The reporting of ethical review and ethical considerations in articles published in medical education journals: A literature review

Health professions education research (HPE‐R) must be ethically sound. Ethical review decisions and other ethical considerations should be clearly reported in journal articles to allow readers to assess the ethical soundness of the research. We explored and evaluated how ethical review decisions and ethical considerations for HPE‐R are reported in health professions education (HPE) journal articles.


| BACKGROUND
The health professions education (HPE) research (HPE-R) community has shared responsibility for ensuring that all performed studies and all published research that involves human participants is ethically sound. In recent decades, HPE journals have adjusted their expectations, policies and requirements for researchers submitting their manuscripts. [1][2][3][4] Currently, all HPE-R must be in accordance with the Helsinki Declaration, and authors are required to seek ethical review or explicitly justify why exemption is appropriate. 1,2,4,5 It stands to reason that ethical review decisions and ethical considerations should be clearly reported in journal articles to allow readers to assess the ethical soundness of the research.
Ethical review decisions include whether an article was submitted for ethical review board (ERB) assessment, which ERB reviewed the research, and what the board's decision was. These points are important because major international and even regional differences exist in how ethical reviews are organised. Moreover, different decision terminology is used to describe the practice of ethical reviewing practice and ERB decisions. [6][7][8][9][10] For example, the meaning of the ethical review decision 'exempt' varies considerably among countries and could encompass an expedited review or no review at all. 8,10,11 Projects that involve HPE-R or quality improvement (QI), without involving patients, often fall outside the scope of general medical ethical rules and regulations in some countries and can be exempted without substantive review by general medical ERBs. 8,10,11 Previous evaluations of the reporting of ethical review decisions revealed that in two major education journals, fewer than 2% of the articles published between 1988 and 1989 and fewer than 10% of the articles published between 1998 and 1999 were reviewed by an ERB. 9 In 2013, the practice of seeking ethical review for articles published in four major medical education journals was evaluated by Hally and Walsh. 11 The authors found that 16% of articles failed to mention an ethics review or stated that ethics approval was not required, without providing further explanation, or that national law did not require ethical review for medical education research. 11  In addition to ethical review outcomes, ethical considerations should be assessable by readers of journal articles. Ethical considerations refer to how authors assess and plan to mitigate the ethical risks to participants, to safeguard participant autonomy. This includes protecting people from psychological stress caused by experienced pressure or coercion to participate. Other risks may arise through discomfort during study activities and procedures, as well as privacy risks. 9,[12][13][14] Privacy risks could include the publication and dissemination of identifiable and sensitive participant information. [7][8][9][11][12][13][14] Roberts et al. assessed the reporting of some of these features and considerations for research articles published in 1988-1989 and 1998-1999 in two major HPE journals. 9 The results revealed sufficient reporting in fewer than 25% of articles. Stovel  All articles were retrieved from the journals' websites. First, we screened all article titles, meta-data and abstracts for eligibility, based on our predefined inclusion and exclusion criteria. If there was any doubt, we read the full article for screening purposes. Articles were included if they (i) were published in one of the included journals within the predefined time frame and (ii) reported on outcomes that were based on research using data derived from human participants.
Articles were excluded if they (i) were published outside the predefined period (2020/01/01 to 2020/12/31), or (ii) were published within the predefined period only as 'ahead-of-print' or similar, or (iii) did not use primary data obtained from individual participants (e.g. reviews), or (iv) used solely publicly available data. We did not limit our search to original research studies. Letters to the editor, perspectives and innovation reports may also report on participant evaluations and use primary data derived from human participants.

| Data extraction
Data were extracted by one researcher (TS, MEF or AW) using a coding sheet for the items (Table 1) Table 1 shows the extracted items.
Regarding the research articles with IMRaD or similar structure, we read the abstract, methods and declarations sections, which usually appear at the end of an article. These are the intended places for reporting ethical review processes and ethical considerations in IMRaD-structured articles. For the other articles, we read the full texts to extract the variables. In addition to the coding sheet, we kept a reflective journal to note illustrative examples, especially regarding researchers' reflective statements about dealing with ethical considerations. The points of interest included power differentials, consent procedures and incentives to participate.

| Data analysis
Extracted data were imported to an IBM SPSS Version 28.0 (IBM Corp., Armonk, NY, USA) database. Descriptive statistics were calculated for the individual variables.

| Ethical considerations
As this is a review study and there were no human participants, we did not seek ethical approval. In the results section, we provide examples of attentive and careful research practice. We choose not to list articles failing to report aspects of ethical considerations, because the goal of this article is not to identify individual researchers. However, we provide anonymised quotes and illustrative examples.

| Search results
From among 2004 articles published in 2020 in the specified HPE journals, 955 articles (47.7%) were eligible for further review. Figure 1 shows a flowchart of the search, selection and review process. In a random selection, a second author reviewed the articles, and there was 96% inclusion concordance and 91% overall concordance.   The most frequently used methodologies were surveys (n = 340), mixed methods (n = 230), interviews (n = 129) and tests or assessments (n = 129). The mixed methods studies usually combined observations or document analysis with interviews or focus groups, or combinations of surveys, interviews and focus groups.

