The ongoing COVID-19 pandemic has underlined the importance of well-trained health care workers (HCWs) in a functioning health care system (1, 2). However, there is a growing global concern regarding the availability of HCWs, as the World Health Organization (WHO) predicts a shortfall of 18 million HCWs by 2030, most of whom will be missing in low-income and middle-income countries (LMICs), especially in rural areas (1, 3). In particular, these countries are faced with a triple burden of disease through the maintained burden of infectious diseases, as well as non-communicable diseases and health impacts through climate change which are likely to negatively affect weak health systems even more (4). Inadequate resources and infrastructure, such as training facilities, educators, and financing for continued professional development (CPD), all contribute to the shortage of HCW in LMICs (5–7). Scaling up and increasing access to quality HCW education is critical to increasing the number of qualified HCWs needed to provide sufficient health care in LMICs.

Over the past two decades, Massive Open Online Courses (MOOCs) have emerged as a mode of electronic learning (e-learning) (8, 9), fostered by steadily increasing internet access (10). MOOCs are courses that can be taken by a theoretically unlimited number of participants (massive), without cost or formal admission (open), that are accessible via the internet (online) whilst following a structured learning plan (course) (11). Due to the capacity of MOOCs to simultaneously reach a large number of learners regardless of geographical location, their open-access nature, and their attraction to students from diverse educational backgrounds, MOOCs may provide a means to increase access to quality education for HCWs (12–14). Since the introduction of the MOOC concept in 2008 (15), interest has increased over the years, reflected in rising course offerings and a continuous high volume of learners (16). MOOCs are now offered by over 950 institutions including many prestigious universities and accessed by over 180 million learners from all parts of the world (16). Overall, MOOCs cover a wide range of topics and in 2020, 7.7% of all available MOOCs covered the field of health and medicine (16).

Several case studies have reported on the successful use of MOOCs for HCWs' education specifically in LMICs. For example, the World Health Organization (WHO) uses the openWHO platform to provide MOOCs that aim to quickly spread knowledge about responses to emerging diseases such as the Plague, Diphtheria, Ebola (17), and COVID-19 (18) to a large number of people. Evaluations on the reach of these courses show high uptake within outbreak countries (17, 18). Similarly, the Latin American Nephrology community has developed a MOOC on kidney immunopathology, targeting all HCWs involved in kidney transplantation in Latin America and concluded that MOOCs are a potential tool for health workers' professional development and reduce heterogeneity in their access to training resources (19). Following a review on the global use of MOOCs for nurses and other health care workers, Longhini et al. (20) concluded that massive open online education may be suitable when face-to-face education is not possible or to reach a broad audience in a short period of time, supporting the notion that MOOCs may be used in HCWs' education to support existing structures by making education more readily available (21). In contrast, Rowe et al. (22) concluded that there is insufficient evidence to advocate their usage for educating medical students. In addition to these findings, we found limitations in the current literature on differences in MOOC use across high- and low- and middle-income countries, as no reviews have differentiated between the use of MOOCs in high and low-and middle income countries. Facilitators and barriers of MOOCs and their capacity to address the educational needs of HCWs in LMIC settings remain unclear. Therefore, to map the state of evidence regarding how MOOCs are used for HCWs' education in these countries and to address the question of the extent to which MOOCs can support the education of HCWs there, we limited the scope of this review to MOOCs accessed by HCWs in LMICs to understand:

Due to the broad scope of the research questions and the aim to include all types of studies, a scoping literature review seemed most adequate (23). This scoping review was underpinned by the methodological framework developed by Arksey and O'Malley (19) and advanced by Levac et al. (24). Accordingly, the review underwent five stages, iteratively: (1) Identify the research question, (2) Identify relevant studies, (3) Select relevant studies, (4) Chart the data, (5) Collect, summarize, and report results (19). The review is reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis – Scoping Review (PRISMA-ScR) (25) framework.

The systematic search of the databases returned 1,813 articles. Following title and abstract screening and removal of duplicates, 332 studies were included in the full-text review; 39 met the inclusion criteria and were included in the scoping review analysis. The main reasons for exclusion were learner location not reported or not clearly stated, MOOC not targeted at HCWs, and publications that did not include original data (Figure 1). The 39 included studies reported on 40 different MOOCs, a summary of the extracted data is available in Supplementary Appendix 2.

We found 40 MOOCs relevant to HCWs education that were accessed by learners in over 90 LMICs. and although overall feedback of study participants was positive, some regions and countries such as Central Africa were not represented, and key causes of diseases were not addressed. Most importantly we found that we were often unable to distinguish between outcomes by cohort due to inconsistent reporting of MOOC characteristics which restricted summarizing of results and conclusions.

