Comparing accuracy of the Yale swallow protocol when administered by registered nurses and speech‐language pathologists

Aims and objectives (1) To describe the results of a web‐based teaching module used by registered nurses to identify patients at risk of aspiration and (2) to determine accuracy of the registered nurse‐administered 3‐ounce water swallow challenge protocol, that is, drinking three ounces of water, a basic cognitive screen and oral mechanism evaluation, when compared with blinded ratings from speech‐language pathology. Background Early identification of potential swallowing problems is important prior to ingestion of food, fluid and medications. Unfortunately, current nurse‐administered screens use a variety of non‐evidence‐based assessments. It would be beneficial to use a valid, reliable and evidence‐based screen, that is, the Yale swallow protocol. Design Prospective, blinded, referral‐based. Methods Fifty‐two registered nurses and 101 inpatients participated. First, each participant was administered the 3‐ounce water swallow challenge protocol by a speech‐language pathologist. Second, a nurse administered the protocol to the same patient within one hour and independently recorded results and diet recommendations. The nurse was blinded to the study's purpose and results of the speech‐language pathologist's initial screening. Out of view, but simultaneous with the nurse‐administered protocol, a speech‐language pathologist rerated the patient's challenge for comparison with initial results and determined the accuracy of the nurse‐administered protocol. Results Intra‐ and inter‐rater protocol agreements for the two speech‐language pathologists were 100%. Inter‐rater protocol agreement between registered nurses and speech‐language pathologists was 98·01%. Conclusions Results confirm the reliability and accuracy of a registered nurse–administered Yale swallow protocol. The consequence of 98% accuracy combined with previously reported 96·5% sensitivity, 97·9% negative predictive value and