A Prediction Model for Post-aortic Dissection Infection Based on Rewarming Time, Drainage Output, and Hemoglobin: A Retrospective Study

Abstract Background There is currently a lack of early and simple tools for assessing the risk of postoperative infection following aortic dissection surgery. This study aimed to develop and validate a prediction model based on early postoperative objective indicators, along with a corresponding simplified scoring system. Methods A total of 243 patients who underwent surgical treatment for aortic dissection were retrospectively enrolled. Independent predictors of postoperative infection were identified using logistic regression analysis. Based on these predictors, a nomogram prediction model was constructed. The model's performance was evaluated in terms of discrimination (area under the receiver operating characteristic curve, AUC), calibration (mean absolute error, MAE), and clinical utility (decision curve analysis, DCA). Furthermore, a simple risk score was established using the optimal cut-off values for each predictor, and its effectiveness for risk stratification was validated using a trend test. Results Rewarming time (≤ 83.5 min), postoperative day 2 drainage output (> 261.5 ml), and postoperative day 1 hemoglobin level (≤ 109.5 g/L) were identified as independent predictors of infection. The constructed nomogram demonstrated good predictive performance (AUC = 0.704, MAE = 0.036). When transformed into a 0–3 point simplified scoring system, patients were stratified into low-, medium-, and high-risk groups. The infection rates for these groups were 6.7%, 21.7%, and 38.6%, respectively, showing a significant increasing trend (P for trend = 0.004). Conclusion The simple scoring system developed in this study can effectively identify patients at high risk for infection in the early postoperative period, providing a practical tool for implementing stratified management and precise intervention.