MOOCs covered a wide range of health-related content over several health disciplines, with the diagnosis and treatment of diseases, particularly infectious diseases, being the most prevalent and their use in disseminating information to HCWs as part of an epidemic or pandemic response being notable. Communication and patient interaction learning aims were incorporated in a few MOOCs, which has been found to be lacking in medical education in some African countries (77). However, we removed articles on other topics such as palliative care (78). Other reviews on MOOCs for HCWs education have retrieved more studies (13, 20) indicating that many courses are not accessed by learners from LMICs. Furthermore, other relevant topics such as cardiovascular disease, maternal and neonatal diseases, or HIV (79) were not yet covered by included MOOCs, suggesting a possible discrepancy that should be explored. In the case of HIV, significant learning needs for HCWs have been reported (80) and it would be of interest why MOOCs are not used to fill this learning gap. Such research may also expand understanding on the limitations on the use of MOOCs for HCWs in LMICs. In addition to covering a broad range of topics, included MOOCs also targeted a broad range of HCWs, and in many cases were not focused on HCWs specifically. This could imply that MOOCs enable interprofessional education and potentially encourage interprofessional cooperation and communication, which could lead to improved patient care (81).

The learning objectives of included MOOCs centered mostly on remembering and understanding, referring to low levels of Bloom's taxonomy (73). This seems to indicate that MOOCs appear to be better suited to encourage knowledge transfer and comprehension (72). Furthermore, it may be that some MOOCs “provide more sound pedagogy that develops higher order thinking, whereas others do not” (82). Bali (82) also argues that the benefit of MOOCs may lie in the engagement between learners made possible through the course. This notion is supported by our finding that some learners reported the exchange with other learners to be particularly helpful (70, 71). Two studies implemented a “flexible pathway” of their MOOC within a blended learning environment, in which the MOOCs taught on theoretical aspects and in-person workshops were used to practice application of the learned material (64) or discuss possibilities for local implementation (52). Blended learning as a teaching and learning format is showing promising results in medical education regarding learner satisfaction and learning gain (83). Another engaging, interactive method are “virtual patients,” which train clinical decision-making skills on simulated patient cases (54, 55). More research is needed to better understand how MOOCs might be used to support skills training for HCWs in LMICs, possibly in combination with other digital learning tools such as virtual reality scenarios using head-mounted devices (84).

We found that the literature does not distinguish facilitators and barriers by learner cohort but rather reports on generic barriers reported by all included learners. Nonetheless, because the majority of included studies reporting on facilitators and barriers were targeted at specific regions and only included learners from LMICs, we can be sure that those identified are experienced by learners from LMICs. However, this means that the identified barriers are likely biased toward MOOC implementations that have taken particular precautions to facilitate uptake in LMICs. Future research should therefore focus on regional barriers experienced by MOOCs with a global audience. Despite these limitations, we found that the facilitators and barriers we identified for learners in LMICs can be compared to those reported elsewhere. Time and internet connection were the main barriers to learning with a MOOC (75, 85). Time constraints related to competing obligations around work, family life, and learning were concerns for some learners (36, 40, 52). Deadlines for assignments (43, 51) and teaching and learning modes such as blended learning approaches (52, 67) were mentioned as effective strategies for overcoming these challenges. Goal setting and contextual structuring, such as deadlines and timeframes, are likely to assist online self-regulated learning (OSRL). This is in line with previous research that has found that OSRL through goal setting is a frequently used strategy in MOOC learning and that students with high OSRL report higher satisfaction (86, 87). Limited internet connection is one of the most widely cited issues related to MOOCs and e-learning (20, 75), and are particularly relevant in the context of LMICs. Internet access continues to be a central issue, particularly in African countries where access is often still comparably expensive (10) and may be a reason for the lack of utilization in central sub-Saharan African countries. Therefore, it is essential to assure that HCWs in LMICs have access to online training materials, either by providing alternative internet access options or by providing course content offline as well, so learners can synchronize content once and then learn offline. The importance of developing an interface that is intuitive for users is highlighted by a study finding that digital literacy was a barrier whilst functional and easy to handle instructional design was identified as a facilitator (65). No barriers in terms of the content being irrelevant or too difficult were mentioned by HCWs, though nurses in an interprofessional MOOC mentioned that not all assessed content applied to them (36). On the other hand, local relevance and integration into a pan- or epidemic response were reported as factors that promote MOOC uptake. This is unsurprising and in line with the Technology Acceptance Model (88), which postulates that technology must have local relevance to be adopted. Relatedly, government and employer endorsement were concluded to be facilitators which is in line with previous research identifying lack of perceived employer support as a barrier to uptake (76). In addition, local promotion may increase awareness for MOOCs amongst potential learners leading to increased uptake, as awareness is a significant factor and has been identified as a key reason for not learning by MOOC (85).

Finally, although the language people learn in greatly impacts learning ease (89), language was not discussed much within the included articles. Further research on the extent to which language is a factor in learning through MOOCs may be useful (90) to support the choices in language options made by MOOC developers.

