Development and validation of a risk prediction model for postoperative gastrointestinal complications in patients undergoing aortic arch surgery

Abstract

Background Postoperative gastrointestinal complications (GICs) were potentially fatal to patients who underwent aortic arch surgery. The aim of this study was to construct a prediction model of GICs. Methods We retrospectively studied the medical records of 3063 patients who underwent aortic arch surgery. Patients were randomly divided into derivation and validation cohorts at a ratio of 4:1. A nomogram was constructed in the derivation cohort. Results A total of 157 patients with GICs were identified. In the derivation cohort, multivariate analysis identified six predictors of GICs including hypertension, ASA classification, preinduction MAP, aortic cross-clamp time, CPB time, and intraoperative transfusion of red blood cells. Compared with the patients without GICs, the patients with GICs had higher mortality, and longer ICU and hospital stays. We also divided the patients into four intervals according to the risk of GICs. Conclusions This study developed a reliable prediction model of GICs after aortic arch surgery. This prediction model had been well verified in our research centre, and further external verification was required before it can be recommended for clinical application.