Understanding Study Designs: Do You Know a Case Series From a Single-Arm Trial?

OBJECTIVES: Single-arm studies are increasingly encountered by researchers conducting intervention reviews. They include observational single-arm studies, commonly called ‘case series’, and interventional single-arm studies, commonly called ‘single-arm trials’ (SATs). We sought to evaluate how review authors understand these study designs and how they treat them in intervention reviews.

METHODS: We conducted a pragmatic literature review for systematic reviews (SRs) of pharmacological interventions published in 2023 to establish whether single-arm study designs are correctly classified by reviewers, what risk of bias (RoB) tools are used for their assessment, and how such studies are incorporated into meta-analyses. A single reviewer identified SRs that included single-arm studies and extracted information on the numbers of included SATs and case series, and any RoB tools used. Possible misclassifications of single-arm studies by review authors were identified.

RESULTS: Across 66 systematic reviews including single-arm study designs, 45% of authors correctly classified single-arm studies as case series or SATs, 15% used non-specific but reasonable classifications (e.g. single-arm or non-comparative study), and 40% misclassified studies (e.g. cohort study for case series or observational study for SAT). In terms of RoB assessment, 35% of authors used tools that were designed for single-arm studies, 41% used tools designed for assessing comparative studies or SRs, and 24% conducted no RoB assessment. For reviews including SATs, 29% of authors used tools designed for SATs. For reviews including case series, 19% of authors used tools designed for case series. For reviews reporting meta-analyses, all authors combined SATs and case series in analyses.

CONCLUSIONS: Findings suggest that there is currently a poor understanding of single-arm studies in the review community and that authors frequently use inappropriate RoB tools to assess them. RoB assessment should aim to discern study quality based on specific domains of bias encountered in different study designs.