The Use and Abuse of Multiple Outcomes in Randomized Controlled Trials of Depression

• Activity Number: 380
• Type: Contributed
• Date/Time: Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Kristin M. Tyler*+ and Nicholas Horton
• Companies: Smith College and Smith College
• Keywords: Multiple outcomes ; Multiple comparisons ; Bonferroni adjustment ; Clinical trials ; Composite outcomes

Abstract

Multiple outcomes are commonly collected and analyzed in randomized clinical trials. However, a fully effective and efficient method for analysis is not yet established. The most common analysis methods for multiple outcome data involves separate testing of each individual outcome, global testing of all outcomes or single testing of a composite outcome. We describe and report the results from a comprehensive review (n=XX) of multiple outcomes from depression trials reported in 2007 from 4 widely read journals. Parallel data collection was undertaken for each paper to determine the number of primary and secondary outcomes, as well as the methods used for analysis. We found that the use of multiple outcomes was quite common, yet the methodology utilized to synthesize these reports was often quite simplistic. Further study of ways to improve trial methodology in this area may be warranted.