Medication adherence with disease modifying treatments for multiple sclerosis among US employees

Medication adherence in chronic diseases like multiple sclerosis (MS) plays an important role in predicting long-term outcomes, yet existing data on adherence in employee populations are not found. The objective of this study is to compare adherence among employees treated with disease modifying treatments (DMTs) for MS in the year following treatment initiation.

A healthcare claims database of US employees from 2001 to 2008 was used to identify patients with MS based on two or more DMT prescriptions or one DMT prescription with an MS diagnosis (ICD-9 340.xx). Employees continuously employed and with health plan coverage for 1 year following DMT initiation were eligible. Two measures were used in estimating adherence after DMT initiation: (1) persistence (the number of days from DMT initiation to the first 30-day gap in supply) and, (2) annual compliance, assessed by the medication possession ratio (MPR = number of days with a medication supply in the year divided by 365 days). Wilcoxon tests on time-to-event data and t-tests were used to compare persistence and MPR, respectively, between DMT groups. Other measures of resource utilization were also compared.

Overall, 358 employees [179 interferon [IFN]-β1a-IM (Avonex* = 'A'); 63 IFN-β1b (Betaseron†  = 'B'); 20 IFN-β1a-SC (Rebif‡  = 'R'); 96 glatiramer acetate (Copaxone§  = 'C')] were eligible for analysis. No significant differences in age, gender, and certain job-related variables existed between cohorts. Persistence was better for 'A' than 'B' (p = 0.039), 'C' (p = 0.0007), and 'R' (p = 0.130). At 1 year, a greater proportion of 'A' employees were persistent (60.34%) than 'B' (42.86%, p = 0.016), 'C' (42.71%, p = 0.0052), and 'R' (45.00%, p = 0.190). 'A' also had the highest MPR (0.782) which was significantly higher than 'C' (MPR = 0.698, p = 0.0160) and statistically equivalent to 'B' (MPR = 0.705, p = 0.0576) and 'R' (MPR = 0.761, p = 0.7347).

The study has limitations characteristic of administrative claims database studies and small sample sizes. The population may not be representative of undiagnosed/untreated MS patients, those not able to maintain employment, and those not using the initial therapy.

Among employees treated with 'A', 'B', 'C', and 'R' for MS, 'A' patients had significantly greater medication adherence.