OBJECTIVES: Adherence comparisons of biologic agents used to treat Crohn’s disease (CD) have typically compared adherent and non-adherent patients.  Non-adherence may be due to efficacy, tolerability, lack of follow-up or patient-centered reasons.  There is a need for a more in-depth examination of non-adherence.  Purpose was to compare healthcare costs of CD patients who were adherent vs. intermittently adherent on infliximab (IFX) therapy. METHODS: Patients with >1 claims for IFX initiated between 1/1/2006-12/31/2009 who had >2 diagnoses of CD (ICD-9-CM: 555.XX) during the pre-index period were identified from Thomson Reuters Marketscan® Databases.  Patients had to be >18 years, continuously enrolled for 12 months before and after IFX initiation, and had no prior use of IFX during 360-days pre-index.  Patients with prior biologic therapy or rheumatoid arthritis (ICD-9-CM: 714.XX) were excluded.  Adherent group was classified as having a medication possession ratio (MPR) of >80%; intermittently adherent group had an MPR<80% with IFX claims spanning 90% or more of the observation period.  Differences between the adherent and intermittently adherent groups were assessed using propensity-weighted general linear models. RESULTS: A total of 599 CD patients were identified (360 adherent; 239 intermittently adherent) with a mean (SD) age of 42.9 (15.5) and 50.9% were female.  Propensity-weighted mean total healthcare costs excluding IFX were $13,097 vs. $20,068 (P<0.0001) for the adherent vs. intermittently adherent groups.  Mean all-cause component costs were $2,764 vs. $6,665 (P<0.001) for hospitalizations, and $7,074 vs. $10,196 (P<0.0001) for outpatient visits among the adherent vs. intermittently adherent groups, respectively; no significant cost differences were observed in ER visits, other prescription or total costs (component + IFX). CONCLUSIONS: Adherence was associated with lower total non-IFX healthcare costs.  Comparison of adherent vs. intermittently adherent patients may provide a conservative test of the effect of adherence since patients that were intermittently adherent have some IFX exposure.