Medication possession ratio: implications of using fixed and variable observation periods in assessing adherence with disease-modifying drugs in patients with multiple sclerosis

BACKGROUND The purpose of this study was to compare two methods of adherence calculation using administrative data for patients with multiple sclerosis (MS) who are prescribed disease-modifying drugs. METHODS Pharmacy-billed disease-modifying drug prescription claims were selected from the 2007-2008 LifeLink™ Health Plan Claims Database. The index date was the first disease-modifying drug prescription claim. Two cohorts were created: all patients with a disease-modifying drug claim in 2007 and a subset with continuous eligibility for 12 months post-index. Adherence was calculated across all disease-modifying drugs for 12 months post-index. Medication possession ratios (MPRs) with variable (start to end of therapy) and fixed (365 days) duration denominators were calculated. Variable MPR was calculated by summing days supply from the first to the last prescription (inclusive) divided by time between the last prescription date plus days supply and the first prescription date. Variable MPR was evaluated for all patients and the continuously eligible cohort. Fixed MPR used the same numerator but divided by 365 days of follow-up and evaluated only for the continuously eligible cohort. RESULTS There were 3405 patients with MS and a disease-modifying drug claim in 2007 and 2145 in the continuously eligible cohort. Means for variable MPR ranged from 87.5% ± 16.6% for the continuously eligible cohort to 90.5% ± 16.0% for the 2007 cohort. The comparable value for fixed MPR was 78.0% ± 28.2% for the continuously eligible cohort. Fixed MPR gave a consistently lower rate of adherence than variable MPR at an 80% adherence threshold. CONCLUSION Different adherence measures can yield different outcomes, especially when using different eligibility criteria. These results demonstrate the importance of full disclosure of methods used for calculations and specification of the study population.