Screening for Dysphagia in Adult Patients with Stroke: Assessing the Accuracy of Informal Detection
This study assessed the accuracy of informal dysphagia detection prior to implementation of a formal screening protocol. We conducted a secondary analysis of data captured between 2003 and 2008 from a sample of 250 adult stroke survivors admitted to a tertiary care centre. Using a priori criteria, patient medical records were reviewed for notation about dysphagia; if present, the date/time of notation, writer ’s profession, and suggestion of dysphagia presence. To assess accuracy of notations indicating dysphagia presence, we used speech language pathology (SLP) assessments as the criterion reference. There were 221 patient medical records available for review. Patients were male (56%), averaged 68 ye ars (SD = 15.0), with a mean Canadian Neurological Scale score of 8.1 (SD = 3.0). First notations of swallowing by SLP were excluded. Of the remaining 170 patients, 147 (87%) had first notations (104 by nurses; 40 by physicians) within a median of 24.3 h from admission. Accuracy of detecting dy sphagia from informal notations was low, with a sensitivity of 36.7% [95% CI, 24.9, 50.1], but specificity was high (94.2% [95% CI, 86.5, 97.9]). Informal identification methods, although timely, are suboptimal in their accuracy to detect dysphagia and leave patients with stroke at risk for poor hea lth outcomes. Given these findings, we encourage the use of psychometrically validated formal screening protocols to identify dysphagia.