Validation of a shortened Taiwanese version of an asthma quality-of-life questionnaire by Rasch model

Abstract

The purpose of this study was to assess the one-dimensionality, item fit, redundancy, and differential item functioning (DIF) of a Taiwanese version of the Asthma Quality of Life Questionnaire (TAQLQ) using Rasch model and to examine the test-retest reliability, validity of a simplified version of the TAQLQ and the original AQLQ. The design of this study was a two-time assessment on a 1-month interval. A total of 143 asthmatic outpatients were requested to administer the original 7-point 32-item TAQLQ questionnaires at the first visit and the 5-point 27-item TAQLQ questionnaires at the second visit. Two versions of the questionnaire were analyzed by using the Rasch model (Rasch G, Probabilistic Models for Some Intelligence and Attainment Tests, Chicago: University of Chicago Press, 1992) to determine the difference in validity and reliability between the original 32-item and the shortened 27-item TAQLQ scales with different categories responded. The analyzed data show that 5 items in the original 32-item AQLQ are not fitting to the Rasch model's expectations. The shortened 27-item TAQLQ was created by removing the misfit and redundant items. The reliability of the shortened 27-item TAQLQ is greater than the original 32-item AQLQ. The quality of life for asthma patients could be improved and stratified into five strata by summed raw scores of 61, 89, 106, and 120. The cutoff thresholds of averaged raw scores are 2.3, 3.3, 3.9, and 4.4. The simplified 5-point 27-item TAQLQ holds better validity and higher reliability than the original 7-point 32-item TAQLQ. Assessment of the quality of life using an abbreviated 5-point 27-item TAQLQ could be applied to Taiwanese asthma patients who are less educated in southern Taiwan than those in the northern part of Taiwan.