Audrey Pivano1, Marine Alessandrini2, Raoul Desbriere3, Aubert Agostini4, Pierre Opinel5, Claude d'Ercole1, Jean-Baptiste Haumonte6. 1. Obstetrics and Gynaecology Unit, Assistance Publique-Hôpitaux de Marseille, Hôpital Nord, Marseille, France. 2. Research Unit EA3279, Laboratory of Public Health, Université Aix-Marseille, France. 3. Obstetrics and Gynaecology Unit, Hôpital Saint Joseph, Marseille, France. 4. Obstetrics and Gynaecology Unit, Assistance Publique-Hôpitaux de Marseille, Hôpital La Conception, Marseille, France. 5. Obstetrics and Gynaecology Unit, Centre hospitalier du Pays d'Aix, France. 6. Obstetrics and Gynaecology Unit, Assistance Publique-Hôpitaux de Marseille, Hôpital Nord, Marseille, France; Research Unit EA3279, Laboratory of Public Health, Université Aix-Marseille, France. Electronic address: jeanbaptiste.haumonte@ap-hm.fr.
Abstract
OBJECTIVE: To identify antenatal events associated with emergency caesarean sections in women presenting with antepartum bleeding and placenta praevia and to establish a score to predict the risk of emergency caesarean after a first bleeding episode has resolved. STUDY DESIGN: This retrospective multicentre study included 250 women presenting with antepartum bleeding and placenta praevia from 20 weeks of gestation until term in three maternity units. The score was constructed from data from 163 women after identification of antenatal risk factors associated with emergency caesareans for profuse bleeding due to placenta praevia. It was validated on a second independent cohort of 87 women. RESULTS: Three variables were significantly associated with emergency caesareans: major or complete praevia, defined as complete or partial praevia (OR=33.15 (95% CI 4.3-257); p=0.001), occurrence of 3 or more episodes of antepartum of uterine bleeding (OR=2.53 (95% CI 1.1-5.86); p=0.03), and a first (sentinel) bleeding episode before 29 weeks of gestation (OR=2.64 (95% CI 1.17-5.98); p=0.02). A fourth variable, moderate or severe antepartum uterine bleeding, was significantly associated with emergency caesareans in the univariate but not the multivariate analysis (p=0.006). These four variables were incorporated into a weighted scoring system that included major praevia (4 points), three or more episodes of antepartum bleeding (3), first bleeding episode before 29 weeks of gestation (3), and bleeding episode estimated as moderate or severe (1). A score ≥6/11 had a sensitivity of 83% and a specificity of 65% for predicting an emergency caesarean in the score development group and 95% and 62% in the validation group. CONCLUSION: A scoring system for placenta praevia with previous bleeding events, based on intensity, gestational age at sentinel bleed (before 29 weeks), number of bleeding episodes (≥3) and type of praevia (major) might be helpful to guide obstetric management and especially to determine the need for admission.
OBJECTIVE: To identify antenatal events associated with emergency caesarean sections in women presenting with antepartum bleeding and placenta praevia and to establish a score to predict the risk of emergency caesarean after a first bleeding episode has resolved. STUDY DESIGN: This retrospective multicentre study included 250 women presenting with antepartum bleeding and placenta praevia from 20 weeks of gestation until term in three maternity units. The score was constructed from data from 163 women after identification of antenatal risk factors associated with emergency caesareans for profuse bleeding due to placenta praevia. It was validated on a second independent cohort of 87 women. RESULTS: Three variables were significantly associated with emergency caesareans: major or complete praevia, defined as complete or partial praevia (OR=33.15 (95% CI 4.3-257); p=0.001), occurrence of 3 or more episodes of antepartum of uterine bleeding (OR=2.53 (95% CI 1.1-5.86); p=0.03), and a first (sentinel) bleeding episode before 29 weeks of gestation (OR=2.64 (95% CI 1.17-5.98); p=0.02). A fourth variable, moderate or severe antepartum uterine bleeding, was significantly associated with emergency caesareans in the univariate but not the multivariate analysis (p=0.006). These four variables were incorporated into a weighted scoring system that included major praevia (4 points), three or more episodes of antepartum bleeding (3), first bleeding episode before 29 weeks of gestation (3), and bleeding episode estimated as moderate or severe (1). A score ≥6/11 had a sensitivity of 83% and a specificity of 65% for predicting an emergency caesarean in the score development group and 95% and 62% in the validation group. CONCLUSION: A scoring system for placenta praevia with previous bleeding events, based on intensity, gestational age at sentinel bleed (before 29 weeks), number of bleeding episodes (≥3) and type of praevia (major) might be helpful to guide obstetric management and especially to determine the need for admission.