Composite Adverse Endpoints In Obstetric Studies: A Systematic Review

INTRODUCTION: The Composite Adverse Obstetric Outcome Study (CAOOS) is currently in the process of standardizing the use of composite outcomes in obstetric trials. This study aims to determine how composite outcomes have been reported and defined, as a preliminary step towards their standardization. METHODS: We conducted a systematic review of English-language randomized controlled trials (RCTs) published between 1999 and 2019 that involved an obstetric condition and reported on a composite outcome. We searched MEDLINE, EMBASE, CENTRAL, www.clinicalTrials.gov and reference lists of included studies. We screened studies and extracted data in duplicate, on study characteristics, and reported on variations on composite outcomes, their components, definitions and/or measurements. RESULTS: Of the 4170 results screened, 156 RCTs, reporting on 183 composite outcomes, were included in the final analysis. Of the composite outcomes, 115 were related to the fetus/neonate, 46 to the mother and 22 to adverse pregnancy outcomes which included maternal and perinatal morbidity. Composites comprised between 2 and 16 components. Maternal composite outcomes included components related to mortality and morbidity (17/46), general morbidity alone (4/46), system-specific morbidity (11/46) and wound-related morbidity (14/46). Perinatal composites did not always include components related to in-utero and neonatal morbidity and mortality. Composite pregnancy outcomes included maternal and perinatal morbidity and mortality components in various combinations. CONCLUSION: There is considerable variation in how composite outcomes are reported and defined in obstetric RCTs. Patients and stakeholders will be invited to participate in a multi-step mixed-methods research project to determine which components should comprise composite adverse obstetric outcomes.