Diagnostic accuracy of the Dysphagia Trained Nurse Assessment tool in acute stroke.

BACKGROUND Comprehensive swallow screening assessments to identify dysphagia and make early eating and drinking recommendations can be used by trained nurses. This study aims to validate the Dysphagia Trained Nurse Assessment (DTNAx) with acute stroke patients. METHODS Participants with diagnosed stroke were prospectively and consecutively recruited from an acute stroke unit. Following a baseline DTNAx on admission they underwent a Speech and Language Therapist bedside assessment of swallowing (SLTAx), a videofluoroscopy (VFS) and a further DTNAx by the same or a different nurse. RESULTS 47 participants were recruited, of which 22 had dysphagia. Compared to the SLTAx in the identification of dysphagia, the DTNAx had a sensitivity of 96.9% (95% confidence intervals CIs 83.8%-99.9%) and specificity 89.5% (95% CIs 75.2%-97.1%). Compared to the VFS in the identification of aspiration, the DTNAx had a sensitivity of 77.8% (CIs 40.0%-97.2%) and specificity 81.6% (CIs 65.7% to 92.3%). Over 81% of the diet and fluid recommendations made by the DTNs were in absolute agreement compared to the SLTAx. Both the DTNAx and the SLTAx had low diagnostic accuracy compared to the VFS definition of dysphagia. CONCLUSIONS Nurses trained to use the DTNAx showed good diagnostic accuracy in identifying dysphagia compared to SLTAx and in identifying aspiration compared to VFS. They made appropriate diet and fluid recommendations in line with SLTs in the early management of dysphagia.