Risk factors for postoperative hypoxemia in Stanford type A aortic dissection: a cohort study from the Hubei Aortic Dissection Registration Database

Abstract

For patients with Stanford type A aortic dissection (STAAD), moderate-to-severe hypoxemia is a common and serious complication after surgery. However, its independent risk factors are not yet well understood, and there is a lack of effective early prediction tools. Therefore, this study aims to analyze the risk factors for postoperative moderate-to-severe hypoxemia in STAAD patients and to develop a nomogram prediction model. This study included patients with STAAD who underwent surgery in Hubei Province between 2020 and 2022 and were registered in the Hubei Aortic Dissection Registration Database. Postoperative hypoxemia was defined as oxygenation index (OI) < 200 mmHg (moderate-to-severe hypoxemia) at 48-h postoperatively. Receiver operator characteristic (ROC) curves analysis and a nomogram prediction model were built based on the results of the multivariable logistic regression analysis. Finally, 379 patients with STAAD were included, with 298 of whom suffered from postoperative moderate-to-severe hypoxemia. Multivariable logistic regression showed that body mass index (BMI) > 25.0 kg/m 2 (OR = 2.163, 95%CI: 1.199–3.902), low glomerular filtration rate (OR = 2.385, 95%CI: 1.276-ཞ4.457), and moderate hypothermic circulatory arrest (MHCA) duration (OR = 1.304, 95%CI: 1.202–1.415) were independently associated with postoperative moderate-to-severe hypoxemia. The prediction model for postoperative moderate to severe hypoxemia was logit (P) =-3.229 + 0.771 × BMI > 25.0 + 0.265 × MHCA time + 0.869 × low glomerular filtration rate. The area under the ROC curve was 0.843 (0.765–0.868), with a sensitivity of 55.4% and a specificity of 91.4%. Decision curve analysis showed that the model provided a net benefit across threshold probabilities ranging from 22.4% to 100%, indicating clinical usefulness within this range. BMI > 25.0, low GFR, and prolonged MHCA duration are independent risk factors for postoperative moderate‑to‑severe hypoxemia in STAAD. The model shows fair discrimination (AUC = 0.843), though not yet sufficient for standalone clinical decision-making. Nevertheless, these associations offer useful clues for risk stratification; future studies incorporating additional perioperative variables are needed to strengthen generalizability.