Prediction Model for Postoperative Pressure Injury in Patients with Acute Type A Aortic Dissection

Abstract

OBJECTIVE To establish a risk assessment model to predict postoperative National Pressure Injury Advisory Panel stage 2 or higher pressure injury (PI) risk in patients undergoing acute type A aortic dissection surgery. METHODS This retrospective assessment included consecutive patients undergoing acute type A aortic dissection surgery in the authors’ hospital from September 2017 to June 2021. The authors used LASSO (logistic least absolute shrinkage and selection operator) regression analysis to identify the most relevant variables associated with PI by running cyclic coordinate descent with 10-times cross-validation. The variables selected by LASSO regression analysis were subjected to multivariate logistic analysis. A calibration plot, receiver operating characteristic curve, and decision curve analysis were used to validate the model. RESULTS There were 469 patients in the study, including 94 (27.5%) with postoperative PI. Ten variables were selected from LASSO regression: body mass index, diabetes, Marfan syndrome, stroke, preoperative skin moisture, hemoglobin, albumin, serum creatinine, platelet, and d-dimer. Four risk factors emerged after multivariate logistic regression: Marfan syndrome, preoperative skin moisture, albumin, and serum creatinine. The area under the receiver operating characteristic curve of the model was 0.765. The calibration plot and the decision curve analysis both suggested that the model was suitable for predicting postoperative PI. CONCLUSIONS This study built an efficient predictive model that could help identify high-risk patients.