The Dysphagia in Stroke Protocol Reduces Aspiration Pneumonia in Patients with Dysphagia Following Acute Stroke: a Clinical Audit

Sarah E. Perry, Anna Miles, John N. Fink, Maggie-Lee Huckabee

Cough reflex testing has been evaluated as a component of the clinical swallowing assessment as a means of identifying patients at risk of aspiration during swallowing. A previous study by our research group found good sensitivity and specificity of the cough reflex test for identifying patients at risk of aspiration post-stroke, yet its use did not decrease pneumonia rates, contrary to previous reports. The aim of this study was to expand on our earlier work by implementing a clinical management protocol incorporating cough reflex testing within the same healthcare setting and compare patient outcomes to those from the original study and to evaluate clinical outcomes in patients with acute stroke who were managed using the Dysphagia in Stroke Protocol (DiSP). Secondarily, to compare those outcomes to historical data prior to implementation of the DiSP. This clinical audit measured outcomes from 284 patients with acute stroke managed per the DiSP, which guides use of videofluoroscopic swallowing study and patient management based on clinical exam with cough reflex testing. Data from our previous trial were included for comparison of pre- and post-DiSP patient outcomes. Data collection took place between November 2012 and April 2016 at four urban hospitals within New Zealand. Following implementation of the DiSP, the rate of aspiration pneumonia (10%) was substantially lower than the pre-DiSP rate (28%), with no pneumonia readmissions within 3 months. Pneumonia-related mortality was unchanged. By 3 months, 81% of the patients were on a normal diet and 67% had returned home, compared to pre-DiSP outcomes of 55% and 55% respectively. Previous work has suggested that simply implementing cough reflex testing in dysphagia management may not be sufficient to improve patient outcomes. The present study adds to this picture by suggesting that the true variable of influence may be the way in which the results of the test are applied to patient care. There is a strong case to support the use of a structured protocol if cough reflex testing is to be implemented in clinical practice.