Ahmad Danial, Serna-Gallegos Derek, Jackson Ariana, Kaczorowski David J, Bonatti Johannes, West David M, Yoon Pyongsoo D, Chu Danny, Squire Joe, Thoma Floyd, Zhu Jianhui, Phillippi Julie, Sultan Ibrahim
Division of Cardiac Surgery, Department of Cardiothoracic Surgery, University of Pittsburgh, Pittsburgh, Pa.
Heart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, Pa.
JTCVS Open. 2024 Dec 18;23:81-88. doi: 10.1016/j.xjon.2024.12.007. eCollection 2025 Feb.
Surgery for acute type A aortic dissection carries a high risk of morbidity and mortality compared with routine cardiac surgical procedures. The German Registry for Acute Aortic Dissection Type A score has been recommended for use as a mortality risk-stratification tool in recent guidelines. We sought to externally validate this score in our local population.
All consecutive patients undergoing surgery for acute type A aortic dissection from 2007 to 2021 were included. Logistic regression analyses were performed. Model discrimination was assessed by C-statistic with 95% CIs as part of the receiver operating characteristic analysis. Model performance was visualized by calibration plot and quantified by the Brier score.
A total of 587 patients were included. The mean age was 61 years (±13.5), with 42.08% of patients aged more than 65 years; 40.37% were female. The mean circulatory arrest time was 30.9 minutes (±16.5). Hemiarch replacement was performed in 62% of patients, and total arch replacement was performed in 35.3% of patients. Thirty-day mortality was observed in 66 patients (11.24%), and stroke was present in 7.16% of patients. The C-statistic revealed good discriminatory ability for predicting 30-day mortality (area under the receiver operating characteristic curve, 0.73; 95% CI, 0.67-0.79; < .0001). Model calibration was good (Brier score = 0.094).
The German Registry for Acute Aortic Dissection Type A score for 30-day mortality showed good discriminatory ability in our local population along with good ability for prediction of mortality, indicating its potential clinical utility in the population with acute type A aortic dissection.
与常规心脏手术相比，急性A型主动脉夹层手术的发病率和死亡率风险较高。德国急性A型主动脉夹层注册评分在最近的指南中被推荐用作死亡风险分层工具。我们试图在本地人群中对该评分进行外部验证。
纳入2007年至2021年期间所有连续接受急性A型主动脉夹层手术的患者。进行逻辑回归分析。通过C统计量及95%置信区间评估模型辨别能力，作为受试者操作特征分析的一部分。通过校准图直观显示模型性能，并通过Brier评分进行量化。
共纳入587例患者。平均年龄为61岁（±13.5），42.08%的患者年龄超过65岁；40.37%为女性。平均体外循环时间为30.9分钟（±16.5）。62%的患者进行了半弓置换，35.3%的患者进行了全弓置换。66例患者（11.24%）出现30天死亡率，7.16%的患者出现中风。C统计量显示预测30天死亡率具有良好的辨别能力（受试者操作特征曲线下面积，0.73；95%置信区间，0.67 - 0.79；P <.0001）。模型校准良好（Brier评分为0.094）。
德国急性A型主动脉夹层注册30天死亡率评分在我们的本地人群中显示出良好的辨别能力以及死亡率预测能力，表明其在急性A型主动脉夹层人群中具有潜在的临床应用价值。
