Construction and validation of an ultrasound-based nomogram model for predicting dysphagia in patients with chronic obstructive pulmonary disease.

INTRODUCTION

Dysphagia is common in chronic obstructive pulmonary disease (COPD), prompting the need for predictive models for this condition. In this study, we aimed to develop a nomogram model for dysphagia prediction in patients with COPD.

METHODS

Data from 300 patients with COPD were divided into the training ( = 210) and validation ( = 90) cohorts. Independent risk factors for dysphagia were identified using logistic regression and used to construct a nomogram model. The model's predictive efficacy, accuracy, and clinical utility were evaluated using receiver operating characteristic curve analysis, calibration, decision curve analysis, and clinical impact curves.

RESULTS

Hypoglossal-hyoid shortening rate, hyoid-larynx shortening approximation distance, chin time of movement of the genioglossus, and distance of movement of the genioglossus were identified as independent risk factors. The nomogram exhibited areas under the curve of 0.834 and 0.804 in the training and validation cohorts, respectively, indicating good predictive efficacy and calibration.

CONCLUSION

The nomogram model effectively predicts dysphagia occurrence in patients with COPD, providing a valuable tool for risk assessment.