Looking Under the Hood of MA

点击翻译按钮获取中文摘要

Project Summary/Abstract Compared to traditional Medicare (TM), Medicare Advantage (MA) health plans have more flexibility in benefit structure, network contracting, and ability to deliver coordinated and integrated care. In addition to these opportunities for potentially improving patient care, the MA program has established a rebate mechanism for passing through efficiencies to enrollees in the form of enhanced benefits. MA may therefore offer greater value to Medicare and its beneficiaries than TM. However, MA plans also engage in costly “gaming” of the risk- adjustment system without improving performance that adds to government expense. The proposed project will leverage two new sources of data. The first are encounter data submitted by MA plans, which recently became available and will be released for subsequent years on a regular basis. The second is the Risk Adjustment Payment System (RAPS) database, which contains information on clinical conditions coded in the prior year for the Centers for Medicare and Medicaid Hierarchical Condition Categories risk adjustment models. The availability of these two datasets will allow us to examine selection, health care use, quality, and risk coding though a more refined clinical lens than in our previous P01 research. Our proposal includes three aims. Aim 1 will leverage the newly available MA encounter data to quantify differences in care delivery both between TM and MA and among MA plans for patients with specific acute and chronic medical conditions. Aim 2 will assess the potential contributions of various mechanisms to the differences in care identified in Aim 1, focusing first on network formation and second on examining within- and between-provider differences in care for MA vs. TM enrollees. We also will be able to examine heterogeneity within the MA program more specifically, as health plans may adopt different approaches to network formation within an area. Aim 3 will examine the magnitude of coding intensity differences between MA and TM and among MA plans and explore alternative approaches to risk adjustment that reduce sensitivity to diagnostic coding practices in performance assessments. By describing provider networks, examining how networks vary across plans operating in the same market, and isolating within-provider differences in care for MA and TM enrollees, we will be able to assess the consistency of various conceptual explanations of care differences between TM and MA with empirical observations now made possible by the new data sources. This project is a natural extension of our current Program Project. We build on our prior work by assessing patterns of health care use and outcomes in greater clinical detail and by exploring potential mechanisms used to achieve those patterns. Understanding the role of private plans in Medicare is crucial for designing systems of financing and delivery that best serve the elderly and disabled.