Medical Education Research Library topics involved subjective judgment that may have influenced the categorization.