Discriminant Ability of the Eating Assessment Tool-10 to Detect Swallowing Safety and Efficiency Impairments.

Abstract

Quick, sensitive dysphagia screening tools are necessary to identify high-risk patients for further evaluation in busy multidisciplinary amyotrophic lateral sclerosis (ALS) clinics. We examined the relationship between self-perceived dysphagia using the validated Eating Assessment Tool-10 (EAT-10) and videofluoroscopic analysis of swallowing safety and efficiency. Prospective, observational, longitudinal study. Individuals with ALS completed the EAT-10 and a videofluoroscopic swallowing study. Duplicate, independent, blinded analyses of the validated Dynamic Imaging Grade of Swallowing Toxicity (DIGEST) scale were performed to index swallowing safety and efficiency (mild dysphagia: DIGEST ≥ 1, moderate dysphagia: DIGEST ≥ 2). A between-groups analysis of variance with Games-Howell test for post-hoc pairwise comparisons was performed to examine EAT-10 scores across dysphagia severity levels. Receiver operator characteristic curve analysis, area under the curve (AUC), sensitivity, specificity, positive-negative predictive values (PPV, NPV), and odds ratios (OR) were derived. Four hundred and thirty five paired EAT-10 and DIGEST scores were analyzed. Mean EAT-10 score was 8.48 (95% confidence interval [CI]: 7.63-9.33). Individuals with dysphagia demonstrated higher EAT-10 scores (mild: 4.1 vs. 11.3, moderate: 6.0 vs. 17.5, P < .001). Mean EAT-10 scores increased across DIGEST levels (D0: 4.1, D1: 7.9, D2: 15.1, D3: 20.4, D4: 39.0). For mild dysphagia, an EAT-10 cut score of 3 was optimal: AUC 0.74 (95% CI: 0.69-0.78; sensitivity: 77%; specificity: 53%; PPV: 71%; NPV: 60%; OR: 3.5). An EAT-10 cut score of 7 optimized detection of moderate dysphagia: AUC 0.83 (95% CI: 0.78-0.87; sensitivity: 81%; specificity: 66%; PPV: 39%; NPV: 93%; OR: 8.1). The EAT-10 is an easy-to-administer dysphagia screening tool with good discriminant ability for use in ALS clinics. 2 Laryngoscope, 132:2319-2326, 2022.