Community pharmacy and mail order cost and utilization for 90-day maintenance medication prescriptions.

BACKGROUND Pharmacy benefit management (PBM) companies promote mail order programs that typically dispense 90-day quantities of maintenance medications, marketing this feature as a key cost containment strategy to address plan sponsors' rising prescription drug expenditures. In recent years, community pharmacies have introduced 90-day programs that provide similar cost advantages, while allowing these prescriptions to be dispensed at the same pharmacies that patients frequent for 30-day quantities. OBJECTIVE To compare utilization rates and corresponding costs associated with obtaining 90-day prescriptions at community and mail order pharmacies for payers that offer equivalent benefits in different 90-day dispensing channels. METHODS We performed a retrospective, cross-sectional investigation using pharmacy claims and eligibility data from employer group clients of a large PBM between January 2008 and September 2010. We excluded the following client types: government, third-party administrators, schools, hospitals, 340B (federal drug pricing), employers in Puerto Rico, and miscellaneous clients for which the PBM provided billing services (e.g., the pharmacy's loyalty card program members). All employer groups in the sample offered 90-day community pharmacy and mail order dispensing and received benefits management services, such as formulary management and mail order pharmacy, from the PBM. We further limited the sample to employer groups that offered equivalent benefits for community pharmacy and mail order, defined as groups in which the mean and median copayments per claim for community and mail order pharmacy, by tier, differed by no more than 5%. Enrollees in the sample were required to have a minimum of 6 months of eligibility in each calendar year but were not required to have filled a prescription in any year. We evaluated pharmacy costs and utilization for a market basket of 14 frequently dispensed therapeutic classes of maintenance medications. The proportional share of claims for each therapeutic class in the mail order channel was used to weight the results for the community pharmacy channel. Using ordinary least squares regression models, we controlled for differences between channel users with respect to the following confounding factors: age, gender, presence or absence of each of the top 11 drug-inferred conditions (e.g., asthma/chronic obstructive pulmonary disease, cardiovascular disease), drug mix, and calendar year. We calculated estimated predicted means holding all covariates at their mean values. For both 90-day dispensing channels, we calculated number of 90-day claims per member per year (PMPY) and cost per pharmacy claim, with all claims counts adjusted to 30-day equivalents (i.e., number of 90-day claims × 3). Differences were compared using t-tests for statistical significance. RESULTS Of 355 PBM clients prior to exclusions, 72 unique employers covering 644,071 unique members (range of approximately 100 to more than 100,000 members per employer) were included in the analysis. On an unadjusted basis, community pharmacies represented 80.8% of 90-day market basket claims (in 30-day equivalents: 3.97 claims PMPY vs. 0.95 in mail order) and 77.2% of total allowed charges. After adjustments for therapeutic group mix and patient characteristics, predicted mean pharmacy claim counts PMPY were 4.09 for community pharmacy compared with 0.85 for mail order (P  less than  0.001). Predicted mean allowed charges per claim for community and mail order pharmacies did not significantly differ ($49.03 vs. $50.04, respectively, P = 0.202). CONCLUSIONS When offered maintenance medications through community and mail order pharmacies on a benefit-equivalent basis, commercially insured employees and their dependents utilized the community pharmacy channel more frequently by a margin of more than 4 to 1 in terms of claims PMPY. Overall allowed charges per claim for community and mail order pharmacy did not significantly differ.