Development and validation of a nomogram model for pneumonia after redo cardiac surgery

Abstract

AimsPostoperative pneumonia (POP) after redo cardiac surgery is prevalent, associated with poor outcome. The aim of this study was to identify independent risk factors for POP after redo cardiac surgery and to develop and validate a prediction model.MethodsAdults undergoing redo cardiac surgery from 2016 to 2019 were identified in a single-institution database. Using a 2: 1 ratio, the patients were randomly divided into training and validation sets. Univariate and multivariate analyses were applied to identify independent predictors for POP in the training set. A nomogram model was constructed for clinical utility and was validated in the validation set.ResultsPOP developed in 72 of the 376 patients (19.1%). Four independent risk factors were identified, including age, chronic obstructive pulmonary disease, serum creatinine level and intraoperative blood transfusion volume. A nomogram based on the four predictors was constructed, with good discrimination in both the training (c-index: 0.86) and validation sets (c-index: 0.78). The model was well calibrated, with a Hosmer-Lemeshow χ2-value of 7.31 (P = 0.50) in the training set and 7.41 (P = 0.49) in the validation set. The calibration was also good by visual inspection. The decision and clinical impact curves of the nomogram indicated good clinical utility. Three risk intervals were identified based on the nomogram for better risk stratification.ConclusionWe developed and validated a nomogram model for POP after redo cardiac surgery. The model may have good clinical utility in risk evaluation and individualized treatment to reduce adverse events. Graphical abstract Incidence, risk factor, and outcomes of postoperative pneumonia after redo cardiac surgery: http://links.lww.com/JCM/A445.