| Ethical review processes and outcomes
The majority (83.4%) of the articles mentioned a review by any ERB, institutional review board (IRB), research ethics board (REB), regional ethics board or specific ethics board for social sciences or HPE-R. Figure 2 and Appendix S1 and Appendix S2 display the ethical review processes and outcomes for the different categories of articles. Overall, more than half of the articles mentioned approval for the reported study, and 18% reported an exempt decision. Among the research articles, 7.7% did not mention the outcome of an ethical review process; for other articles, 67.6% did not mention the outcome (a nearly 10-fold rate).
Reported ethics assessment was mostly performed by general institutional or regional ERBs and in a few studies by national ERBs.
Specialised ERB assessment for HPE was described for one specific board in the Netherlands.
In 24 research articles, the study could not be classified as reporting an approved or exempt status. Either there was inconsistency in reporting the outcome of an ethical review, such as reporting the outcome as 'exempt' in one place and elsewhere as

| Key ethical considerations
The key ethical considerations assessed are described in the following section and presented in Figure 2 and Appendix S1 and Appendix S2, stratified for the different methods and article types. Additional examples of well-described considerations are provided in Appendix S3.
Adherence to the Declaration of Helsinki was mentioned in a few research articles (9.5%) and in even fewer other articles. In addition to the outcomes of ethical review in HPE research, specific ethical considerations-such as informed consent and incentives-are increasingly reported in research articles (Figure 3).
We found a higher rate of reporting incentives (15%) than did Sto- Interestingly, we noted that some articles that reported on focus group interviews stated that data collection was anonymous; however, the interpersonal nature of interviews precludes that possibility. This discrepancy suggests that researchers may not always be fully aware of the ethical consequences of certain research practices.
Apart from research articles, we also reviewed 'other' articles, such as letters to the editor, perspectives and innovation reportswhich regularly report on QI projects. The requirements and our HPE-R community's expectations regarding ethical considerations and review board assessment for QI projects are less clearly defined than they are for research articles. 5,24 In a recent survey of HPE journal editors, almost half the editors stated that their journals required ERB assessment for all submitted QI projects, versus 87% for research projects. 5 This discrepancy suggests that different standards exist for research articles and QI projects. However, in another study, editors emphasised that innovation reports are 'not a path for publishing suboptimally conducted research'. 25 Possible reasons for not seeking ethical approval-from a researcher perspective-include that ethical considerations and procedures can be perceived as detrimental in designing, performing and describing innovative research or QI projects. 13,24 Moreover, seeking ethical approval is generally only possible for prospective evaluations.
The latter point may hinder the reporting of retrospective evaluation of QI projects, despite these projects having specific merits.
In practice, the conceptual difference between an innovation reported as a QI project or formally evaluated and described in the format of a research paper is quite theoretical. Indeed, the difference between QI evaluations and research articles seems to be a continuum. Moreover, in some educational case reports, 26 ethical considerations are described with better rigour than in research articles.
Therefore, if a researcher's intention is to prospectively evaluate a QI project, IRB or ERB review and approval would be required. 27  In this review, we aimed to provide insight about the practices of reporting ethical considerations in HPE-R. Limitations to our study included the journals we selected for inclusion. The HPE community has many journals, and the journals we selected had high impact ratings in the HPE field. Hence, there was potential selection bias, since articles in high-impact journals may contain particularly thorough descriptions. Articles with less well-described methods or unreported ethical considerations are relatively likely to be rejected and submitted to lower ranking journals that were not included in our review.
In addition, the studied journals mainly publish research from the Global North. 28 Hence, reporting of ethical review and considerations in research from countries in the Global South remains underexposed in HPE-R. 28 Another plausible confounder in our research is the formal ethical reporting requirements among journals and potential differences in the requirements for research articles and other articles.
We do not know what the ethical reporting requirements were for the selected journals at the time of submission of the manuscripts we reviewed.
Another limitation was that for IMRaD-structured articles, our data extraction was limited to specific sections of the articles. Therefore, we could have missed ethical considerations reported in other sections (i.e. introduction or discussion section), although these sections are not typically used to report such considerations.
F I G U R E 3 Ethical considerations and reported outcomes of ERB assessments of research articles in medical education/health professions education journals over time. The data points are the year(s) wherefore the data are collected and not of publication of the article (e.g. Roberts et al report on two periods (1988-1989 and 1998-1999) and the present paper (Schutte) report on articles in 2020). *The referenced data points originate from Roberts et al., 9 Stovel et al., 15 Hally and Walsh 11 and the present paper (Schutte) and entail the percentages of reported ethical considerations and outcomes of the research articles (other articles are here omitted, for comparability). In the study of Stovel et al., 15 8% explicitly reported the use of incentives.
We demonstrated high concordance among our assessors in a random selection; this result was comparable with previous research on this subject, which achieved particularly high concordance. 9 The discrepancies among our assessors were almost all due to interpretation and the difficulties of using a dichotomised scale, namely 'reported' versus 'not reported'.
The strengths of this research include the sample of assessed articles, which was larger than in previous studies, 9,11,15,22  In this article, we underline the importance of ethical reviews.