A scoping review methodology was selected “to ‘map’ relevant literature in the field of interest,” following the framework proposed by Arksey and O’Malley [24]. The protocol was publicly registered on Open Science Framework [25]. Results are reported in line with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) [26]. The PRISMA-ScR checklist is included as Multimedia Appendix 1.

We defined our research question as “What is known in the existing literature about the barriers and corresponding enablers to developing, producing, and maintaining open access medical education platforms containing material from multiple sources?”

Searches were conducted on April 17, 2023, using PubMed and EBSCOhost, which includes Academic Search Complete, ERIC, and CINAHL Plus with Full text. The search string used was ((resources) OR (material)) AND ((((medical education) OR (training)) AND (((((identify) OR (evaluate)) OR (integrate)) OR (compile)) OR (collect))) AND (open-access)). The SnowGlobe tool [27] was then used for citation chaining (snowballing), automatically retrieving references of included studies as well as studies citing them. Manual reference searching additionally contributed a number of papers. Search results were imported into Covidence (Veritas Health Innovation), and duplicates were removed. Gray literature, such as institutional websites, preprints, and reports, was not searched.

Studies were eligible for inclusion if they described open access medical education or training platforms that curate or compile content from multiple sources or institutions for use by health care providers. To be included, studies also needed to provide information on factors that facilitate or hinder the design, production, implementation, or ongoing operation of such platforms. All paper types, including reviews, were considered. No date restrictions were applied, given the emerging nature of this field. Studies were excluded if they focused solely on a single course or on materials developed by a single institution. Conference abstracts and non-English language studies were also excluded.

All identified papers underwent a 2-stage screening process. In stage 1, a total of 2 authors (AAEA and AB) independently reviewed the title and abstract of each paper against the predefined inclusion and exclusion criteria. Discrepancies were resolved with the input of EPOF. In stage 2, the full texts of potentially relevant papers were assessed for final inclusion by AAEA and AB, with EPOF adjudicating on discrepancies.

Beginning with an a priori framework of “barriers” and “enablers,” the data chart form was then developed through an inductive process, with input from all authors. Following initial immersion in the data, coding for each barrier and enabler was developed. Data from included papers were then independently extracted by AAEA and AB into this form. Any disagreements were resolved through discussion or with the input of EPOF. Finally, the coding for barriers and enablers was reviewed, and minor coding edits were made. To derive practical recommendations for the various stakeholder groups involved in the production, management, and use of open access educational platforms, barriers and enablers were then mapped to the appropriate stakeholder group: (1) learners and training programs, (2) content designers and creators, and (3) platform managers.

Thematic analysis was undertaken. Once authors were familiar with the data, and coding agreed, themes were then proposed, explained, and clarified through iterative team discussion. Both barriers and enablers were then grouped under these themes, and the themes were reviewed for appropriateness. Descriptive statistics were used to analyze characteristics of the included studies.

The review involved analysis of publicly available secondary data only. No human or animal participants were involved; therefore, ethics approval was not required.

Our search strategy identified a total of 1108 records from multiple sources. PubMed contributed 502 records, followed by 253 from EBSCOhost databases (including Academic Search Complete, ERIC, and CINAHL Plus with Full Text). Snowball sampling added 332 records, and 21 were identified through manual reference searching. After removing duplicates, 1064 records were screened by title and abstract. Of these, 64 full-text papers were assessed for eligibility.

A total of 29 full-text studies were excluded because, although they addressed open access content, they did not examine barriers or enablers or did not relate to platform-level considerations. Ultimately, 34 studies were included in the final review. The study selection process is shown in Figure 1.

