Decision Analysis Studies Involving the Management of Pregnant Women

INTRODUCTION: This study systematically reviewed Decision Analysis (DA) and economic evaluation (EE) studies involving the management of pregnant women and described methodological variations and limitations. METHODS: Three databases were searched to include articles from inception to December 2014. DA and EE studies describing conditions involving the management of pregnant women were included and screening and data-extraction were conducted in duplicate. Outcomes included type of analysis, model type, input variables (“probabilities,” “utilities” and “costs”) and outcome variables (clinical or cost-based, time horizon). Reporting quality was assessed by a five-point scale. RESULTS: 6,055 titles were screened and 303 full texts retrieved, of which 57 describing management of clinical conditions were included. Five used Markov Models, while 52 used decision trees. Most studies (45) involved EE. Probabilities were obtained from multiple publications in 48 studies. Utilities were used in 26 studies [maternal and offspring (18), offspring (5) and combined (3)]. Utilities were obtained from physicians (6), physician and patient (3), mother and offspring (2), mother (2), public (2), parents (2) and offspring (1). Costs were obtained from perspectives of society (10), health care system (7), third-party payer (6), hospital (5) and multiple sources (2). 53 studies described clinical outcomes and 4 only described costs. Time-horizon involved maternal and/or offspring lifetime (19), pregnancy duration (19) and other reported time-horizons (6). Only five studies had a high risk-of-bias. CONCLUSION: There is considerable variation in the conduct and reporting of DA and EE studies in obstetrics. A consensus statement to standardize conduct of these studies in obstetrics is required.