Composite adverse outcomes in obstetric studies: a systematic review

• Published in: BMC pregnancy and childbirth 2021-02, Vol.21 (1), p.107-12, Article 107
• Main Authors: Herman, Dylan, Lor, Kar Yee, Qadree, Abdul, Horn, Daphne, D'Souza, Rohan
• Format: Article
• Language: English
• Subjects: Adverse pregnancy outcomes; Bias; Clinical trials; Composite endpoints; Composite outcomes; Eclampsia; Female; Hemorrhage; Humans; Hypertension; Infant, Newborn; Infant, Newborn, Diseases; Infections; Maternal Health; Maternal Mortality; Morbidity; Obstetric trials; Obstetrics; Outcome Assessment, Health Care; Perinatal Mortality; Preeclampsia; Pregnancy; Pregnancy Complications; Randomized Controlled Trials as Topic; Sepsis; Systematic review
• ISSN: 1471-2393; 1471-2393
• DOI: 10.1186/s12884-021-03588-w

Composite outcomes are increasingly being used in obstetric trials. The aim of this systematic review is to critically appraise the use of composite outcomes in obstetric RCTs with an intention of identifying limitations and providing potential solutions for future research. The study protocol was prospectively registered. Medline, Embase, Cochrane Databases and www.clinicaltrials.gov were searched for randomized controlled trials (RCTs) published in English between 1999 and 2019, using search terms related to pregnancy and composite outcomes. RCTs involving an obstetric condition that reported on a composite outcome. Screening and data extraction were performed in duplicate, and a descriptive synthesis and critical appraisal of composite obstetric outcomes, is presented. Of the 4170 results screened, we identified 156 RCTs, reporting on 181 composite outcomes. Of these, 158 composite outcomes related to general morbidity and mortality, either exclusively maternal (n=20), fetal-neonatal [perinatal] (n=116) or maternal and perinatal (n=22) were included in the final analysis. Obstetric composite outcomes included between two and 16 components. Components that comprised these composite outcomes were often dissimilar in terms of severity and frequency of occurrence, unlikely to have similar relative risk reductions and sometimes unrelated to the study's primary objective - important pre-requisites to consider while constructing composite outcomes. In addition, composite adverse obstetric outcomes often do not incorporate the perspectives of pregnant persons, embrace a holistic view of health or consider outcomes related to both members of the mother-fetus dyad. Composite outcomes are being increasingly used as primary outcomes in obstetric RCTs, based on which study conclusions are drawn and clinical recommendations made. However, there is a lack of consistency with regard to what components should be included within a composite adverse obstetric outcome and how these components should be measured. The use of novel research methods such as concept mapping may be able to address some of the limitations with the development of composite adverse obstetric outcomes, to inform future research.