Content area
Full text
The 'minimal important difference'(MID) is a little phrase with big appeal in a field struggling to interpret health-related quality of life (HRQOL) and other patient-reported outcomes (PROs). It is a deceptively simple term; a nuanced understanding of terminology and methods is needed to avoid oversimplification and misuse as the phrase gains popularity in a field looking for a simple solution to a complex problem.
This article critiques the terminology and methods of the MID, providing a historical context for the various 'how to'-focused papers, which summarize methods and provide recommendations [1-4]. It is presented in six sections, addressing this series of questions: how are various MID-related terms defined and what is their historical sequence? What is the MID used for? Why are HRQOL results difficult to interpret? How is the MID usually determined? How does the MID differ from the smallest statistically detectable difference, and how does it link clinical importance with statistical significance, sample size and power? It concludes by speculating on future directions for the MID in the field of HRQOL and PRO research and practice. The articles selected are not based on a systematic search, but on the author's personal experience, reading and a literature search that grew organically from that.
Evolution of definitions & terminology
Table 1 summarizes the evolution of MID-related definitions and terminology. In 1987, Guyatt et al. proposed the minimal clinically important difference (MCID) as the appropriate benchmark of important change against which to assess the responsiveness of an instrument or scale [5]. They did not define the MCID, and acknowledged the difficulty of quantifying it, suggesting that the change induced by an intervention of known efficacy could provide an initial estimate [5]. A total of 2 years later, in perhaps the most influential paper in MID history, the MCID was defined by Jaeschke, Singer and Guyatt as "the smallest difference which patients perceive as beneficial and which would mandate, in the absence of troublesome side effects and excessive cost, a change in the patient's management"[6]. This definition planted the MCID firmly in a shared decision-making context. In 1993, in one of the most widely cited papers on HRQOL interpretation, Lydick and Epstein commended Jaeschke et al. on their 'wonderful'definition of the MCID, but noted that they...