Medicare beneficiaries and the impact of gaining prescription drug coverage on inpatient and physician spending

UMMS Affiliation

Meyers Primary Care Institute; Department of Medicine, Division of Geriatric Medicine

Publication Date


Document Type



Aged; Aged, 80 and over; Centers for Medicare and Medicaid Services (U.S.); Fees, Pharmaceutical; Female; Health Care Surveys; Health Expenditures; Hospitalization; Humans; Insurance Claim Review; Insurance, Pharmaceutical Services; Male; Medicare; Middle Aged; Models, Econometric; Office Visits; United States


Health Services Research | Life Sciences | Medicine and Health Sciences | Primary Care


OBJECTIVE: To assess whether gaining prescription drug coverage produces cost offsets in Medicare spending on inpatient and physician services.

DATA SOURCE: Two-year panels constructed from 1995 to 2000 Medicare Current Beneficiary Survey, a dataset of Medicare claims and health care surveys from the Medicare population.

STUDY DESIGN: We estimated a series of fixed-effects panel models to calculate adjusted changes in Medicare spending as drug coverage was acquired (Gainers) relative to the spending of beneficiaries who never had drug coverage (Nevers). Explanatory variables in the model include age, calendar year, income, and health status.

PRINCIPAL FINDINGS: Assessments of inpatient and physician services spending provided no evidence of overt selection behavior prior to the acquisition of drug coverage (i.e., there were no preswitch spikes in Medicare spending for Gainers). After enrollment, the medical spending of Gainers resembled those of beneficiaries who never had drug coverage. Overall, the multivariate models showed no systematic postenrollment changes in either inpatient or physician spending that could be attributed to the acquisition of drug coverage.

CONCLUSIONS: We found no consistent evidence that drug coverage either increases or reduces spending for hospital and physician services. This does not necessarily mean that drug therapy does not substitute for or complement other medical treatments, but rather that neither effect predominates across the Medicare population as a whole.

DOI of Published Version



Health Serv Res. 2005 Oct;40(5 Pt 1):1279-96. Link to article on publisher's site

Journal/Book/Conference Title

Health services research

Related Resources

Link to Article in PubMed

PubMed ID