Join now to get access to this content and more.
Become a SOAP member and have access to our benefits.
- For Review: SOAP Consensus Statement on Neuraxial Procedures in Thrombocytopenic Parturients
- Sample Centers of Excellence Applications
- ASA Corner
- SOAP Policy and Procedure Manual (P&P Manual)
- SOAP Expert Opinions
- SOAP's Learning Modules
- 2019 Annual Meeting Lecture Videos
- December 2018 - SOAP Unofficial Guide to ASA Committees Webinar
- Submit a Position
- View Job Postings
- Previous Meeting Archives
- Previous Meeting Abstract Search
- CMS Guidelines
- Member Benefits
- Newsletter Clinical Articles
- ACOG Documents
- Search our Patient Safety Archive
- Ask SOAP a Question
- Global Health Opportunities
- And more…
Collecting data on intrapartum characteristics yields worthwhile improvements in risk prediction for operative delivery
Abstract Number: S 26
Abstract Type: Meta Analysis/Review of the Literature
The use of risk factors to predict the probability of operative delivery (forceps or cesarean section) in a laboring woman could improve obstetrical decision-making. Recently Schuit et al. (2012, BJOG) constructed two risk prediction models for operative delivery: Model 1, a baseline model involving only antepartum characteristics, and Model 2, an extended model that added intrapartum characteristics. We questioned whether the improvement in prediction with Model 2 is worth the “cost” in time and money of increased data collection on intrapartum characteristics ( induction of labor, oxytocin augmentation, intrapartum fever, rupture of membranes > 24 hours, epidural analgesia, and meconium-stained amniotic fluid).
A standard approach to answering this question is to compute receiver operating characteristic (ROC) curves for Model 1 versus Model 2 (Figure 1, left). The ROC curve plots true positive rate (TPR) versus false positive rate (FPR). To incorporate costs and benefits, we applied the following decision-analytic approach (Baker and Kramer, 2012, Discovery Medicine). Let T denote the risk threshold, here the risk of operative delivery at which a women would be indifferent between receiving elective Cesarean section or not. Because the risk threshold T summarizes the tradeoff between the benefit of a true positive and the cost of a false positive, it can be used to compute relative utility (RU) which is the anticipated clinical utility of prediction relative to the anticipated clinical utility of perfect prediction. The relative utility curve is a plot of RU versus T and is used to compare the performances of Models 1 and 2 (Figure 1, right).
Based on the maximum difference in RU curves between Models 1 and 2 (at T=.40), Model 2 is worthwhile over Model 1 if it is acceptable to trade the collection of data on intrapartum characteristics among at least 76 women for every correct prediction of operative delivery. Because this tradeoff is reasonable we recommend that clinicians collect data on intrapartum characteristics and use Model 2 to estimate the risk of operative delivery.