A Statistician's Viewpoint of Responders in the Treatment of Neuropathic Pain

Abstract

In many clinical studies, the efficacy of a new drug is measured by the mean difference in primary endpoint between the new drug and an active control drug or between the new drug and the placebo. Clinicians sometimes do not believe that the mean difference in primary endpoint between two treatments is clinically meaningful for some drug indications, such as a drug for neuropathic pain associated with diabetic peripheral neuropathy. They prefer to investigate the difference between responder rates. The responder for a treatment for neuropathic pain is defined as a patient who experiences a specific level, often at least 50% in pain reduction from baseline. In this article, the disadvantages of using such a definition are discussed. Some new ideas of defining a responder are proposed. This is illustrated with a real data example.