Key points are not available for this paper at this time.
OBJECTIVE: We sought to evaluate the relationship between different levels of medication possession ratio (MPR) attained and achievement of clinically meaningful reductions in lipid levels. RESEARCH DESIGN AND METHODS: This was a retrospective cohort study of 4691 new statin users from the Department of Veteran Affairs (VA). Subjects were required to be eligible for VA medical and pharmacy services throughout the 1 year study period from index date and to have complete data for exposure, outcome, and adjustment variables. MPR was defined as number of days supplied with prescription medication divided by days of observation. MAIN OUTCOME MEASURES: Achieving 25% or greater reduction from baseline in lipid levels for three lipid outcomes: non-high-density lipoprotein (non-HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, and total cholesterol (TC). RESULTS: We observed a statistically significant trend of an increasing proportion of study subjects achieving a 25% reduction or more for all three lipid outcomes (p-values <0.001 for each of the three outcomes using the Cochran-Armitage trend test). Using multiple logistic regression, odds ratios (ORs) for each of the three outcomes were at a maximum for the 0.9-1.0 MPR category with ORs of 12.90 (95% confidence interval (CI), 9.60, 17.35) for the non-HDL cholesterol outcome; 11.29 (95% CI, 8.61, 14.80) for the LDL cholesterol outcome; and 9.11 (95% CI, 6.62, 12.53) for the TC outcome. Direct comparison of the 0.9-1.0 MPR category versus the 0.8-0.89 MPR category demonstrated an increase in odds of achieving 25% or more reduction for all three lipid outcomes. CONCLUSIONS: We conclude that significant improvements in outcomes are achieved with higher MPR thresholds than commonly targeted. The authors propose consideration of an MPR-adherence threshold of 0.9 MPR. Limitations include the possible modification of study findings in non-VA settings. MPR is a secondary adherence measure based on refill frequency.
Watanabe et al. (Tue,) studied this question.