Development of a scoring system for prediction of placenta accreta and determine the accuracy of its results

Background: Optimal management of women with placenta accreta requires accurate preoperative diagnosis. Therefore, this study was conducted with the aim to evaluate a new prediction scoring items for risk assessment on placenta accreta and determine its accuracy ratio. Methods: This prospective cohort study was carried out on 159 suspected pregnant women morbidly adherent placenta (MAP) in Shariati, Imam Khomeini, and Yas Hospitals in Tehran from October 2016 to May 2018. The number of previous cesarean deliveries; lacunae stage, location of placenta; Doppler assessment; and loss of clear zone were used for review and scoring of ultrasound images. Ultimately after collecting scores, subjects fall into one of the following three categories: low (<= 5 points), moderate (6-7 points), or high (8-10 points) probability for placenta accreta. Ultimately, diagnosis of accreta was based on hysterectomy during surgery or reports of pathology. A logistic regression model was used to calculate the probability of placenta accreta on univariable analysis, to assess the discriminant power of all explanatory variables assessed by the receiver operating characteristic (ROC) curve. Results: The area-under-the-ROC curve of the composite scores was 98% and the overall sensitivity, specificity, and positive and negative predictive values of our developed scoring system were 91.84%, 87.27%, 86.54%, and 92.31%, respectively. Conclusion: Combination of several simple ultrasound and clinical characteristics in a scoring system may be highly effective for prenatal risk assessment and prediction of placenta accreta. Output of scoring system helps medical staff to prepare appropriately before surgery and avoid perinatal mortality and morbidity.