Machine learning predictive model for aspiration screening in hospitalized patients with acute stroke.

Dysphagia is a fatal condition after acute stroke. We established machine learning (ML) models for screening aspiration in patients with acute stroke. This retrospective study enrolled patients with acute stroke admitted to a cerebrovascular specialty hospital between January 2016 and June 2022. A videofluoroscopic swallowing study (VFSS) confirmed aspiration. We evaluated the Gugging Swallowing Screen (GUSS), an early assessment tool for dysphagia, in all patients and compared its predictive value with ML models. Following ML algorithms were applied: regularized logistic regressions (ridge, lasso, and elastic net), random forest, extreme gradient boosting, support vector machines, k-nearest neighbors, and naïve Bayes. We finally analyzed data from 3408 patients, and 448 of them had aspiration on VFSS. The GUSS showed an area under the receiver operating characteristics curve (AUROC) of 0.79 (0.77-0.81). The ridge regression model was the best model among all ML models, with an AUROC of 0.81 (0.76-0.86), an F1 measure of 0.45. Regularized logistic regression models exhibited higher sensitivity (0.66-0.72) than the GUSS (0.64). Feature importance analyses revealed that the modified Rankin scale was the most important feature of ML performance. The proposed ML prediction models are valid and practical for screening aspiration in patients with acute stroke.

Clinical Usefulness of the Korean Version of the Dysphagia Handicap Index: Reliability, Validity, and Role as a Screening Test.

This study aimed to measure the validity and reliability of the Korean version of the Dysphagia Handicap Index (K-DHI) and evaluate its diagnostic efficacy for predicting aspiration. We enrolled 104 patients with dysphagia symptoms (D group) and 88 controls (ND group). Among controls, there were 43 patients without dysphagia symptoms (ND patient group). All subjects completed the K-DHI survey. The D and ND group patients underwent the Gugging Swallowing Screen (GUSS) and videofluoroscopic swallowing study (VFSS). Two weeks later, the D group completed the second session of the K-DHI survey. The internal consistency of the K-DHI was good to excellent (Cronbach's α: 0.79-0.95). The test-retest reliability of the K-DHI survey was also high (interclass correlation coefficient = 0.88). There were moderate correlations between the K-DHI and GUSS (r = - 0.65, p < 0.001) as well as findings of VFSS-videofluoroscopic dysphagia scale (r = 0.55, p < 0.001) and American Speech-Language-Hearing Association National Outcome Measurement System swallowing scale (r = - 0.55, p < 0.001). For predicting aspiration, the K-DHI cutoff value was 11 (sensitivity, 0.82; specificity, 0.72; positive predictive value, 0.34; and negative predictive value, 0.96). K-DHI ≥ 11 [odds ratio (OR), 6.43; 95% Confidence Interval (CI) (1.87-22.16); p = 0.003] and GUSS ≤ 15 [OR 4.73; 95% CI (1.59-14.07); p = 0.005] were independent risk factors for aspiration on VFSS. The K-DHI is a reliable and valid self-reporting instrument for evaluating patient's quality of life associated with dysphagia among the Korean language population. It is also useful for the screening of aspiration.