Predicting the risk of postoperative death risk in Acute Type A Aortic Dissection: development and evaluation of a new predictive nomogram

Abstract Background Acute type A aortic dissection (ATAAD) is an emergency condition characterized by severe chest pain and back pain, with rapid disease progression and a very high mortality rate. The current methods for predicting postoperative mortality rate in acute type A aortic dissection are inadequate, necessitating the urgent need for new prediction methods. Methods This study is a retrospective analysis of 309 patients with ATAAD in The First Affiliated Hospital Zhejiang University of Medicine. By utilizing the LASSO and logistic regression analysis, we have developed a novel predictive model for postoperative mortality rate. The model incorporates factors such as platelet count (PLT), lactic acid (LA), hydroxybutyrate dehydrogenase (HBDH) , activated partial thromboplastin time (APTT) , deep hypothermic circulatory arrest (DHCA) time to predict the risk of mortality in patients. Results The predictive nomogram included predictors such as PLT, LA, HBDH, APTT, and DHCA time. With a C-index of 0.9787, the model demonstrated good discrimination power, calibration, and ROC curve. It was able to maintain a high C-index value of 0.984 even during interval verification. Conclusions We have developed and validated a novel predictive model for assessing postoperative mortality risk in Chinese ATAAD patients. This predictive tool demonstrates good discriminatory ability and calibration, which can assist clinicians in making more accurate risk assessments and devising personalized treatment plans.