Establishment and validation of a nomogram to predict thirty-day unplanned reoperations of primary anastomosis in neonates with intestinal atresias.

Background/purpose: Unplanned reoperation rates becoming a critical metric for evaluating healthcare quality and have received increasing attention in recent years. Intestinal atresia (IA) has a high rate of unplanned reoperations. The purpose of this study is to evaluate the thirty-day unplanned reoperation rates and their risk factors in neonates with intestinal atresias after primary anastomosis surgery, and to construct a predictive nomogram. Methods: We developed and internally validated a predictive model from a retrospective cohort of 200 neonates admitted to our hospital for primary anastomosis surgery. The primary outcome was thirty-day unplanned reoperation rates. Independent factors significantly associated with thirty-day unplanned reoperation rates were identified using multivariable logistic regression analysis. The effectiveness of the developed nomogram was evaluated through calibration, discrimination, and clinical utility. Results: The incidence of thirty-day unplanned reoperation rates was 11%. Multivariable analysis identified the type of bowel anastomosis and combined meconium peritonitis as independent factors predicting thirty-day unplanned reoperation rates. The derivation model showed good discrimination, with a C-index of 0.791 (95% CI, 0.685-0.897), and good calibration (Hosmer-Lemeshow test P = 0.231). The analysis of the decision curve showed that the nomogram was beneficial in clinical practice. Conclusion: We developed a nomogram to predict thirty-day unplanned reoperations of primary anastomosis in neonates with IA. This prediction model may enable assist in clinical decision-making, patient counseling, and treatment planning.