A comparison of traditional and Rasch cut points for assessing clinically important change in health-related quality of life among patients with asthma

Abstract

Patient-perceived change in health-related quality of life (HRQoL) domains has often been classified using a 15-point patient transition rating scale. However, traditional change levels of trivial ( - 1, 0, or 1), minimal (2, 3 or - 2, - 3), moderate (4, 5 or - 4, - 5) and large (6, 7 or - 6, - 7) on this scale have been arbitrarily defined and originally assumed that change related to an improvement was the same as that for a decline. To compare traditional and Rasch partial credit model-derived cut points and the mean changes for each change categorization when assessing clinically important change in asthma-specific HRQoL. Our sample included 396 asthmatic outpatients who completed bimonthly telephone interviews on the Asthma Quality of Life Questionnaire and transition rating items over 1 year of participation. We employed item response theory in a novel approach to identify cut points on domain-specific HRQoL change data and transition ratings. After determining natural cut points for minimal, moderate, and large differences on the transition rating anchor, we calculated mean changes under change categorizations for both improvements and declines for the two transition rating classification approaches. Although traditional and Rasch categorizations for small, moderate, and large changes slightly differed and displayed a lack of symmetry between improvements and declines, nearly all mean changes between classification approaches were comparable. In this study, traditional transition rating cut points remain suitable to assess HRQoL clinical significance in outpatients with asthma.