Transesophageal Echocardiography – Dysphagia Risk in Acute Stroke (TEDRAS): a prospective, blind, randomized and controlled clinical trial

Abstract

Background and purpose: Dysphagia is common in acute stroke and leads to worse overall outcome. Transesophageal echocardiography (TEE) is used in the diagnostic evaluation of stroke with regard to its etiology and is a known cause of postoperative dysphagia in cardiac surgery. The prevalence of dysphagia in acute stroke patients undergoing TEE remains unknown. The aim of the Transesophageal Echocardiography – Dysphagia Risk in Acute Stroke (TEDRAS) study was to assess the influence of TEE on swallowing among patients who have experienced acute stroke. Methods: The TEDRAS study was a prospective, blind, randomized, controlled trial that included two groups of patients with acute stroke. Simple unrestricted randomization was performed, and examiners were blinded to each other's results. Swallowing was tested using flexible endoscopic evaluation of swallowing (FEES) at three different time points in the intervention group (24 h before, immediately after and 24 h after TEE) and in the control group (FEES on three consecutive days and TEE earliest after the third FEES). Validated scales were used to assess dysphagia severity for all time points as primary outcome measures. Results: A total of 34 patients were randomized: 19 to the intervention group and 15 to the control group. The key findings of the repeated-measures between-group comparisons were significant increases in the intervention group for the following dysphagia measures: (1) secretion severity score (immediately after TEE: P < 0.001; 24 h after TEE: P < 0.001) and (2) Penetration-Aspiration Scale score for saliva (immediately after TEE: P < 0.001; 24 h after TEE: P = 0.007), for small (immediately after TEE: P = 0.009) and large liquid boli (immediately after TEE: P = 0.009; 24 h after TEE: P = 0.025). Conclusion: The results indicate a negative influence of TEE on swallowing in acute stroke patients for at least 24 hours.