Psychometric Evaluation of the Assessment of the Burden of COPD Scale According to the Rasch Model

To facilitate shared decision-making in Chronic obstructive pulmonary disease (COPD), the Assessment of the Burden of COPD (ABC) tool was developed. An integral part of the ABC tool is the ABC scale, a questionnaire to determine COPD burden. The psychometric evidence of the ABC scale is limited. If the data would fit the Rasch model, more accurate scoring would be possible. This would improve the use of the ABC tool. Therefore, the aim of this study is to validate the ABC scale using the Rasch model. We conducted a secondary analysis based on data collected in a pragmatic RCT using Rasch model theory. The ordering of thresholds, reliability and targeting were analysed, and we tested unidimensionality, local independence of items (LD), fit of the data with the model and measurement invariance between groups (DIF). We analysed 551 samples from 176 patients. The analyses confirmed that the ABC scale is a reliable instrument. However, initial analysis showed a poor fit of the data to the Rasch model. After recoding the response options and analysing the fit of the data for the subdomains of the ABC scale, we found ordered thresholds and a good overall model fit. This analysis indicates that the ABC scale assessed three distinct sub-domains: symptoms, physical functioning, and emotional health. We conclude that the ABC scale is a valid instrument to assess COPD burden. While the scale is considered suitable for use in clinical practice and research, we suggest rewording the response options to ensure uniform understanding by respondents.

The used data were derived from a previously conducted trial (NTR3788).