ABSTRACT
BACKGROUND: 5-aminosalicylic acid (5-ASA) is the recommended first-line treatment for active mild-to-moderate ulcerative colitis (UC) and for maintenance of UC remission. However, persistence and adherence to prescribed 5-ASAs are often suboptimal. OBJECTIVE: To evaluate 5-ASA medication use patterns and assess risk factors associated with nonpersistence and nonadherence to oral 5-ASA medications in UC patients. METHODS: IMS LifeLink Health Plan claims data (January 2007 to June 2011) were analyzed. We identified adult patients (18 years or older) with at least 1 diagnosis of UC (ICD-9-CM code = 556.x [ulcerative colitis]) and at least 1 pharmacy claim for an oral 5-ASA (balsalazide disodium, sulfasalazine, mesalamine delayed-release, and Multi-Matrix System mesalamine) during the study period. Patients were required to have continuous eligibility on both health and pharmacy plans for 6 months pre- and 12 months post-initial pharmacy claim (index date). Medication use patterns (discontinuation, time to discontinuation [days], switch, and nonadherence) in the 12 months following the index date were evaluated. Nonpersistence or discontinuation with the index medication was defined as a treatment gap ≥ 60 days. Switch was identified as patients changing to another 5-ASA product after discontinuing the index medication. Nonadherence to index medication was determined by medication possession ratio (MPR) less than 0.8 for the index medication. Nonadherence to any 5-ASA treatment was determined by a proportion of days covered (PDC) less than 0.8 for any 5-ASA. A Cox model was used to assess the relative hazards associated with discontinuation. Multiple logistic regression models were used to assess risk factors associated with nonadherence to either the index or any 5-ASA medications. RESULTS: A total of 5,664 patients met selection criteria. The median time to discontinuation of index drug differed significantly across index medications (range, 98.5 days [sulfasalazine] to 177.5 days [Multi-Matrix System mesalamine], P less than 0.0001). Patients on Multi-Matrix System mesalamine were less likely to discontinue (63.3% vs. ≥ 68.6%, P = 0.001) and more likely to adhere to their medication (MPR ≥ 0.8; 23.1% vs. ≤ 17.4%, P less than 0.0001) than patients on other medications. Patients on mesalamine delayed-release (13.8%) or Multi-Matrix System mesalamine (14.3%) had lower switch rates than the patients on balsalazide (17.2%) or sulfasalazine (17.8%), P = 0.01. Significant predictors of nonpersistence included index medication versus Multi-Matrix System mesalamine (balsalazide disodium: HR = 1.21, 95% CI = 1.07-1.36; mesalamine delayed-release: HR = 1.21, CI = 1.11-1.32; sulfasalazine: HR = 1.40, CI = 1.25-1.57), female gender (HR = 1.16, CI = 1.09-1.23), never receiving specialist care (HR = 1.14, CI = 1.07-1.21), preferred provider organization (PPO) versus health maintenance organization (HR = 1.14, CI = 1.04-1.24), and Medicare fee for service or self-insured health plan versus commercial plan (HR = 1.29, CI = 1.10-1.52). Significant variables associated with nonadherence with 5-ASA treatment (PDC less than 0.8) included not switching medication (OR = 1.90, CI = 1.58-2.29), age less than 65 (OR = 1.90, CI = 1.56-2.31), index medication as compared with Multi-Matrix System mesalamine (balsalazide disodium: OR = 1.43, CI = 1.10-1.85; mesalamine delayed-release: OR = 1.41, CI = 1.19-1.68; sulfasalazine: OR = 1.66, CI = 1.30, 2.12), female gender (OR = 1.33, CI = 1.17-1.52), residing in different regions as compared with the Midwest region (the South [OR = 1.40, CI = 1.20-1.64] and Northeast [OR = 1.29, CI = 1.05-1.58]), no use of rectal forms during the post-index period (OR = 1.28, CI = 1.08-1.50), no use of immunosuppressive/biologic agents during the post-index period (OR = 1.70, CI = 1.35-2.14), never receiving specialist care (OR = 1.25, CI = 1.08-1.44), and Medicaid/Medicare versus commercial plan (OR = 1.48, CI = 1.03-2.13). CONCLUSIONS: Patients on once-daily dosed Multi-Matrix System mesalamine had the lowest risk of discontinuation and the highest adherence rate. Multiple factors were associated with either nonpersistence or nonadherence, including multiple-daily dosed index medication, younger age, female gender, residing in the South region, PPO plan, noncommercial payer, not using immunosuppressive/biologic agents, not using rectal 5-ASA, and never receiving specialist care.
