Determinants of emergency delivery in pregnancies complicated by placenta previa or placenta accreta spectrum disorders: analysis of ADoPAD Cohort

Objectives: To report the rate and the outcomes of unplanned caesarean delivery (CD) delivery in women with placenta accreta spectrum disorders (PAS) and placenta previa without PAS, and to elucidate the diagnostic accuracy of ultrasound in predicting this outcome. Methods: Secondary analysis of a multicenter prospective study involving 14 referral hospital in Italy (ADoPAD Study). Inclusion criteria were women with a low-lying placenta (< 20 mm from the internal cervical os) or placenta previa (covering the os), aged ≥ 18 years, who underwent trans-abdominal and transvaginal ultrasound assessment at ≥ 26 + 0 weeks of gestation. The primary outcome was the occurrence of emergency CD, defined as the need of immediate surgical intervention performed for emergency maternal or fetal indications, including active labour, cumulative maternal bleeding >500 ml, severe and persistent vaginal bleeding such that maternal hemodynamic stability cannot be achieved or maintained or category III fetal heart rate tracing unresponsive to resuscitative measures in women with PAS and in those with placenta previa with no PAS. The primary outcome was reported in the population of women with placenta previa and with no PAS confirmed after birth and in those with PAS separately. The secondary aim was to report the strength of association and to test the diagnostic accuracy of ultrasound in predicting emergency delivery. Univariate, multivariate, and diagnostic accuracy analyses were used to analyse the data. Results: 450 women (97 with PAS and 353 with placenta previa but not PAS) were included in the analysis. In women with PAS disorders, emergency CD was required in 21% (95% CI 14-30%) and 60% women delivered before 34 weeks of gestation. Mean gestational age at delivery was 32.3±2.7 weeks in women undergoing emergency and 34.9±1.8 weeks (p<0.001) in those undergoing elective CD. Women undergoing emergency CD had a higher median estimated blood loss (2500 ml, IQR 1350-4500 vs 1100 ml, IQR 625-2500, p=0.012) and mean units of blood transfused (4.3±1.6 vs 0.8±2.2; p= 0.02) compared to those undergoing elective delivery. At univariate analysis, the presence of interrupted retroplacental space, bladder line and placental lacunae were more common in women not experiencing emergency CD. At multivariate analysis, only maternal BMI (OR: 0.83. 95% CI 0.69-0.99, p= 0.045) was independently associated with emergency delivery in women with PAS. However, ultrasound signs of PAS, including presence of interrupted retroplacental space, bladder line and placental lacunae, were not associated neither predictive of emergency CD. In women with placenta previa but not PAS, emergency CD was required in 31.1% (95% CI 26.6-36.2) and 32.8% delivered before 34 weeks of gestation. Mean gestational age at delivery was lower in women undergoing compared to those not undergoing emergency delivery (34.2±2.9 vs 36.7±1.6; p<0.001). Pregnancies complicated by emergency CD, had new-borns with a lower birthweight (2330±620 g vs 2800±620 g, p<0.001) and had a higher risk of receiving blood transfusions (22.7% vs 10.7%, p= 0.003) compared to those who underwent elective CD. At multivariate analysis, only placental thickness (p= 0.046) and a cervical length < 25 mm (OR: 3.89, 95% CI 3.89-11.33, p= 0.01) were associated with emergency CD. However, a short CL showed a low diagnostic accuracy for predicting emergency CD in these women. Conclusion: Emergency CD complicated about 20% of women with PAS disorders and 30% of those with placenta previa and not PAS and is associated with a worse maternal and perinatal outcome compared to elective intervention. Prenatal ultrasound cannot entirely predict the risk of emergency delivery in women with these disorders. This article is protected by copyright. All rights reserved.