All 34 studies described educational resources targeted at doctors or medical students. Nurses were also targeted in 14 (41%) studies, and allied health professionals in 11 (32%) studies. Most studies did not explicitly focus on a particular geography. Some focused on LMICs [2,28], while others were explicitly “global” [29] or “international” [30] in their scope. However, most first authors were based in high-income countries, with a majority from the United States. Other first authors were from Canada, Germany, Belgium, Italy, the United Arab Emirates, the United Kingdom, Australia, Botswana, Papua New Guinea, and South Africa. In total, 16 (47%) studies were categorized as reviews or expert opinion, 11 (32%) were primarily qualitative research, 5 (15%) were primarily quantitative research, and 2 (6%) were reports on educational innovations. While all included studies discussed open access medical education platforms, this was rarely their primary focus. Most included studies focused on e-learning in medical education in general, or in specific medical education and training contexts, or specific topics and specialties, or for specific cohorts of learners. Other studies again focused on the FOAM education movement. Characteristics of the included studies are given in Table 1.

Across the 34 included studies, we identified 127 instances of barriers and 122 instances of enablers. These were coded into 22 unique barriers and 33 unique enablers, and then, mapped to 1 of 3 stakeholder-based themes. A full list of the coded barriers and enablers, and the frequency with which they were identified, is presented in Tables 2 and 3.

As has already been seen in the specialties of emergency medicine and critical care [21], it seems likely that open access medical education resources in other areas and specialties are becoming increasingly consolidated in a smaller number of larger online platforms. Such compilations of open access medical resources have the potential to improve the learner experience by offering a “1-stop shop,” where learners are confident that they can find easy-to-use, quality-assured, up-to-date, comprehensive learning resources appropriate for their context.

Our review finds numerous barriers reported to the design, production, and operation of such platforms. The literature also shows that there is much that open access content designers, authors, and platform administrators can do to address these barriers; and indeed, there is much that can be done by training programs and learners themselves. We can reasonably expect that some of these barriers will diminish in severity over time. Some of the practical barriers identified, particularly in resource-limited settings [2], will partially be addressed by improving internet connectivity, while growing learner familiarity with interactive online learning methodologies can be expected to partially address some of the cultural and motivational barriers.

The 3 most cited barriers in this review were a perceived lack of content quality control, incompleteness of content, and concerns that the demand on human and financial resources required to maintain and update content are unsustainable.

While informal peer and expert recommendations of resources may be helpful, neither learners nor experts are consistently able to make an accurate “gut feeling” appraisal of the quality of online educational resources [60]; therefore, informal recommendation of resources alone may not suffice as a quality control measure. Objective, validated content quality evaluation tools such as the “Approved Instructional Resources” tool [45] or the revised “Medical Education Translational Resources: Impact and Quality” tool [59] should be routinely used by content creators and open access platform curators.

Where e-learning content is targeted at students or trainees in formal training programs, such content should be integrated with learners’ other educational activities, rather than existing apart from them. The connection between content and curricula should be made clear and explicit through a mapping of content to appropriate established curricula. Training programs should identify appropriate open access resources, which fulfill the requirements of their curricula and should guide their students and trainees toward them. The curation of content into platforms, and the linking of content and curricula, will make clear which training needs are met and which are not met by currently available open access content. Identification of content area gaps will reduce duplication of effort and allow “holes” [44] in open access medical education content to be filled by appropriate content.

The barrier presented by the human and financial resources required to maintain and update content is the least amenable to the solutions found in this review. Enablers addressing the sustainability of human resources, such as paying content creators, may negatively impact the financial sustainability of such platforms. Cost and effort may be contained to some degree by the enablers identified in this review; however, creating, hosting, and updating high-quality content will continue to cost money and require significant effort. Hosting content on curated platforms, such as the United Nations Global Surgery Learning Hub [61] and OpenCriticalCare [62], which are free for both content providers and learners, may address some of the issues associated with long-term hosting of content as well as partially address issues of quality control and trust. Ultimately, the value of material and platforms will have to be acknowledged for the associated costs to be met.

This review differs from prior published work by focusing on open access e-learning platforms compiling content from multiple sources, whereas previously published literature has predominantly focused on specific courses and cohorts. Many of the barriers and enablers identified relating to platforms have previously been found in studies of individual courses, and meta-analyses relating to e-learning in health sciences in general [23] and e-learning in medical education in LMICs [15].

