Risk of bias assessments in systematic reviews and meta-analyses of behavioral interventions for substance use outcomes.

Abstract

OBJECTIVE\nThis study aims to explore the current practice of risk of bias assessment in systematic reviews of behavioral clinical trials published in substance use journals and how assessment results were incorporated into meta-analysis.\n\n\nSTUDY DESIGN AND SETTING\nThe authors searched for systematic reviews and meta-analyses of behavioral interventions published from 2016 to 2020 in 40 substance use journals. Two authors independently screened and extracted relevant information from each article. Different tools for risk of bias assessment and approaches of incorporating the risk of bias assessment results into meta-analysis were summarized.\n\n\nRESULTS\nThe study identified 35 systematic reviews and meta-analyses of behavioral clinical trials. Among the 35 reviews, 31 (89%) assessed the risk of bias of their included studies. Twelve (39%) of the 31 reviews incorporated these assessment findings into their meta-analysis of intervention effects (e.g., conducted meta-regression or subgroup analysis, sensitivity analysis, limited the synthesis only to the high quality studies).\n\n\nCONCLUSION\nPerforming and reporting risk of bias assessment remain inconsistent in published systematic reviews. Future systematic reviews and meta-analyses are encouraged to connect their risk of bias assessment findings with meta-analysis and follow the most updated PRISMA guidelines in reporting the methods and results of risk of bias assessment.