ABSTRACT

Objective : The purpose of this study was to develop a Swallowing Function Test (SFT) for stroke patients byapplying a Rasch analysis based on the item response theory. Methods : The subjects were 153 stroke patients with dysphagia in 5 categories based on 23 items of apreliminary SFT. The data were analysed using a Rasch analysis to investigate whether the persons anditems fit, the distribution of item difficulty, separation reliability, and the appropriateness of the rating scale. Finally, the items were categorized using a factor analysis. Results : Eleven out of 153 subjects were found to not fit. Six out of 23 items were also found to be misfit. Themisfitted items included‘ history of pneumonia’,‘ oral sensation’,‘ gag reflex’,‘ nasal/oral regurgitation’, ‘swallowing delayed’and ‘stridor after swallowing’. Of the appropriate items, the most difficult was‘laryngeal elevation’, and the easiest was‘ alertness’. The SFT rating scale was shown to be appropriate forstroke patients with dysphagia. The item separation reliability was .95, and the person separation reliabilitywas .98. Finally, four categories for 17 items were completed for the SFT through a factor analysis. Conclusion : The SFT for stroke patients with dysphagia was proven to be reliable and valid. It is expected thatSFT may help identify the clinical features of patients, and the recovery level may present interventionguidelines, considering the level of difficulty of the items.