Development and validation of a nomogram for predicting the probability of serious complication in esophageal cancer patients after esophagectomy

Abstract Background Esophagectomy remains a procedure with one of the highest complication rates. This study aimed to develop and validate a model of nomogram for predicting the probability serious postoperative complications for the postoperative esophageal cancer (EC) patients. Methods An observational study was conducted at Daping Hospital, China, encompassing 529 postoperative EC patients from December 2014 to November 2023. We integrated independent prognostic factors to craft a predictive model, specifically a nomogram, designed to forecast the risk of severe postoperative complications. Results The incidence of serious complications for postoperative EC patients was 58.1%. A total of seven variables, including age, gender, bleeding volume during the operation (P = 0.035), operating time (P = 0.031), the Nutritional Risk Screening 2002 (NRS 2002) score (P = 0.027), the perioperative nutrition screen (PONS) score (P = 0.018), and cystatin C levels (P < 0.001), were applied to draw the nomogram of predicting serious complications after esophagectomy. The accuracy of predictive value for the model was quantified by Harrell’s C index at 0.690 (95% confidence interval = 0.596–0.785, P < 0.001). Conclusion We developed a nomogram system to demonstrate exceptional predictive capabilities for postoperative EC patients in forecasting the risk of serious complications. These results also emphasize the predictive value of the preoperative nutritional risk screening with NRS 2002 and PONS in EC patients undergoing esophagectomy.