How Biased are Indirect Comparisons, Particularly When Comparisons are Made over Time in Controlled Trials?

Abstract

Indirect comparisons are undertaken when a comparison is made between two regimens, usually where the regimens have never been given concurrently in any controlled trial investigating the same general patient population. We highlight the issues of making indirect comparisons when there has been a period of time between the studies, particularly when indirect comparison is being made to placebo. We discuss the impact of any bias in indirectly estimating any effect over placebo in context with noninferiority trials. © 2008, Drug Information Association. All rights reserved.