Drug Benefit Restrictions: Economic Effects of Generic-Only Coverage for Patients on Statin Drugs

Abstract

Rationale: With rising pharmacy costs, there is growing interest in encouraging greater use of generic drugs. Many health plans are placing limitations on drug benefits, which include restricting coverage to only generic drugs. There is limited empirical data on the effects of generic-only policies.

Objectives: We examined the impact of generic-only drug coverage on statin drug expenditures and choice of medication in a prepaid, integrated delivery system (IDS).

Methodology: In a natural experiment with 53,583 Medicare beneficiaries in 2003-2004, we examined the effects of a new restricted, generic-only policy, initiated in 2004 for some patients on statin drug therapy for hyperlipidemia; there was one generic statin available during the study period. There were 41,330 subjects with restricted coverage and 10,253 with unrestricted drug coverage. Subjects with unrestricted coverage paid $10 for generic drugs and $15-35 for brand drugs; subjects with restricted coverage had unrestricted coverage up to $1,000 in 2003, and switched to generic-only drug coverage in 2004. We compared monthly statin drug expenditures, and patient out-of-pocket costs, using a patient random-effects approach; we used the coefficients from the models and the mean values of the covariables to find monthly adjusted estimates of expenditures in each group. All analyses adjusted for year, month, coverage-month interactions, age, gender, race/ethnicity, years of health system membership, other copayments, comorbidity, and medical center. To test the sensitivity of our findings, we repeated the analyses using a two-part model and using generalized estimating equation approach. All results yielded comparable findings.

Results: In 2004, subjects with generic-only coverage had lower annual statin drug expenditures (mean=$120/year, relative cost [RC]=0.44, 95% CI: 0.42-0.46), compared with subjects with unrestricted coverage. The two coverage groups had similar statin expenditures at the beginning of 2003, but expenditures decreased in the restricted coverage group during 2003 (the year with an annual benefit cap). There was a sharp decrease in expenditures soon after the introduction of generic-only coverage in 2004 (p<0.0001 for difference in slopes in 2003 and 2004). In 2004, expenditures were lower for brand drugs and higher for generic drugs in the restricted coverage group compared with the unrestricted coverage group (RC=0.25 for brand statins, 95% CI: 0.22-0.28; RC=1.28 for generic statins, 95% CI: 1.23-1.33). In addition, the restricted coverage group had higher annual out-of-pocket statin costs (mean difference=$29/year, RC=1.75, 95% CI: 1.56-1.95).

Conclusions: The introduction of generic-only coverage resulted in substantial reductions in statin drug costs. These drug coverage policies with substantial patient financial incentives can be highly effective in reducing pharmacy expenditures through encouraging greater use of lower cost generic drugs. Average increases in annual out-of-pocket statin costs were modest. Information and attention also is needed to insure that patients have access to and coverage for clinically necessary brand-name drugs when these are appropriate.

Keywords: Consumer and Doctor Behavior , Economic Evaluation, Health & Health Services Causality

JEL Classification: I11