Reducing Bias in Studies of Comparative Effectiveness: Applications from Mental Health

• Activity Number: 594
• Type: Invited
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Health Policy Statistics Section
• Author(s): Yulei He*+ and James O'Malley and Alisa Busch and Katya Zelevinsky
• Companies: Harvard University and Harvard Medical School and Harvard Medical School and Harvard Medical School
• Keywords: randomized trial ; observational study ; generalizability ; bipolar ; patient selection ; comparative effectiveness

Abstract

Most randomized clinical trials (RCTs) are conducted on highly selective patients, leading to questions of their generalizability for usual care populations (e.g., observational studies). Additionally, studies for the same intervention but with different study type are often conducted independently, making it difficult to understand differences between their results nor generalize their findings. The Systematic Treatment Enhancing Program for Bipolar Disorder (STEP-BD) targeted patients in usual care setting. However, some patients in the observational study arm were selected into a RCT for Acute Depression pharmacotherapy, at the end of which they could reenter the usual care setting. This provides an opportunity for studying multiple objectives including factors behind patient selection into RCTs, distinction between efficacy estimates from RCT and effectiveness estimates from observational studies, and methodology for combining inference from multiple study study types. In this talk, we provide some empirical analyses using STEP-BD data and discuss some of the implications which are relevant for comparative effectivenes research aiming to inform healthcare decisions and policy.