Risk factors for postpartum hemorrhage in a cohort of 6011 Italian women

INTRODUCTION: Postpartum hemorrhage is responsible for 25% of maternal pregnancy-related deaths and it is the first cause of maternal morbidity and mortality worldwide. OBJECTIVE: To define the prevalence of postpartum hemorrhage and associated risk factors after vaginal birth and to develop a risk model that improves postpartum hemorrhage prediction. PATIENTS AND METHODS: All women who underwent a vaginal delivery at the Obstetric Unit of a large University hospital in Milan (Italy) between July 2007and September 2009 were enrolled. Postpartum hemorrhage was defined as ≥500mL blood loss. A nomogram tailored to predict postpartum hemorrhage was developed, summarizing the impact of each covariate on the probability of postpartum hemorrhage. RESULTS: 6011 women were studied (24% had blood loss ≥500mL and 4.8% ≥1000mL). Nulliparity, episiotomy, retained placenta and high neonatal body weight were confirmed as risk factors for postpartum hemorrhage. The odds ratio of postpartum hemorrhage was 0.86 (95%CI 0.78-0.90) for each 1gr/dL increase in ante-partum hemoglobin. An extensive internal validation of the developed nomogram demonstrated a good stability of the risk model. CONCLUSIONS: Low ante-partum hemoglobin is a new potentially modifiable risk factor for postpartum hemorrhage. A nomogram to predict the probability of postpartum hemorrhage is now available for external validation.