These include learner motivation and familiarity with e-learning, incompleteness and inappropriateness of content, and challenges in maintaining and updating content. In LMICs, basic challenges in getting online were highlighted [15]. Enablers related to all of these barriers were similar to those found in this study. Some challenges related to courses are not applicable to platforms. Barteit et al [15] found a preponderance of pilot or small-scale e-learning interventions for resource-limited settings that never produced content at scale. De facto compilations of medical education contain significant volumes of content.

This study found the financial sustainability of open access medical education compilations, and the sustainability of the level of effort required, to be challenges with few straightforward solutions. This concern is echoed throughout the broader literature on open access medical education. Lee et al [52] calculated the median overall value of a FOAM website to be US $22,815 and concluded that there is “substantial value being generated from these resources and ... this should be recognized by academia.” Platform costs will either have to be met by funding from academic institutions and funding bodies or by charging fees—to content providers, to learners in high-income settings, to training programs, or for learners to access “premium” content. Lin et al [21] found that the majority of the 109 FOAM sites that they analyzed generated income, with 18% embedding advertisements; 27.5% asking for donations or payment for merchandise, continuing medical education credits, books, or web-based courses; and 44% being affiliated with a sponsoring institution, such as a professional society, hospital, or journal. Lee et al [52] draw parallels between the FOAM movement and the open access publication movement, noting that in making publications available open access, publishers shifted the cost from the content consumer to the content producer. They note that grant funding mechanisms exist to fund these publication costs and call for similar support for medical education content creation. Mncube and Mthethwa [49] quote an unnamed academic, “When we write and publish the research articles or book chapters we are incentivized, however in the development of [open access resources], there are no incentives.” Where it is not possible to fund content creators, and until more sustainable models are developed, greater acknowledgment and recognition of content creators and others involved in this work may be a simple way to help sustain their motivation.

Two important barriers found relating to compilations of medical education resources, that were either not discussed or not given prominence in the literature on individual courses, are issues of trust and language. It may simply be that analyses at a course level exclude potential learners who either do not trust the content or are not comfortable learning in the course language of instruction—they would never take the courses in the first place. We found that managers of such platforms aspiring to make content sourced from multiple different institutions available to a global audience must give much higher priority to issues of trust and language than content creators may have previously done. We found several means by which platform managers can build trust—most prominently, the transparent use of objective, validated content quality evaluation tools. Translation of content into appropriate languages was found to be a key enabler.

In addition to mapping existing literature, future empirical research should include systematic surveys of open access platform websites and interviews with their developers to provide deeper insight into their design, governance, and real-world implementation. There is a notable lack of longitudinal research evaluating the lasting impact of open access medical education platforms. Additionally, limited evidence exists on the structured integration of these resources into formal medical training programs. Sustainable models—particularly those detailing financial and operational strategies for the long-term maintenance of such platforms—are also rarely published.

This review identifies from the literature actions that each stakeholder group can take to improve open access medical platforms and increase their impact. Recommendations are summarized in Table 4.

Only studies published in English were included, which may introduce language bias and limit the representation of perspectives from non-English–speaking regions. Furthermore, the majority of included studies originated from high-income countries, potentially narrowing the cultural and economic contexts considered. We recognize the possibility of funding bias influencing the evidence base. The exclusion of gray literature may also have led to the omission of relevant nonindexed reports or institutional documents. Changes in medical education technology are fast-paced, and the literature necessarily lags behind changes in learner behavior. Nevertheless, this review establishes a foundation for future research, implementation, and policy development in the field of open access medical education.

Open access medical education has the potential to add significant value to preservice and in-service medical education worldwide. However, numerous challenges are preventing open access medical education platforms from achieving their potential, particularly a lack of content quality control, incompleteness of content, and the challenge of sustaining the necessary human and financial resources.

Ultimately, if open access medical education is to achieve its potential, much greater coordination and collaboration on a global scale is required—between learners, educational institutions, content creators, and platform managers.