Postpartum Hemorrhagic Morbidities With Livebirth versus Stillbirth

ACOG publications on stillbirth or postpartum hemorrhage (PPH) do not consider stillbirth as a risk factor for postpartum hemorrhagic morbidity. Objective: To ascertain the likelihood of composite maternal hemorrhagic outcome (CMHO) among individuals that delivered vaginally with livebirth versus a stillbirth. Study Design: This was a retrospective cohort study of all parturients greater than 20 weeks gestation who delivered vaginally at a single level IV site within 24 months. Demographic differences and baseline PPH risks were analyzed. CMHO included any of the following: estimated blood loss > 1000 mL, use of uterotonics (beyond prophylactic oxytocin), Bakri balloon, surgical management of postpartum hemorrhage, blood transfusion, hysterectomy, venous thromboembolism (VTE), admission to the intensive care unit (ICU) or maternal death. Statistical analysis included Chi-square, Kruskal-Wallis, and Poisson regression with robust error variance for risk ratios, adjusting for gestational age (GA), bleeding on admission, chorioamnionitis, and prior uterine surgery. Results: Of 8,623 consecutive vaginal births > 20 weeks, 89 (1.9%) were stillbirths. Maternal age, marital status, gestational age (GA) at delivery, and PPH risk stratification at admission differed significantly. Bleeding at admission (p<0.001), prior uterine surgery (p<0.001), magnesium sulfate use (p=0.006), chorioamnionitis (p<0.001), platelet count <100 (p=0.001), platelet count <50 (p<0.001) and retained products of conception (p<0.001) were different in the two groups. CMHO was significantly higher with a stillbirth delivery (32.6% vs 16.8%; aRR 1.56 95%CI 1.01-2.46). After adjustment, the components of the CMHO that differed significantly were estimated blood loss > 1000mL and ICU admission. Tamponade, surgical intervention, VTE, hysterectomy and maternal death did not differ between the two groups. Pregnancies with stillbirth, compared to livebirth, had an increased risk of hemorrhagic related morbidity. In addition to being useful in shared decision making, our results can be nidus for intervention trials to decrease the hemorrhagic morbidity associated with stillbirth