Construction and Validation of a Risk Prediction Model for Postoperative Dysphagia in Patients with Head and Neck Cancer
Abstract:
Objective: This study aimed to construct a nomogram and web calculator for predicting the risk of dysphagia in patients with head and neck cancer after surgery.
Design: A prospective study
Methods: patients who met the exclusion criteria in a class III a hospital in Zhejiang Province from October 2023 to June 2024 were selected as the modeling group, and patients in the same hospital from July 2024 to October 2024 were selected as the validation group. SPSS software was used for single factor and multi factor analysis to build the prediction model, and R software was used to draw the nomogram and web cal culator. Internal and external validation of the constructed prediction model.
Results: the incidence of dysphagia in the modeling group and the validation group were 56.4% and 53.8%, respectively. Age (OR = 2.332), smoking history (OR = 2.423), tumor T stage (OR = 1.818), primary tumor location, neck lymph node dissection (OR= 2.342), flap transplantation (OR = 2.954), nutritional risk (OR = 1.772) and Beck oral score (OR = 2.482) were independent predictors of dysphagia in patients with head and neck cancer after surgery. The areas under the receiver operating characteristic curve of the modeling group and the validation group were 0.893 and 0.890, respectively.
Conclusion: the incidence of postoperative dysphagia in patients with head and neck cancer is high. The risk prediction model for early postoperative dysphagia in patients with head and neck cancer constructed in this study has good discrimination, calibration and clinical applicability. It can be used as an evaluation and prediction tool to help clinical medical staff early identify high-risk groups of postoperative dysphagia in patients with head and neck cancer.
