BACKGROUND: Corticosteroids are effective for induction of clinical remission in inflammatory bowel disease (IBD), but are neither safe nor effective for maintenance of remission. Given their adverse effects and association with poor long-term outcomes, minimization of steroid use is a goal of therapy, and the use of steroid-sparing agents is indicated in patients that are steroid-dependent. We examined the patterns of steroid utilization within Canada.METHODS: Health claims data was purchased from IMS Brogan (including regional-private and national-private aggregate, and two different public plans). These databases do not contain ICD-9 codes and did not allow for differentiation of IBDs. As 5-ASA is a unique therapy for IBD, patients were indexed based on a claim for 5-ASA in the first month and followed for 12 consecutive months. We measured steroid utilization amongst 5-ASA claimants as follows: Fraction with >= 2 prednisone claims/year, fraction with a first and last prednisone claim >90 days apart, and median days elapsed between first and last prednisone claim in those with >= 2 prednisone claims. Utilization was characterized amongst different database types (public/private), genders, calendar years, age categories, and concomitant IBD therapies.RESULTS: The databases comprised 19,613 patients with a claim for a 5-ASA product in January 2011. The group was ~50% male, and 70% were aged 20-65 years old. The median number of 5-ASA claims was 8/year. The patterns of steroid utilization were generally consistent between the databases (see Table 1), with little variation across database type, gender, or calendar years. However, geriatric claimants (aged >65) in the Quebec-public database had higher rates of steroid use than the other groups. Overall, 58%-73% of patients with a claim for 5-ASA and prednisone have >= 2 claims for prednisone. Within this group, at least half the claimants (53%-69%) had >90 days between their first and last prednisone claim, with a median time between the first and last prednisone claim ranging from 103 to 222 days. Some prednisone claimants (~40%) also had claims for steroid-sparing therapies. After excluding those with any immunosuppressive or biologic claim, the rates of steroid dependence remained largely unchanged.DISCUSSION: Multiple treatment guidelines define steroid-dependence as: An inability to completely taper steroids within 90 days of initiation, frequent use of steroid courses (>1 course/year), and/or relapse within 3 months of stopping. The 3 measures of steroid utilization used in this analysis approximate that definition, although not without caveats. After excluding those with claims for steroid-sparing therapies, prednisone-claimants had a median of 2-4 prednisone claims per year. They also had a median time between first and last prednisone claim of at least 100 days, and more than 52% had a claim for prednisone >90 days apart. Any of these measures taken individually suggests a prevalence of steroid-dependence despite access to steroid-sparing therapies.CONCLUSIONS: The data suggest that >50% of patients on 5-ASA that are prescribed prednisone become steroid-dependent within 1 year. Steroid-sparing therapies appear underused in this group of patients, and show little change despite a shifting treatment paradigm.(C) Crohn's & Colitis Foundation of America, Inc.