Abstract
There is a long history of discussions about what is a minimal clinically important difference (MCID) and how this term applies to clinical research. This paper deals with a practical framework for MCID and its applicability to clinical trials. A literature review on the topic confirmed the fundamental role of MCID for the clinical research, although no guide on how to best use the MCID in clinical trials was identified. We propose a framework that takes into account (1) the definition of MCID as a term when random variable is discussed, (2) a 4-level approach for classifying the MCID evidence to be considered in a clinical development program, and (3) a method of MCID evaluation, defined in a scientifically sound protocol. The proposed framework can prompt and steer stakeholders to improve the methodological sense of clinical trials based on the definition of MCID at the level of efficacy or safety, increase the quality of data derived from clinical trials and reporting of results, and allow effective planning of drug development programs.