Feasibility of a Network Meta-Analysis Assessing Relative Effectiveness and Safety of Remdesivir Versus Other Treatments for COVID-19 Inpatients

OBJECTIVES: Remdesivir is an intravenous ribonucleic acid polymerase inhibitor that targets viral replication of SARS-CoV-2 and is used in the treatment of hospitalized COVID-19 patients. There is limited head-to-head evidence for remdesivir versus other treatments for COVID-19 inpatients, therefore a feasibility assessment (FA) for a network meta-analysis (NMA) was undertaken.

METHODS: A systematic literature review (SLR) was conducted to identify randomized controlled trials (RCTs) and observational studies involving searches of databases, conference literature and trial registries up to March 2023. As part of the FA, study, treatment and patient characteristics were assessed for heterogeneity. Outcomes of interest were all-cause mortality, length of hospital stay, proportion of patients recovered from COVID-19, proportion of patients discharged from hospital and number of patients experiencing at least one treatment-emergent adverse event at 2 and 4 weeks. Network connectivity was assessed separately for mild-moderate, mixed or severe COVID-19 inpatients.

RESULTS: The SLR identified 135 publications reporting on 74 unique studies on patients hospitalized with COVID-19. Heterogeneity and network connectivity were assessed for 23 comparative studies (12 RCTs, 11 observational) connecting to remdesivir for outcomes of interest. While networks connected via standard of care (SoC) for some outcomes, due to differences in geographic region, study start date, and local treatment guidelines, substantial heterogeneity was observed in SoC received and disease severity (oxygen supplementation levels). Where reported, substantial differences were also seen in age, race, body mass index, and smoker status. Additionally, comparisons with remdesivir were mostly among treatments that are not used in clinical settings to treat patients of the same severity. Findings were similar in networks containing only observational studies.

CONCLUSIONS: Networks connected via SoC consisted of treatments indicated for patients of differing severity levels, limiting clinical relevance of available comparisons. Furthermore, due to substantial heterogeneity, it was concluded that NMAs would not be appropriate.