Screening for oropharyngeal dysphagia in stroke: insufficient evidence for guidelines.

There is no evaluation of the evidence for the screening of oropharyngeal dysphagia in stroke. We reviewed the literature on clinical screening for oropharyngeal dysphagia in adults with stroke to determine (a) the accuracy of different screening tests used to detect dysphagia defined by abnormal oropharyngeal physiology on videofluoroscopy and (b) the health outcomes reported and whether screening alters those outcomes. Peer-reviewed English-language and human studies were sought through Medline (from 1966 to July 1997) by using the key words cerebrovascular disorders and deglutition disorders, relevant Internet addresses, and extensive hand searching of bibliographies of identified articles. Of the 154 sources identified, 89 articles were original, peer-reviewed, and focused on oropharyngeal dysphagia in stroke patients. To evaluate the evidence, the next selection identified 10 articles on the comparison of screening and videofluoroscopic findings and three articles on screening and health outcomes. Evidence was rated according to the level of study design by using the values of the Canadian Task Force on Periodic Health Examination. From the identified screening tests, most of the screenings were related to laryngeal signs (63%) and most of the outcomes were related to physiology (74%). Evidence for screening accuracy was limited because of poor study design and the predominant use of aspiration as the diagnostic reference. Only two screening tests were identified as accurate: failure on the 50-ml water test (likelihood ratio = 5.7, 95% confidence interval = 2.5-12.9) and impaired pharyngeal sensation (likelihood ratio = 2.5, 95% confidence interval = 1.7-3.7). Limited evidence for screening benefit suggested a reduction in pneumonia, length of hospital stay, personnel costs, and patient charges. In conclusion, screening accuracy needs to be assessed by using both abnormal physiology and aspiration as diagnostic markers for dysphagia. Large well-designed trials are needed for more conclusive evidence of screening benefit.