Nomogram for predicting of postoperative delirium in elderly patients undergoing radical esophagectomy

Abstract

Background Postoperative delirium (POD) commonly occurs in elderly patients with esophageal cancer after curative surgery and is associated with serious outcomes. This study aimed to construct and validate a nomogram to predict POD after esophagectomy for carcinoma.Methods 1107 patients who underwent esophageal cancer surgery at the Fujian Cancer Hospital from January 2018 to September 2022 were retrospectively enrolled from a prospective database as the training cohort. The least absolute shrinkage and selection operator (LASSO) method was performed to screen the potential risk factors. Multivariate logistic regression was used to identify the independent predictors and to develop the the nomogram in the training cohort. The receiver operating characteristic (ROC) curve, calibration plot, and decision curve analysis (DCA) plots were adopted to assess the predictive accuracy of the nomogram. The validation cohort consisted of 338 patients treated at First Affiliated Hospital of Fujian Medical University was performed to externally validate the nomogram.Results The rate of POD was 16.6%(184/1107) in the training cohort. Age, diabetes mellitus, ASA classification, albumin level and surgical approach were identified as independent predictors for POD in multivariate analyses. The nomogram yielded great discrimination, with the area under the ROC curve were 0.788(95%CI: 0.752, 0.824) and 0.812(95%CI: 0.762, 0.862) in the training cohort and the validation cohort, respectively. The calibration plots and Hosmer–Lemeshow tests demonstrated a good correlation between the nomogram prediction and actual observation. The DCA further illustrated a high clinical application value of the model.Conclusions This study established a reliable nomogram based on five clinical parameters that could can be accurate and applicable in the probability of POD in elderly patients undergoing esophageal cancer surgery.