Measures of effectiveness in medical research: Reporting both absolute and relative measures

Abstract

Biomedical research, especially pharmaceutical research, has been criticised for engaging in practices that lead to over-estimations of the effectiveness of medical treatments. A central issue concerns the reporting of absolute and relative measures of medical effectiveness. In this paper we critically examine proposals made by Jacob Stegenga to (a) give priority to the reporting of absolute measures over relative measures, and (b) downgrade the measures of effectiveness (effect sizes) of the treatments tested in clinical trials (Stegenga, 2015a). After exposing significant flaws in a central case study used by Stegenga to bolster his first proposal (a), we go on to argue that neither of these proposals is defensible (a or b). We defend the practice, in line with the New England Journal of Medicine, of reporting both absolute and relative measures whenever feasible.

Keywords: Absolute measures; Alendronate; Base-rate fallacy; Clinical trials; Measures of effectiveness; Philosophy of medicine; Randomised controlled trials; Relative measures.