Prediction of Intra-abdominal Adhesions and Uterine Scar Grade Based on Abdominal Scar Characteristics in Women With a Previous Cesarean Section: A Diagnostic Accuracy Study.

Background: This study aims to investigate the relationship between cutaneous scar morphology and severe intra-abdominal adhesions and to predict uterine scar grade in repeat cesarean sections. It could be a valuable tool to plan elective repeat cesarean sections in patients with predicted weak uterine scars and also to have an experienced surgeon for repeat cesareans in patients predicted to have dense intra-abdominal adhesions.

Methods: Preoperatively, the external scar was assessed using the Manchester score in 260 women. Intraoperatively, the Knightly score was used for adhesions at 5 different sites. The total adhesion score was categorized into scores ≤5 and >5 (severe adhesions). The uterine scar was graded using the Qureshi method and grouped into intact scar and weak scar (grades 3 and 4). Fischer exact and chi-square tests were used to compare the groups. The receiver operating characteristic curve calculated a cutoff score for predicting severe adhesions and weak scars.

Results: Of the 260 patients, 63.5% had adhesions and 36.5% had no adhesions. The distortion of the skin scar had 71.3% specificity, and the texture had 84.8% sensitivity in predicting severe adhesions. Similarly, the texture had 72.2% sensitivity, and the distortion had 68.3% specificity in predicting weak uterine scar. A Manchester Scar Scale score of 9 includes the area under the curve of 0.72 for predicting severe adhesions and 0.62 for predicting weak uterine scar. A score of 9 was 66.6% sensitive and 66.5% specific, with a negative predictive value (NPV) of 93.2% for predicting severe adhesions. The same score was 55.5% sensitive and 65.1% specific, with an NPV of 90.1% for predicting weak uterine scar.

Conclusions: A Manchester Scar Scale cutoff score of 9 has a high NPV for predicting severe adhesions and weak uterine scar. The texture had high sensitivity, and distortion had high specificity for predicting severe adhesions.