Impact of adherence to disease modifying therapies on long-term clinical and economic outcomes in multiple sclerosis: A claims analysis of real-world data

Background: Multiple sclerosis (MS) is a chronic neurodegenerative inflammatory disease that requires long-term commitment to treatment for optimal outcomes. A variety of disease-modifying therapies (DMTs) are now available that reduce relapses and delay disease progression in people with MS. However, adherence remains a significant issue, with a variety of mental, physical, and emotional factors contributing to non-adherence. In a large number of studies, non-adherence has been associated with worse clinical outcomes (relapses and disease severity), a higher economic burden, and loss of work productivity. However, many of these studies were shortterm (1-2 years) or cross-sectional studies; thus, more data are needed on the long-term clinical and economic impacts of DMT non-adherence. The objective of this study was to determine the longer-term impact of adherence to DMTs on disease activity and healthcare resource utilization (HCRU) in people with MS. The study hypothesis was that non-adherence to DMTs would be associated long-term with worse clinical outcomes and a higher economic burden. Methods: A retrospective administrative claims analysis of the US MarketScan (R) Commercial database (2011-2017) in individuals (18-65 years) with MS (based on International Classification of Disease coding) was conducted. Adherence was classified by proportion of days covered (PDC) =0.8 and non-adherence by PDC <0.8; sensitivity analyses helped further categorize as moderately (PDC =0.6-<0.8) or highly (PDC <0.6) nonadherent. Cohorts were matched using propensity score matching. Time to first relapse, annualized relapse rate (ARR), time to use of assistive devices (cane/walker or wheelchair), and annual HCRU (inpatient, emergency room [ER], outpatient, and MRI visits and costs) were compared between cohorts. Results: 10,248 MS cases were identified; 58% met adherence criteria, and 42% met non-adherence criteria. Mean follow-up from diagnosis or first DMT claim was 5.3 years. Adherent individuals had a longer time to first relapse (hazard ratio [HR] 0.83; 95% confidence interval [CI]: 0.77-0.90; p<0.0001), a lower ARR (0.13 vs. 0.18, respectively; rate ratio [RR] 0.75 [95% CI: 0.71-0.79]; p<0.0001), and longer lag times to cane/walker use (HR 0.79 [95% CI: 0.66-0.94]; p=0.0067) and wheelchair use (HR 0.68 [95% CI: 0.55-0.83]; p=0.0002) than nonadherent individuals. Adherent individuals had fewer annual inpatient and ER visits and lower total costs than those who were non-adherent (p<0.0001). Sensitivity analyses showed that differences in disease activity and HCRU were generally more pronounced between matched adherent and highly non-adherent pairs than between matched adherent and moderately non-adherent pairs. Conclusion: Significant differences in MS disease activity and HCRU were observed based on adherence to DMTs. Our study underscores the negative impact of non-adherence to DMTs on long-term clinical and economic outcomes in MS.