Enhancing the understanding of clinically meaningful results: A clinical research perspective

Abstract

Published research often address aspects related to "statistical significance" but fail to address the clinical and practical importance and meaning of results. Our main objectives in this article are to investigate the merit of common measures of Effect Size in statistical research and to highlight the importance of the simple Relative Risk ratio. In this article we present data where we consider two widely utilized effect size measures (Cohen's d and Pearson's r) in relations to relative risk. We conclude that probability analyses of risk surpass the most commonly used statistical approach used in clinical trials today and should thus be the preferred compared to the misuse and misunderstanding of reporting for instance p-values alone.

Keywords: Clinical significance; Effect size; P-value; Relative risk and risk ration; Statistical hypothesis inference testing.