Emergency delivery in case of suspected placenta accreta spectrum: Can it be predicted?

Hanulikova, Petra Savukyne, Egle Fox, Karin A Sobisek, Lukas Mhallem Gziri, Mina [UCL] van Beekhuizen, Heleen J Stefanovic, Vedran Braun, Thorsten Paping, Alexander Bertholdt, Charline Morel, Olivier IS‐PAS et al.[show all]

INTRODUCTION: The main goal of placenta accreta spectrum (PAS) screening is to enable delivery in an expert center in the presence of an experienced team at an appropriate time. Our study aimed to identify independent risk factors for emergency deliveries within the IS-PAS 2.0 database cohort and establish a multivariate predictive model. MATERIAL AND METHODS: A retrospective analysis of prospectively collected PAS cases from the IS-PAS database between January 2020 and June 2022 by 23 international expert centers was performed. All PAS cases (singleton and multiple pregnancies) managed according to local protocols were included. Individuals with emergent delivery were identified and compared to those with scheduled delivery. A multivariate analysis was conducted to identify the possible risk factors for emergency delivery and was used to establish a predictive model. Maternal outcomes were compared. RESULTS: Overall, 315 women were included in the study. Of these, 182 participants (89 with emergent and 93 with scheduled delivery) were included in the final analysis after exclusion of those with unsuspected PAS antenatally or who lacked information about the urgency of delivery. Gestational age at delivery was higher in the scheduled group (34.7 vs. 32.9, p < 0.001). Antenatal bleeding (OR 2.9, p = 0.02) and a placenta located over a uterine scar (OR 0.38, p = 0.001) were the independent predictive factors for emergent delivery (AUC 0.68). Ultrasound (US) markers: loss of clear zone (p = 0.001), placental lacunae (p = 0.01), placental bulge (p = 0.02), and presence of bridging vessels (p = 0.02) were more frequently documented in the scheduled group. None of these markers improved the predictive values of the model. Higher PAS grades were identified in the scheduled group (p = 0.01). There were no significant differences in maternal outcomes. CONCLUSIONS: Antenatal bleeding and the placental location away from the uterine scar remained the most significant predictors for emergent delivery among patients with PAS, even when combining more predictive risk factors, including US markers. Based on these results, patients who bleed antenatally may benefit from transfer to an expert center, as we found no differences in maternal outcomes between groups delivered in expert centers. Earlier-scheduled delivery is not supported due to the low predictive value of our model.