Risk Score for Predicting Dysphagia in Patients After Neurosurgery: A Prospective Observational Trial

Background: Acquired dysphagia is common in patients with tracheal intubation and neurological disease, leading to increased mortality. This study aimed to ascertain the risk factors and develop a prediction model for acquired dysphagia in patients after neurosurgery. Methods: A multicenter prospective observational study was performed on 293 patients who underwent neurosurgery. A standardized swallowing assessment was performed bedside within 24 h of extubation, and logistic regression analysis with a best subset selection strategy was performed to select predictors. A nomogram model was then established and verified. Results: The incidence of acquired dysphagia in our study was 23.2% (68/293). Among the variables, days of neurointensive care unit (NICU) stay [odds ratio (OR), 1.433; 95% confidence interval (CI), 1.141-1.882; P = 0.005], tracheal intubation duration (OR, 1.021; CI, 1.001-1.062; P = 0.175), use of a nasogastric feeding tube (OR, 9.131; CI, 1.364-62.289; P = 0.021), and Acute Physiology and Chronic Health Evaluation (APACHE)-II C score (OR, 1.709; CI, 1.421-2.148; P < 0.001) were selected as risk predictors for dysphagia and included in the nomogram model. The area under the receiver operating characteristic curve was 0.980 (CI, 0.965-0.996) in the training set and 0.971 (0.937-1) in the validation set, with Brier scores of 0.045 and 0.056, respectively. Conclusion: Patients who stay longer in the NICU, have a longer duration of tracheal intubation, require a nasogastric feeding tube, and have higher APACHE-II C scores after neurosurgery are likely to develop dysphagia. This developed model is a convenient and efficient tool for predicting the development of dysphagia.