Abstract

There are several concepts that define a threshold for interpreting differences in scores on a health outcome measure, including the minimally important difference (MID), the clinically important difference, the minimum clinically important difference, and the meaningful change threshold. In this Commentary, we do not concern ourselves with the definitions or differentiating features of these concepts (and, for simplicity, we use MID as an umbrella term). The similar underlying premise—that there is a difference in scores that can be considered “unimportant”—is sufficient for our exposition. Our focus is solely on whether estimates for this type of threshold are meaningful when applied to health state values. For many, publishing MID estimates for health state values is considered a worthwhile pursuit, with the rationale that understanding what constitutes an important difference is “vital for decision making and policy purposes.” The existence of this literature has inevitably led to systematic reviews that seek to synthesize and appraise methodologies associated with MID estimation, often providing recommendations for users and calls for further research to advance the field. As is apparent from the title of this article, we do not believe that this line of research has merit.