Premium
Analyzing the 22‐item Sino‐Nasal Outcome Test using item response theory
Author(s) -
Crump R. Trafford,
Liu Guiping,
Janjua Arif,
Sutherland Jason M.
Publication year - 2016
Publication title -
international forum of allergy and rhinology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.503
H-Index - 46
eISSN - 2042-6984
pISSN - 2042-6976
DOI - 10.1002/alr.21770
Subject(s) - chronic rhinosinusitis , medicine , differential item functioning , item response theory , endoscopic sinus surgery , sinusitis , patient reported outcome , clinical psychology , psychometrics , physical therapy , test (biology) , quality of life (healthcare) , surgery , paleontology , nursing , biology
Background The 22‐item Sino‐Nasal Outcome Test (SNOT‐22) is a widely applied patient‐reported outcome instrument used to assess the severity of symptoms associated with chronic rhinosinusitis. The purpose of this study was to evaluate the measurement performance of the SNOT‐22 instrument on an item‐level basis, in a sample of patients awaiting elective surgery for chronic rhinosinusitis. Methods This study involved secondary analysis of SNOT‐22 data that was prospectively collected from patients diagnosed with chronic rhinosinusitis and awaiting endoscopic sinus surgery in Vancouver, Canada. This study used classic test theory and a 2‐parameter graded‐response model to evaluate the SNOT‐22 items’ abilities to measure the severity of chronic rhinosinusitis in terms of patients’ self‐reported symptoms. This approach models each item's discriminability and difficulty, which provides insight into how well they respectively measure symptoms related to chronic rhinosinusitis. Results Factor analyses indicated that there are 5 domains of measurement in the SNOT‐22. The majority of items demonstrated strong discriminability between symptom severities. Likewise, most of the items demonstrated strong difficulty measuring the symptoms across their range of levels. The exception was those items related to psychological symptoms. Differential item functioning demonstrated that very few of the SNOT‐22 items were answered significantly differently by gender or age subgroups. Conclusion This item‐level analysis demonstrates that, in general, the SNOT‐22 is a strong instrument. Items related to psychological symptoms require further investigation and warrant a supplemental patient‐reported outcome instrument.