Work

Development and validation of a perioperative prediction model for in-hospital mortality after surgery for type A aortic dissection

Year: 2026

Type: article

Abstract: Type A aortic dissection (TAAD) remains a life-threatening cardiovascular emergency with substantial risk of postoperative mortality. Accurate perioperative risk stratification is essential to guide clinical decision-making and optimize patient outcomes. To develop, internally validate, and visualize a robust perioperative nomogram for predicting in-hospital mortality in patients undergoing surgical repair for TAAD. We conducted a retrospective cohort study of 998 patients who underwent open surgical repair for TAAD between May 2020 and April 2025. LASSO regression was performed for variable selection, followed by multivariable logistic regression to construct the predictive model. A visual nomogram was then developed based on the independent predictors. The model’s performance was evaluated using the area under the receiver operating characteristic curve (AUC) for discrimination, and calibration curves with the Hosmer–Lemeshow test. Internal validation and optimism correction of the nomogram were performed using bootstrap resampling with 1,000 iterations. The overall in-hospital mortality rate was 4.3% (43/998). In the multivariable model, inter-hospital transfer (OR = 3.86, 95% CI: 1.91–7.82), lower limb malperfusion (OR = 3.09, 95% CI: 1.32–6.96), and chest pain were identified as independent risk factors for in-hospital mortality. Conversely, admission hypertension (OR = 0.20, 95% CI: 0.08–0.48) and the presence of collateral circulation served as significant independent protective factors. The derived nomogram demonstrated strong discrimination, with a bias-corrected AUC of 0.824 (95% CI: 0.752–0.897) and exhibited robust calibration. We established and internally validated a parsimonious perioperative nomogram for TAAD surgical mortality. Our findings highlight that inter-hospital transfer and lower limb malperfusion are critical risk factors, while admission hypertension and collateral circulation are protective. This intuitive, paper-based nomogram provides clinicians with a practical tool for individualized, bedside risk estimation, though future external validation is warranted.

Source: Journal of Cardiothoracic Surgery

DOI: 10.1186/s13019-026-04713-6