Most systematic reviews on medical education present critically low quality

Article type
Authors
Martimbianco A1, Delbone R2, Sá K3, do Nascimento Y4, Yarak S4, Riera R5
1Hospital Sírio-libanês, Sao Paulo, Sao Paulo, Brazil; Universidade Metropolitana de Santos (Unimes), Santos, Sao Paulo, Brazil
2Universidade Federal de Minas Gerais , Belo Horizonte, Minas Gerais, Brazil
3Universidade Metropolitana de Santos (Unimes), Santos, Sao Paulo, Brazil
4Universidade Federal de São Paulo (Unifesp), Sao Paulo, Sao Paulo, Brazil
5Hospital Sírio-libanês, Sao Paulo, Sao Paulo, Brazil; Universidade Federal de São Paulo (Unifesp), Sao Paulo, Sao Paulo, Brazil
Abstract
"Background: In medical education, systematic reviews have occupied a growing niche in the literature and are a powerful tool to inform decision-making in the educational environment. However, how these reviews were planned, conducted, and reported is still unclear.

Objective: To assess the methodological quality of systematic reviews on medical education.

Methods: This meta-research study considered systematic reviews on medical education available in MEDLINE (via PubMed) in the last ten years. The following outcomes were assessed: (i) characteristics of systematic reviews, (ii) frequency and type of methodological tool and/or guidance used for conducting the systematic review, (iii) adequacy to the AMSTAR-2 items, (iv) relation between the methodological tool/guide used and the adequacy of the AMSTAR-2, and (v) frequency of systematic reviews that referred the use of PRISMA statement for their reporting.

Results: 48 systematic reviews published between 2004 and 2023 were included; 87.5% had an interventional question of interest, and the outcomes assessed were predominantly knowledge, attitude, skills, and behavior. According to the AMSTAR-2 tool assessment, 93.8% of the systematic reviews were classified as critically low quality; 45.8% erroneously used the PRISMA as methodological guidance and 18.8% used this tool as a reporting guideline as it was designed (Figure 1). Eight reviews (16.6%) followed the methodological recommendations of Cochrane or BEME, and three of them were classified as low or moderate quality. The mean percentage of adequacy to the AMSTAR 2 items was higher among reviews that reported the use of an appropriate methodological guide than among those that did not report this use or that used PRISMA for this purpose (57.62% versus 33.62%, MD 23.50, 95% CI 11.86-35.13, p = 0.0008).

Conclusion: Considering that most of the assessed systematic reviews are of critically low methodological quality, there is evident a need for guidance and training on the methodological process of a systematic review in the medical education area.

Relevance and importance to patients: These findings enhance the methodological quality of evidence supporting teaching strategies in medical education. This improvement can lead to innovations in medical training by providing more reliable methods for educating future healthcare professionals.
"