Adherence, Persistence, and Expenditures for High-Cost Anti-Inflammatory Drugs in Rheumatoid Arthritis: An Exploratory Study.

Background: Drugs for inflammatory conditions are one of the highest expenditure therapeutic classes for health plans. Published literature for adherence, persistence, nonadherence risk factors, and health care costs are incomplete for newer biologic agents.

Objectives: To (a) examine differences in adherence, persistence, switch patterns, and health care costs among high-cost specialty anti-inflammatory medications and (b) suggest risk factors for nonadherence in rheumatoid arthritis.

Methods: In this exploratory retrospective cohort study, we used medical and pharmacy claims from 1.2 million enrollees in commercial health plans administrated by Premera Blue Cross, the largest not-for-profit health plan in the Pacific Northwest. We included members with rheumatoid arthritis who used the following high-cost disease-modifying antirheumatic drugs: abatacept, adalimumab, anakinra, apremilast, certolizumab, etanercept, golimumab, infliximab, rituximab, sekukinumab, tocilizumab, tofacitinib, and ustekinumab. Adherence was calculated via medication possession ratio. Persistence was calculated as the amount of days between the initial fill and final fill plus days supply. Switch rates for adalimumab and etanercept were calculated as the percentage of members who switched to another target drug during the observation period. Direct medical costs (total health care costs) and health care costs excluding specialty agents were calculated using the net allowable amount per claim for the duration of each therapy. Adherence, persistence, and costs of care were also examined for concurrent methotrexate use for the most used target drugs.

Results: The most commonly used drugs were abatacept (n = 47), adalimumab (n = 226), and etanercept (n = 252). Nonadherence in certain subgroups was associated with higher mean monthly health care costs, excluding specialty agents (etanercept cohort: +$1,063 for nonmethotrexate users; +$492 for nonadherent methotrexate users), but adherence was associated with higher total health care costs (+$883 for etanercept). Relative to specialty pharmacies, retail was associated with 9% higher nonadherence. Concurrent methotrexate use was associated with higher persistence (+307 and +192 days with adalimumab and etanercept). The most commonly switched-to drug after adalimumab/etanercept was abatacept (n = 39).

Conclusions: This exploratory study raises signals suggesting that retail pharmacies may be associated with higher nonadherence; nonadherence may be associated with increased health care costs, excluding specialty agents; adherence may increase total health care costs; and methotrexate use may be associated with increased persistence. Future research should confirm these findings.

Disclosures: This research was part of an internship awarded to Khilfeh by the AMCP Foundation/Pfizer Summer Internship Program and funded by Pfizer. Gross is an employee of Pfizer. The other authors have nothing to disclose. A portion of this research was presented at the AMCP Managed Care & Specialty Pharmacy Annual Meeting as a continuing education session entitled "The Evolving Role of Real-World Data in Health Care Decision Making" on March 29, 2017, in Denver, CO, and at AMCP Nexus 2016 as a poster on October 3-6, 2016, in National Harbor, MD.