A step-by-step approach for selecting an optimal minimal important difference

Researchers have proposed that the minimal important difference (MID), the smallest change or difference that patients perceive as important, could aid the interpretation of patient reported outcomes measure (PROM) scores. When multiple MIDs for a given PROM differ substantially, the selection of an optimal MID to aid interpretation could prove challenging. This article describes a systematic, step-by-step selection approach developed to resolve this problem. An optimal MID, at least, should be methodologically sound and should, as far as possible, match the intended application contexts. Therefore, this approach is geared to explaining the variability of the MIDs for the PROM of interest by the methodological rigor and contextualised factors influencing the MID application, and where appropriate, provides one optimal MID (ie, the median of the selected estimates in a relatively narrow range).