Implementation of Medicare Part D and Nondrug Medical Spending for Elderly Adults With Limited Prior Drug Coverage

CONTEXT: Implementation of Medicare Part D was followed by increased use of prescription medications, reduced out-of-pocket costs, and improved medication adherence. Its effects on nondrug medical spending remain unclear. OBJECTIVE: To assess differential changes in nondrug medical spending following the implementation of Part D for traditional Medicare beneficiaries with limited prior drug coverage. DESIGN, SETTING, AND PARTICIPANTS: Nationally representative longitudinal survey data and linked Medicare claims from 2004-2007 were used to compare nondrug medical spending before and after the implementation of Part D by self-reported generosity of prescription drug coverage before 2006. Participants included 6001 elderly Medicare beneficiaries from the Health and Retirement Study, including 2538 with generous and 3463 with limited drug coverage before 2006. Comparisons were adjusted for sociodemographic and health characteristics and checked for residual confounding by conducting similar comparisons for a control cohort from 2002-2005. MAIN OUTCOME MEASURE: Nondrug medical spending assessed from claims, in total and by type of service (inpatient and skilled nursing facility vs physician services). RESULTS: Total nondrug medical spending was differentially reduced after January 1, 2006, for beneficiaries with limited prior drug coverage (-306/quarter [95% confidence interval CI, -586 to -$51]; P = .02), relative to beneficiaries with generous prior drug coverage. This differential reduction was explained mostly by differential changes in spending on inpatient and skilled nursing facility care (-$204/quarter [95% CI, -$447 to $2]; P = .05). Differential reductions in spending on physician services (-$67/quarter [95% CI, -$134 to -$5]; P = .03) were not associated with differential changes in outpatient visits (-0.06 visits/quarter [95% CI, -0.21 to 0.08]; P = .37), suggesting reduced spending on inpatient physician services for beneficiaries with limited prior drug coverage. In contrast, nondrug medical spending in the control cohort did not differentially change after January 1, 2004, for beneficiaries with limited prior drug coverage in 2002 ($14/quarter [95% CI, -$338 to $324]; P = .93), relative to beneficiaries with generous prior coverage. CONCLUSION: Implementation of Part D was associated with significant differential reductions in nondrug medical spending for Medicare beneficiaries with limited prior drug coverage.