The Minimal Clinically Important Difference Changes Greatly Based on the Different Calculation Methods: Response

We thank Shuntaro Tamura and coauthors for their letter to the editor and the interest shown in our work, which provided important insights on the minimal clinically important difference (MCID), a parameter increasingly reported in the literature to evaluate the results of clinical trials and determine the clinical relevance of a specific treatment. The large number of calculation methods used in this study led to a big heterogeneity of MCID threshold values (14× difference) in relation to the International Knee Documentation Committee subjective score, confirming the important implications in choosing one approach or the other to determine this type of value.1 To answer the comments of Tamura et al, it is important to underline that this study was not aimed at determining the best MCID calculation method, but at drawing attention to the limitations of the MCID approach itself, questioning the real usefulness of the MCID, as currently available, in the clinical research. Since no consensus has been reached for the most suitable approach to calculate the MCID, in this study different methodologies available in the literature were described and employed, both anchor and distribution based, each with advantages and disadvantages. Tamura et al provided one more MCID calculation approach, confirming the existence of many different methods, which are all published and used, thus creating the heterogeneous situation with regard to MCID use addressed by our study. While considering the absence of an ascertained superiority of one method over the others, in a previous letter to this journal we already stated that the anchor-based methods should be preferred, as recommended by the Food and Drug Administration and later by Terwee et al.2,3 Moreover, we criticized the current mindset that identifies the MCID with a threshold to be crossed. Given the limitations of the MCID, caution is needed when reporting and interpreting the MCID of a given patient-reported outcome measure as a measure of a treatment's effectiveness. Failure to acknowledge these limitations runs the risk of misclassifying a treatment. Thus, we confirm and strengthen our conclusions, as agreed by the other authors, that guidelines are necessary to converge on a common language to assess the results in our patients and provide a more consistent body of knowledge on how to evaluate the minimal benefits provided by the different treatments that are perceived as clinically important.