The Attempt To Establish A Predicting Model For Severe Postoperative Hypoxemia After Surgery For Standford Type A Aortic Dissection

Objective To study the risk factors of severe postoperative hypoxemia after surgery for Standford type A aortic dissection and fabricate a predicting model.Method Data of 411 consecutive patients from January 2014 to April 2015, who underwent surgery for Standford type A aortic dissection in the department of cardiovascular surgery of Beijing Anzhen Hospital, were collected retrospectively. All the cases were divided into two groups according to the appearance of severe postoperative hypoxemia. All the data about potential risk factors was put into the database and analyzed by Logistic regression. The predicting model was made up by using the variables with statistical difference, which were regarded as independent risk factors. The discrimination and calibration of the predicting model were assessed with ROC curve and Hosmer-Lemeshow goodness-of-fit statistic.Result The in-hospital mortality was 6.57%. Severe postoperative hypoxemia(Pa O2/Fi O2<100mm Hg) happened on 69 cases within 48 hours after procedures, with the incidence rate of 17.1%. The Logistic regression demonstrated Body Mass Index, age, preoperative serum myoglobin, preoperative serum creatinine, preoperative serum alanine aminotransferase, the time of cardiopulmonary bypass, postoperative re-exploration in 48 hours are the independent risk factors of the incidence for severe postoperative hypoxemia. The area under ROC curve of the model is 0.8031, and the P value in Hosmer-Lemeshow test is 0.554.Conclusion The Logsitic model built in this reseach succeeded to predict the incidence of severe postoperative hypoxemia after surgery for Standford type A aortic dissection, and it could fullfill the clinical doctors’ requrement with its excellent discrimination and calibration.