Abstract

Objective: The aim of this study was to ascertain the risk factors of major postpartum haemorrhage (PPH), as substantiated from diagnoses and procedures recorded in the French hospital administrative database. Materials and methods: This nationwide populationbased study looked at 723 905 deliveries in mainland France in 2011, from the hospital database. The outcome of interest was major PPH identified by the International Classification of DiseasesICD-10 code for PPH in association with i) coding procedures for advanced measures for haemostasis or ii) need for intensive care or iii) death. Determinants studied included maternal and pregnancy characteristics, birth environmental factors and home-to-hospital distance. Adjusted odds ratios (aOR) for the outcome were calculated using a multilevel random-intercept model of logistic regression. Results: Major PPH was associated with age ≥ 35 years (aOR 1.41; 95% confidence interval, 95% CI [1.25–1.59]), multiple pregnancy (aOR 3.40 [2.85–4.05], pre-eclampsia (aOR 2.80 [2.32–3.38]), chorioamnionitis (aOR 2.57 [1.64– 4.03]), caesarean section (aOR 4.80, [4.27-5.39]) and delivery in a level III maternity unit. Home-to-hospital distance was not a significant risk factor for PPH. The AUC was 0.8 [0.83-0.85]. Conclusion: The model used gave accurate predictions of PPH occurrence according to risk factors identifiable from a French hospital database.