• Steven B Leder and Julian F Espinosa.
• Department of Surgery, Section of Otolaryngology, Yale University School of Medicine, New Haven, Connecticut 06504, USA. steven.leder@yale.edu
• Dysphagia. 2002 Jan 1; 17 (3): 214-8.

AbstractAspiration is an important variable related to increased morbidity, mortality, and cost of care for acute stroke patients. This prospective systematic replication study compared a clinical swallowing examination consisting of six clinical identifiers of aspiration risk, i.e., dysphonia, dysarthria, abnormal gag reflex, abnormal volitional cough, cough after swallow, and voice change after swallow, with an instrumental fiberoptic endoscopic evaluation of swallowing (FEES) to determine reliability in identifying aspiration risk following acute stroke. A referred consecutive sample of 49 first-time stroke patients was evaluated within 24 hours poststroke, first with the clinical examination followed immediately by FEES. The endoscopist was blinded to results of clinical testing. The clinical examination correctly identified 19 subjects with aspiration risk, when compared with the criterion standard FEES, but incorrectly identified 3 patients as having no aspiration risk when they did. The clinical examination incorrectly identified 19 subjects with aspiration risk but determined correctly no aspiration risk in 8 patients who did not exhibit aspiration risk on FEES. Clinical examination sensitivity = 86%; specificity = 30%; false negative rate = 14%; false positive rate = 70%; positive predictive value = 50%; and negative predictive value = 73%. It was concluded that the clinical examination, when compared with FEES, underestimated aspiration risk in patients with aspiration risk and overestimated aspiration risk in patients who did not exhibit aspiration risk. Careful consideration of the limitations of clinical testing leads us to believe that a reliable, timely, and cost-effective instrumental swallow evaluation should be available for the majority of patients following acute stroke.

20 keywords...

hide…