Diagnostic Accuracy of Cervical Auscultation for Aspiration Detection in Adult Dysphagia: Implications for FEES Prioritization.

Objectives: To determine whether cervical auscultation (CA) provides a high-specificity, rule-in bedside signal to prioritize fiberoptic endoscopic evaluation of swallowing (FEES) in adults at risk of aspiration. Methods: Single-center, cross-sectional diagnostic-accuracy study in a tertiary hospital. Adults referred for clinically indicated swallow evaluation underwent the same-session, mutually blinded CA (index) and FEES (reference). The unit of analysis was the individual bolus; each bolus received paired CA and FEES ratings. We estimated sensitivity, specificity, predictive values, likelihood ratios (LR±), diagnostic odds ratio (DOR), and area under the ROC curve (AUC), with 95% confidence intervals; McNemar's test compared paired classifications. Results: Thirty-two adults contributed 256 analyzable boluses (aspiration prevalence 19.5%). CA sensitivity was 62.0% and specificity 99.5%, with accuracy 92.2%, PPV 96.9%, and NPV 91.5%. LR+ = 127.7 and LR- = 0.38; DOR = 334.5; AUC = 0.808. McNemar's test (FP = 1 vs. FN = 19) was significant, indicating more false negatives than false positives. At pretest probabilities of 10%, 20%, and 40%, a CA-positive result yielded post-test probabilities of 93.4%, 97.0%, and 98.8%, respectively; CA-negative results yielded post-test probabilities of 4.1%, 8.7%, and 20.3%. Conclusion: CA offers a high-specificity rule-in signal to triage and prioritize FEES rather than to exclude aspiration. Integrating CA into dysphagia pathways may improve access to instrumental assessment while maintaining safety. Level of Evidence: 2 (Diagnosis).