E-learning has been introduced in many African countries to improve HCW education (91). The fact that MOOCs are currently readily available, typically free, and often appropriate for use in diverse countries makes them a strong contender for inclusion in HCW strengthening strategies. However, the findings of our review suggest that simply the availability of MOOCs is not sufficient, rather to achieve optimal effect, they should be integrated into existing educational structures and endorsed by stakeholders and top management. For example, MOOCs could be formally included in curricula as a partial requirement, this would strengthen MOOCs' credibility in the respective countries as a serious learning and teaching tool (91). Overall, it is of note that only 11 of the included studies reported on facilitators and barriers faced by learners; also these facilitators were not formally assessed or evaluated in the included studies. Instead, they were deduced by respective authors based on experience. Although this generates valuable insights, formal research in this area would be necessary to make these findings more credible. Although barriers are more formally investigated through post course surveys, studies rarely use empirical approaches to assess barriers. A better understanding of factors that support online MOOC learners in LMICs is needed to identify how MOOCs may align with national education and training strategies for HCWs' education.

The evaluations were overwhelmingly positive and though only a few reported on barriers to taking the MOOC, there were no reports of the MOOC failing. This may be due to a response bias of learners with a positive experience being more likely to complete the evaluation (92) or publication bias of interventions evaluated as successful (93). A key finding within the evaluations is that only one study (68) reported on cohort specific evaluations indicating that learners from developing countries experienced a higher benefit from the MOOC compared to learners from developed countries. Thus the available data does not allow us to conclude whether learners in LMICs achieve similar positive outcomes as the average learners. As learners from HICs often make out the greater share, it is possible that greater usefulness or even unsuccessful participation for a sub-cohort was missed. Furthermore, included evaluations, focused mainly on short term impacts, such as learner satisfaction and knowledge gain. However, as theoretical knowledge and understanding of a procedure are not always sufficient to ensure changes in clinical practice and patient care (90, 91), these evaluations do not allow for deductions of MOOCs effectiveness in improving the quality of care. Nor do they show whether gained understanding can be retained over time.

We identified two key factors in the available literature that limited the extent to which conclusions can be drawn and summarized for our target cohort learners in LMICs. Firstly, only few articles described outcomes such as facilitators and barriers or evaluation criteria by cohort. This is likely a result of the inclusive nature of MOOCs, however, by only reporting on overall findings, conclusions for specific cohorts cannot be drawn making it difficult to suggest this teaching method to specific cohorts. Secondly, we found large heterogeneity in what MOOC characteristics the included articles reported on which impacts the extent to which conclusions can be drawn and generalized. For example, learning objectives were only reported for 25% of the MOOCs; the geographical location of learners was reported by 28% of included studies, and the delivery language was reported for 45% of MOOCs. The development and use of reporting guidelines for MOOC-based research, or research on e-learning in general, could foster progress in this field. Based on this review, the reporting guidelines for mobile health interventions (94) and the guidelines for reporting evidence-based practice education and teaching (95), we suggest that evaluations should include basic characteristics of the MOOC such as the institution or company who developed the MOOC, the teaching language including subtitles, the target audience, and learning objectives (also in respect to Bloom's taxonomy), as well as basic demographic characteristics of all learners to enable synthesis research to summaries which populations may benefit most from which characteristics. Finally, reports should include details on their evaluation (method, duration, objective, sample size, outcomes) and success criteria (rationale, evaluation methods of measuring success) where appropriate.

We found that MOOCs relevant to HCWs education are accessed by learners from over 90 LMICs and that these are evaluated positively in terms of learner satisfaction and learning gain for all learners in these courses. Insufficient internet connectivity and time constraints were the most commonly reported barriers, whilst support from the government and employer as well as promotion through academic sources were identified as facilitators. Included MOOCs covered a range of topics, with management and treatment of infectious diseases being the most common. However, we also identified two key limitations within the literature. Firstly, the large amount of heterogeneity between the characteristics that are reported on limits our analysis. Secondly, as only few studies reported on cohort specific findings, we were limited in the extent to which we could draw cohort specific conclusions. Therefore, we suggest the development and application of reporting guidelines for future studies.

Finally, we identified gaps in the literature including a lack of research on the actual impact of MOOCs on HCWs' practice and the subsequent benefits for patients and a mismatch between the topics covered by MOOCs and the largest burden of diseases present in LMICs.

Our findings suggest that MOOCs constitute a supplementary tool to strengthen the quality and coverage of HCW education in LMICs. While current MOOCs primarily focus on lower levels of learning (remembering and understanding) future MOOCs could strive to build on this with higher levels of learning (apply, analyze, evaluate, create).

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

The objective and research question were developed by JN, PN, and SB. The first round of data screening was conducted by JN and PN. The second round was conducted by JN. Data extraction, analysis, and interpretation were conducted predominantly by JN with support from PN and SB. JN wrote the manuscript draft. SB, PN, and RS all contributed to further manuscript revision. The final version of the manuscript was approved by all authors.

This study was part of a research unit funded by the German Research Foundation (DFG). The DFG supported this research but was not involved in study design, collection, management, analysis or interpretation of data, neither in the writing of this report or in any decision to submit this report for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.