Validation of the Yale Swallow Protocol: a prospective double-blinded videofluoroscopic study

The purpose of this prospective, double-blinded, multirater, systematic replication study was to investigate agreement for aspiration risk, in the same individual, between videofluoroscopic swallow studies (VFSS) and the Yale Swallow Protocol. Participants were 25 consecutive adults referred for dysphagia testing who met the inclusion criteria of completion of a brief cognitive assessment, oral mechanism examination, and no tracheotomy tube. First, all participants were administered the Yale Swallow Protocol by two experienced speech-language pathologists trained in protocol administration. Failure criteria were inability to drink the entire amount, interrupted drinking, or coughing during or immediately after drinking. Second, all participants completed a VFSS within 5-10 min of protocol administration. A speech-language pathologist, blinded to protocol results, reviewed the VFSS to determine aspiration status in a binary (yes/no) manner. Inter-rater agreement between two speech-language pathologists was 100 % for identification of aspiration risk with the Yale Swallow Protocol. Inter-rater agreement between the speech-language pathologist and the radiologist for identification of aspiration status with VFSS was 100 %. Twenty percent of VFSS recordings were viewed again 3-6 months after initial data collection, and intrarater agreement for identification of thin liquid aspiration was 100 %. Sensitivity for the Yale Swallow Protocol = 100 %, specificity = 64 %, positive predictive value = 78 %, and negative predictive value = 100 %. Importantly, all participants who passed the protocol did not aspirate during VFSS. Multiple, double-blinded raters and VFSS as the reference standard agreed with previous research with a single, nonblinded rater and FEES as the reference standard for identification of aspiration risk. The clinical usefulness and validity of the Yale Swallow Protocol for determining aspiration risk in a small sample size of male participants has been confirmed. Future research is needed with a larger and more heterogeneous population sample.