Item analysis and evaluation on chronic obstructive pulmonary disease quality of life scale based on classical test theory and item response theory

In recent years, the incidence of chronic obstructive pulmonary disease (COPD) has been increasing and the quality of life of patients with COPD has received more and more attention. The QLICD-COPD (V2.0), a widely used quality of life scale for patients with chronic obstructive pulmonary disease, warrants rigorous item-level analysis to evaluate its psychometric properties, and to inform future scale refinement and potential item reduction. 321 patients with chronic obstructive pulmonary disease were investigated by QLICD-COPD (V2.0) scale. The items of the QLICD-COPD (V2.0) scale was evaluated by some statistical methods based on CTT including correlation coefficient method, variance method, factor analysis method and Cronbach’s α coefficient method. Meanwhile, Samejima hierarchical response model of item response theory was utilized to calculate the difficulty, information amount and differentiation coefficient of each item in the scale. CTT results showed that 6 items failed meet at least three or more statistical requirements, including 5 items in the general module and 1 item in the disease specific module. IRT results showed that the range of item differentiation was 1.03 ~ 1.41, which was suitable. The difficulty coefficient increases monotonously with the increase of the difficulty level (B1→B4), and some items exceed the standard value range. The average information amount of each item ranges from 0.250 to 0.546. It concluded that most items of the QLICD-COPD (V2.0) scale have good performance and good differentiation, but a few items still need further revision.