Using a Predictive Model of Summative, Stratified Variables to Assess Risk for Postpartum Hemorrhage [5B

Abstract

INTRODUCTION: Postpartum hemorrhage (PPH) is a leading cause of maternal mortality. A previous study (Dilla et al) showed statistically different hemorrhage rates among groups rated low, medium, and high risk, but each group had low overall hemorrhage rates. METHODS: From 9/2012-5/2013, women who delivered at Advocate Lutheran General Hospital at gestational age greater than 20 weeks were included in this retrospective case control study. Risk factors present on admission and during labor were given score of 1, 2, or 3 based on existing literature and recommendation by an expert panel. Each patient was labeled low, moderate, or high risk based on total score at time of delivery. Definition of PPH was considered estimated blood loss greater or equal to 500 cc for vaginal delivery, greater or equal to 1000 cc for cesarean section, or if uterotonic medications beyond standard Pitocin protocol were administered. RESULTS: Of 1227 women, 115 (9.37%) experienced a PPH. The incidence of PPH in each risk group was: low (7%), moderate (22%), and high (31%). The proportion of the total hemorrhages occurring in each group was: low (64%), moderate (32%), and high (3%). CONCLUSION: This predictive model was better at predicting PPH than previous studies. Future studies with larger patient populations are needed to validate these findings. This type of model using a summation of all risk factors prior to delivery may represent an improved way to predict and thus prevent PPH.