Synthesis of comparisons of multiple interventions in infectious diseases research: a scoping review

Abstract Background Evidence synthesis in infectious diseases research often involves comparing multiple interventions to assess relative efficacy or safety, usually through pairwise or network meta-analyses. However, the statistical methods used and whether they are applied appropriately in practice are frequently unclear. This scoping review provides an overview of how comparisons of multiple interventions are synthesised in infectious diseases research. Methods A systematic search was conducted in PubMed, Embase, Web of Science, and the Cochrane Database of Systematic Reviews using search terms related to meta-analyses and infectious diseases. Systematic reviews were included if they quantitatively compared three or more interventions for the prevention or treatment of infectious diseases. Results Of the 8160 records identified, 420 systematic reviews were eligible after full text screening. A random sample of 200 systematic reviews was selected for data extraction: 100 each from 2015–2019 and 2020–2024, to ensure representation of pre- and post-COVID infectious diseases. Systematic reviews of interventions for COVID-19 (n = 32), HIV/AIDS (n = 22), and hepatitis B (n = 22) were most common. Therapeutic interventions were assessed in 153 (76.5%) systematic reviews and preventive treatments in 47 (23.5%). For statistical comparison of interventions, 34 (17.0%) systematic reviews used only pairwise meta-analyses, 88 (44.0%) used only network meta-analyses, and 78 (39.0%) used both. Of the 166 network meta-analyses, 83 (50.0%) were implemented using a Bayesian framework. Only 41 (24.7%) assessed transitivity, that is, whether it is appropriate to combine evidence from direct and indirect comparisons of interventions, and 112 (67.5%) evaluated consistency, which quantifies the agreement between the effect estimates from direct and indirect comparisons. Conclusion Network meta-analysis is widely adopted in infectious diseases research. However, the key assumption of transitivity is rarely assessed. Improved adherence to publicly accessible protocols, thorough evaluation of assumptions, and transparent reporting is needed to improve the validity and reproducibility of evidence synthesis in infectious diseases research.