We present the largest retrospective observational study on drug-dependent risk of psychiatric hospitalization in bipolar disorders. Using the Truven Health Analytics MarketScan® administrative claims database, we analyzed 190,894 patients newly treated with one of 29 medications. The data were transformed to the OMOP common data model v5.0.1. Competing risk regression was used to compare monotherapies for the risk of psychiatric hospitalization, adjusting for patient age, sex, inpatient/outpatient status, comorbidities, and concomitant drugs.

Learning Objective 1: The attendee will be able to apply a study design based on competing risks that enables time-to-event analysis of multiple outcomes of interest, with specific application to hospitalization outcomes in persons with bipolar disorder. This approach is of particular importance for mental health informatics as an alternative to the widely used “intent-to-treat” model, which analyzes the effect of the first medication prescribed on the outcome of interest, despite significant gaps between the re-fills and high rates of drug switching/discontinuation. The participant will learn some of the benefits and limitations of the Observational Medical Outcomes Partnership (OMOP) common data model for large scale comparative safety and effectiveness research. This study is the largest and most comprehensive comparison of drugs used to treat bipolar disorder to date, and will add to the body of evidence clinicians require for evidence-based psychiatric care.


Anastasiya Nestsiarovich (Presenter)
University of New Mexico Health Sciences Center

Aurélien Mazurie, TwoFoldChange Consulting
Nathaniel Hurwitz, New Mexico Behavioral Health Institute
Berit Kerner, University of California
Stuart Nelson, University of New Mexico
Annette Crisanti, University of New Mexico Health Sciences Center
Mauricio Tohen, University of New Mexico Health Sciences Center
Ronald Krall, University of Pittsburgh
Douglas Perkins, University of New Mexico Health Sciences Center
Christophe Lambert, University of New Mexico Health Sciences Center

Presentation Materials: