Value of ultrasonic scoring system for predicting risks of placenta accreta

Objective To explore the value of ultrasound scoring system in predicting the type and risk of placenta accreta. Methods Clinical data of 180 placenta accreta patients who delivered in the Peking University Third Hospital between January 2005 and November 2014, were retrospectively analyzed. Prenatal ultrasonographic features were analyzed, including position and thickness of the placenta, disappearance of hypo-echoes in posterior placenta, continuousness ofbladder line, existence of lacuna, condition of the subplacental vascularity, completeness of cervical morphology, existence of cervical sinus, and history of cesarean section. A score of 0, 1 or 2 was given to each item, and a sum-up was calculated for each patient. The cut-off scores of patients with placenta accreta, placenta increta and placenta percreta were calculated by receiver operating characteristic carve, respectively. At the same time, blood loss and hysterectomy rate were compared among the three groups. Variance analysis, rank sum or Chi-square tests were used for statistical analysis. Results Among the 180 cases, there were 115 cases of placenta accreta, 38 of planceta increta and 27 of placenta percreta. Placenta increta and percreta were defined as the severe type. Blood loss in placenta accreta was lower than in the severe type [200 (100-4 000) ml vs 3 025 (100-15 000) ml, P<0.01]. There was no difference in blood loss between patients with placenta increta or percreta (P=0.350). No hysterectomy was performed for patients with placenta accreta, the rate being lower than in the severe type [0 vs 29.2% (19/65), P<0.01]. Among the severe type, 18.4% (7/38) of the placenta increta patients underwent hysterectomy, the rate being lower than in placenta percreta patients [44.4% (12/27), P<0.01]. The score in placenta accreta was lower than in the severe type [(1.88±1.45) vs (7.01±2.15) scores, P<0.01]. In the severe type, the score in placenta increta was lower than in placenta percreta [(6.08±2.62) vs (8.74±2.75), P<0.01]. The receiver operating characteristic curve showed that the cut-off score of placenta accreta and the severe type was 5 [area under the curve (AUC)=94.3%, the score ≥4.5, the sensitivity=81.5%, and the specificity=95.7%], the cut-off score of placenta accreta and increta was 3 (AUC=91.1%, score ≥2.5, the sensitivity=92.1%, and the specificity=75.7%), and the cut-off score of placenta increta and percreta was 10 (AUC=74.6%, score ≥9.5, the sensitivity=55.6%, and the specificity=89.5%). Conclusions Ultrasound scoring system is effective in assessing types of placenta accreta and predicting its associate risks, and alerting the possibility of hysterectomy. It also facilitates preoperative planning and guides physicians in formulating subsequent treatment plans. Placenta accreta and the severe type (placenta increta and percreta) can be distinguished by cut-off scores ≥5, and a score ≥10 implies a higher risk of placenta percreta. Key words: Placenta accreta; Ultrasonography, prenatal; Severity of illness index; Blood loss, surgical; Forecasting