OBJECTIVES: Claims data collected by third-party payers contain information on medications, diagnoses, and procedure codes, but generally lack information on clinical outcomes for complicated disease states such as rheumatoid arthritis (RA). However, a validated, claims-based algorithm for estimating effectiveness of biologic agents in treating RA has recently been developed. This algorithm has been applied to various insurance databases in order to estimate the effectiveness of biologics in treating RA. The objective of this study was to compare the results of a claims-based algorithm for RA biologic effectiveness in different large retrospective databases. METHODS: In the various databases, patients were categorized as effectively treated using the claims-based algorithm if they met all of the following 6 criteria in the 12-month post-index period: (1) a medication possession ratio (MPR) ≥ 80% for subcutaneous biologics, or at least as many infusions as specified in U.S. labeling for intravenous biologics; (2) no increase in biologic dose; (3) no switch in biologics; (4) no new non-biologic disease-modifying antirheumatic drug; (5) no new or increased oral glucocorticoid treatment; and (6) no more than 1 glucocorticoid injection. RESULTS: Five research projects used the claims-based effectiveness algorithm for commercial insurance plans (Humana, Medco, Optum, Marketscan and Pharmetrics) for four biologics in RA patients. The average effectiveness across the 5 plans was estimated at 28% for abatacept, 20% for infliximab, 29% for adalibumab, and 31% for etanercept. The most common criterion not met was MPR