Risk factor assessment and adverse outcome prediction of placenta accreta in pregnant women after cesarean section complicated with placenta previa: a national multicenter retrospective study

Abstract

Objective: To study the risk factors of adverse pregnancy outcomes for the diagnosis and treatment of pregnancy after cesarean section complicated with placenta previa. Methods: A national multicenter retrospective study was conducted to select a total of 747 pregnant women with the third trimester singleton pregnancy after cesarean section complicated with placenta previa from 12 tertiary hospitals in January 1st to December 31st, 2018. The risk factors of severe adverse outcomes [hysterectomy, intraoperative blood loss ≥1 000 ml, intraoperative diagnosis of placenta accreta spectrum disorders (PAS)] in pregnant women with second pregnancy complicated with placenta previa after cesarean section were investigated by logistic regression analysis. The roles of prenatal ultrasonography and magnetic resonance imaging (MRI) in the prediction of PAS and severe adverse outcomes were observed. According to whether vascular intervention was performed (uterine artery embolization or abdominal aortic balloon occlusion), the pregnant women were divided into the blocked group and the unblocked group, and the maternal and infant perinatal outcomes between the two groups were compared. Results: (1) General information: the hysterectomy rate of 747 pregnant women with second pregnancy complicated with placenta previa after cesarean section was 10.4% (78/747), the intraoperative blood loss ≥1 000 ml in 55.8% (417/747), and PAS was confirmed in 47.5% (355/747). The incidence of uterine rupture was 0.8% (6/747). (2) Analysis of risk factors for severe adverse outcomes: based on binary unconditioned logistic regression univariate and multivariate analysis, the risk factors for hysterectomy were the mode of vascular embolization and intraoperative blood loss. The probability of hysterectomy with uterine artery embolization was 5.319 times higher than that with abdominal aortic balloon occlusion (95%CI: 1.346-21.018). The risk factors of intraoperative blood loss ≥1 000 ml were the number of cesarean section delivery, ultrasonography indicated PAS and suspected PAS, intraoperative PAS and complete placenta previa. The risk factors for intraoperative PAS were uterine scar thickness, ultrasonography indicated PAS and suspected PAS, MRI indicated PAS and suspected PAS, and complete placenta previa. (3) The roles of ultrasonography and MRI in predicting PAS: the sensitivity and specificity of ultrasonography in predicting PAS were 47.5% and 88.4%; the kappa value was 0.279 (P<0.001), with fair agreement. The sensitivity and specificity of MRI to predict PAS were 79.2% and 97.8%, respectively. The kappa value was 0.702 (P<0.001), indicating a good agreement. The intraoperative blood loss and hysterectomy rate of pregnant women with PAS indicated by ultrasonography and MRI were significantly higher than those with PAS only by ultrasonography or MRI. (4) Influence of vascular occlusion on pregnancy outcome: there were no significant differences in intraoperative blood loss and incidence of intraoperative bleeding ≥1 000 ml between the blocked group and the unblocked group (all P>0.05). There was no significant difference in intraoperative blood loss between the pregnant women with abdominal aortic balloon occlusion, uterine artery embolization and those without occlusion (P=0.409). The hysterectomy rate of pregnant women with uterine artery embolization was significantly higher than those with abdominal aortic balloon occlusion [39.3% (22/56) vs 10.0% (5/50), P=0.001]. Conclusions: In the third trimester of pregnancy with placenta previa after cesarean section, MRI examination has better consistency in predicting PAS than ultrasonography examination. Ultrasonography examination combined with MRI examination could effectively predict the hysterectomy rate and intraoperative blood loss. Vascular occlusion could not reduce the amount of intraoperative blood loss. The hysterectomy rate of pregnant women with uterine artery embolization is higher than those with abdominal aortic balloon occlusion.