Premium
A validated model for the 22‐item Sino‐Nasal Outcome Test subdomain structure in chronic rhinosinusitis
Author(s) -
Feng Allen L.,
Wesely Nicholas C.,
Hoehle Lloyd P.,
Phillips Katie M.,
Yamasaki Alisa,
Campbell Adam P.,
Gregorio Luciano L.,
Killeen Thomas E.,
Caradonna David S.,
Meier Josh C.,
Gray Stacey T.,
Sedaghat Ahmad R.
Publication year - 2017
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.22025
Subject(s) - medicine , cronbach's alpha , confirmatory factor analysis , chronic rhinosinusitis , exploratory factor analysis , construct validity , statistics , physical therapy , structural equation modeling , clinical psychology , psychometrics , mathematics
Background Previous studies have identified subdomains of the 22‐item Sino‐Nasal Outcome Test (SNOT‐22), reflecting distinct and largely independent categories of chronic rhinosinusitis (CRS) symptoms. However, no study has validated the subdomain structure of the SNOT‐22. This study aims to validate the existence of underlying symptom subdomains of the SNOT‐22 using confirmatory factor analysis (CFA) and to develop a subdomain model that practitioners and researchers can use to describe CRS symptomatology. Methods A total of 800 patients with CRS were included into this cross‐sectional study (400 CRS patients from Boston, MA, and 400 CRS patients from Reno, NV). Their SNOT‐22 responses were analyzed using exploratory factor analysis (EFA) to determine the number of symptom subdomains. A CFA was performed to develop a validated measurement model for the underlying SNOT‐22 subdomains along with various tests of validity and goodness of fit. Results EFA demonstrated 4 distinct factors reflecting: sleep, nasal, otologic/facial pain, and emotional symptoms (Cronbach's alpha, >0.7; Bartlett's test of sphericity, p < 0.001; Kaiser‐Meyer‐Olkin >0.90), independent of geographic locale. The corresponding CFA measurement model demonstrated excellent measures of fit (root mean square error of approximation, <0.06; standardized root mean square residual, <0.08; comparative fit index, >0.95; Tucker‐Lewis index, >0.95) and measures of construct validity (heterotrait‐monotrait [HTMT] ratio, <0.85; composite reliability, >0.7), again independent of geographic locale. Conclusion The use of the 4‐subdomain structure for SNOT‐22 (reflecting sleep, nasal, otologic/facial pain, and emotional symptoms of CRS) was validated as the most appropriate to calculate SNOT‐22 subdomain scores for patients from different geographic regions using CFA.