Estimating the Potential Medicare Savings from Comparative Effectiveness Research

Harvard Medical School, Boston MA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Comparative effectiveness research (CER) is intended to measure how alternative approaches to health care affect health outcomes. CER has the potential to reduce expenditures by limiting the use of both ineffective care and care that is no more effective than less costly alternatives. We propose to study the benefits that might accrue to Medicare from the application of CER to six high- priority medical areas. Our specific aims are to: 1. Identify areas of CER that have the greatest potential to improve the efficiency and appropriateness of care delivered to Medicare beneficiaries. 2. Estimate (long-term) incremental expenditures attributable to the use of competing strategies in these high-priority clinical areas. 3. Determine the expenditure implications for Medicare that would result from the adoption of clinical strategies supported by CER criteria, including cost-effectiveness, for each high-priority area, and to describe policy options (e.g. reference pricing, bundled payments, etc.) for implementing these findings for Medicare. Our analyses are principally based on complete Medicare claims files for Parts A and B from years 1991 to 2009 and Part D from 2006 to 2009. Data for the analysis of cancer management (e.g., localized prostate cancer), will include SEER-Medicare linked data. We will first select the medical areas for further research based upon a prioritization process considering, for each potential area of investigation, effects on total Medicare expenditures, potential for growth, and evidence of variability or uncertainty in clinical approaches. A distinguished expert Advisory Committee, representing health care, health insurance, and consumer perspectives, will provide close guidance for this component and several other aspects of the project. For each high-priority medical area selected, we will identify the relevant cohort from the Medicare claims files (or SEER-Medicare files) and assign individuals to management strategies based upon treatment or diagnostic interventions received. In addition to descriptive statistical analyses (frequency of the use of each strategy, rates of important outcomes, and total and related health expenditures), the project will include multivariate statistical analysis to investigate the relationship between strategy, outcome (including expenditures) and key variables. Where possible, we will use instrumental variable analysis and related techniques to control for bias in the selection of management strategies. Separate models will be estimated for important subgroups. We will estimate the potential consequences to Medicare of implementing our findings for each medical area by constructing a population model that shifts (with scenarios of varying degrees of adoption) patients to the strategy(ies) supported by our findings (least costly with equivalent outcomes or additional cost worth the incremental benefit based on cost-effectiveness analysis). Finally, we will investigate approaches that Medicare might use to implement these findings, including payment reform options such as bundled payments, reference pricing and value-based copayments.