Nomogram for Predicting Early-Tracheostomy Decannulation in Patients Undergoing Partial Laryngectomy

This study aimed to explore the predictive factors of early-tracheostomy decannulation in patients with laryngeal cancer undergoing partial laryngectomy and to develop a nomogram for predicting the probability of early decannulation (within 1 month).

This study conducted a retrospective analysis of patients with laryngeal cancer who underwent laryngeal function preservation surgery at the First Affiliated Hospital of Wenzhou Medical University from January 2012 to December 2021. Univariate analysis and multivariate logistic regression analysis were used to identify independent predictors factors, and the nomogram is constructed based on the results of multivariate logistic regression analysis. Receiver operating characteristic (ROC) curves, calibration curves, and decision curves were employed to validate the clinical utility of the model.

A total of 141 patients were analyzed. Among them, 82 patients successfully achieved early decannulation, while 59 patients failed to undergo early decannulation, the multivariate logistic regression analysis showed age (odds ratio [OR] 0.954, 95% CI 0.912-0.999), history of smoking (OR 0.975, 95% CI 0.951-0.998), surgical modality (OR 0.590, 95% CI 0.404-0.861), drainage tube (OR 0.216, 95% CI 0.08-0.580) is the independent risk factor for early-tracheostomy decannulation. The area under the curve of the ROC curve is 0.756 (95% CI 0.673-0.840) indicating good discriminative capacity. The calibration curve and the Hosmer-Lemeshow test (P = .138) demonstrate good model fitted and high calibration. The incidence of early decannulation was 58.2%, and at the threshold of 58.2%, the decision curve was above the none and all lines, so the model has clinical utility.

A nomogram model based on age, history of smoking, surgical modality, and drainage tube was successfully developed to predict the probability of early-tracheostomy decannulation.