Evidence-based medicine is built on a foundation of randomized trials. Interpretation of trials depends on whether a predetermined difference (effect size) between intervention and control groups was observed and the likelihood of this being due to chance (statistical significance). It is now generally agreed that the effect size for which a trial is powered should be just large enough to be clinically significant. Introduction of the concept of minimal important difference (MID) in the 1980s was followed by an emerging stream of research aimed at assessing this quantity in a variety of disease conditions, followed by rapidly increasing use of MID among researchers and regulators.To date, however, MID methods focus on benefit magnitude alone, ignoring side effects, adverse events, and other harms. This article 1) argues that MID should be replaced with "sufficiently important difference" (SID), a concept that includes balancing of benefits against harms, and 2) outlines a few pressing research questions and challenges related to this incipient field. Sufficiently important difference was originally defined as "the smallest amount of patient-valued benefit that an intervention would require to justify associated costs, risks and other harms." SID is understood in the context of existing evidence and is measured at the individual level.
展开▼
