Exploration de l'utilité d'un bref protocole de dépistage des troubles de déglutition en comparaison avec une vidéofluoroscopie simultanée

This study involved a direct blinded comparison between the results of a brief, standardized swallowing screening protocol and videofluoroscopy of exactly the same swallows. Forty adults participated. Each participant completed a brief swallow screening protocol involving tongue lateralization, voluntary cough, a voice task, and 2 swallowing tasks (3 swallows of 5cc thin liquid barium suspension and a cup-drinking task). We collected time-linked radiographic data and a high-definition movie of the participant's face, head and neck. The movie data were rated by 7 blinded clinicians (nurses and speech-language pathologists) for evidence of clinical signs associated with dysphagia. The videofluoroscopy data were rated by a separate panel of blinded speech-language pathologists for evidence of penetration-aspiration and post-swallow pharyngeal residues. Predictive statistics were calculated for the movie rating results, compared to the videofluoroscopy results. The results showed that none of the screening questions met our criteria for adequate predictive power: sensitivity, specificity and negative predictive values > 0.6, a false negative rate < 0.2 and a positive likelihood ratio > 1.0. We conclude that swallow screening decisions based on a series of 3-4 thin liquid swallows do not have good clinical utility for detecting dysphagia or penetration-aspiration. We discuss a number of issues in swallow screening research that may have contributed to the difference in these results compared to other studies.