Background/Purpose:  Adherence to disease-modifying antirheumatic drugs (DMARDs) in patients with rheumatoid arthritis (RA) is varied; rates range from 30% to 80%.¹ Adherence to methotrexate (MTX), the most commonly prescribed DMARD in patients with RA, has been incompletely studied; available data come from third-party rather than patient-level sources. We aimed to describe self-reported MTX adherence among an online community of patients with RA and to explore the patient characteristics associated with MTX nonadherence.

Methods:  US-based adult patients self-reporting RA were recruited from an online health-sharing network, PatientsLikeMe. Patients were invited to fill out a cross-sectional Web-based survey if they reported an RA diagnosis and current or past (within 6 months) treatment with MTX. Eligible patients were grouped by treatment regimen into 4 MTX groups (MTX alone, MTX with biologic DMARD[s] [bDMARD], MTX with nonbiologic DMARD[s] [nbDMARD], MTX with both bDMARD[s] and nbDMARD[s]), and they completed the 4-item Morisky Medication Adherence Scale (MMAS-4). MTX adherence was classified as high, moderate, or low based on MMAS-4 scoring criteria. Descriptive statistics were used to describe the sample, bivariable statistics were used to test for associations between MTX use and adherence, and multivariable logistic regression modeling with a backward selection procedure was used to identify factors predictive of moderate or low adherence.

Results:  Of 745 patients who viewed the invitation, 232 eligible patients participated in the survey, for an online response rate of 31%. Participants were demographically similar to nonparticipants but were slightly older (54.7 vs. 50.5 years; p<0.0001). Most survey completers (n=210) were female (90%), white (93%), and, on average, 55 years of age. Thirty-four percent (n=60) took MTX with bDMARD, 28% (n=49) took MTX alone, 25% (n=44) took MTX with nbDMARD, and 13% (n=23) took MTX with both bDMARD and nbDMARD. There was a significant relationship between MTX medication group and nonadherence; of patients taking MTX with bDMARD, 58% were highly adherent compared with 92%, 75%, and 87% in the other MTX medication groups, respectively (overall χ² p=0.0003). In the final model adjusted for demographics, years since RA diagnosis, duration of MTX treatment, and mode of administration, MTX with bDMARD use was most strongly predictive of lower levels of MTX adherence; the MTX with bDMARD group had 4 times the odds of moderate/low adherence (OR, 4.1; 95% CI, 1.9-8.6; p=0.0002) compared to those taking MTX alone.

Conclusion:  In this study, there was a strong relationship between MTX with bDMARD use and lower levels of self-reported adherence to MTX. Future research should explore how adherence is related to disease activity and should test whether these findings are replicated in other patient samples. Health care providers should be aware of this potential association and should appropriately counsel patients to optimize RA medication adherence. Reference: 1. van den Bemt BJ et al. Expert Rev Clin Immunol.2012;8:337-351.

« Back to ACR/ARHP Annual Meeting 2016

ACR Meeting Abstracts - https://acrabstracts.org/abstract/methotrexate-adherence-in-an-online-network-of-patients-with-rheumatoid-arthritis/