Construction of a Postoperative Complication Risk Model for Patients With Stanford Type A Aortic Dissection

Scritto il 09/05/2026
da Xiaohui Liu

J Clin Hypertens (Greenwich). 2026 May;28(5):e70248. doi: 10.1111/jch.70248.

ABSTRACT

To identify risk factors for postoperative complications in Stanford Type A aortic dissection (TAAD) patients, a risk prediction model was developed and validated. The study included a specific cohort of TAAD patients (aged 32-66 years) who underwent surgery with general anesthesia and cardiopulmonary bypass. A risk model was created using multivariate logistic regression analysis. Of the 522 patients, 174 (33.33%) experienced at least one major postoperative complication, including neurological, respiratory, renal, or other organ dysfunction. Univariate analysis identified several preoperative and intraoperative factors significantly associated with complications. Five independent predictors-age, BMI, stroke history, aortic cross-clamp time, and deep hypothermic circulatory arrest time-were retained in the multivariate logistic regression model as risk factors for postoperative complications in TAAD patients. Based on the logistic regression equation, logistic (P) = -7.458 + 0.056 × 1 + 0.102 × 2 + 0.374 × 3 + 0.018 × 4 + 0.026 × 5, a multivariate logistic regression risk model was developed to calculate the risk of postoperative complications, p = 1/1 + exp (-7.458 + 0.056 × 1 + 0.102 × 2 + 0.374 × 3 + 0.018 × 4 + 0.026 × 5). Model validation results showed that the area under the ROC curve was 0.879, the Hosmer-Lemeshow test p-value was 0.178, with a sensitivity of 77.59% and specificity of 85.34%.The incidence of postoperative complications following surgery for TAAD is relatively high. Both patient physical condition and surgical factors are associated with these complications. Establishing a risk model can help assess the risk of major postoperative complications in TAAD, providing clinically relevant information for perioperative monitoring and early intervention. Due to the specific age distribution of the study population, the results should be applied with caution to elderly patients (> 66 years).

PMID:42104739 | DOI:10.1111/jch.70248