Better Care Coordination for High-Need Beneficiaries: What Works, and for Whom? SNP Alliance Annual Leadership Forum A National Conversation on Integration and Specialized Managed Care ” November 2, 2012 Randy Brown

 Medicare has to bend the cost curve  The best opportunity  reduce avoidable hospitalizations for chronically ill and dual eligibles (Willie Sutton, 1976)  Some extravagant savings projections for duals  What does the hard evidence tell us? The Problem and the Claims 2

1.What are the savings claims for duals? 2.Does evidence support them? 3.Targeting—one size doesn’t fit all 4.What features distinguish effective care coordination programs? 5.New initiatives being tested 6.New research to look for in March 7.A word of caution Overview 3

 Several authors claim managed care would save Medicare $10–25 billion per year –Cite no evidence or only cite evidence that is methodologically weak and for nonduals –Assume reductions in hospitalizations = savings –Eliminating 40% of preventable hospitalizations for duals would only save $3 billion per year  Medicare Advantage plans’ own costs > FFS in most counties (Biles 2012) Savings Projections for Dual Eligibles 4

 MedPAC repeatedly shows Medicare managed care plans are paid more than FFS costs  Some fully integrated plans reduced hospitalizations, but costs were higher than with FFS –Evercare (serving nursing home residents) –PACE (day centers for adults needing LTSS) –Minnesota Senior Health Options –Community Care Alliance Senior Care Option plan –Wisconsin Partnership Program  A few may yield cost savings –SCAN health plan in southern California –Disability Care Plan in Massachusetts Little Evidence that Managed Care Saves Money 5

 CBO review of 30+ programs (1/12) found little favorable evidence –Telephonic-only disease management programs didn’t work –More personal care coordination programs didn’t save enough  Other studies show some significant favorable effects—but only for high risk patients –Medicare Coordinated Care Demonstration—4 sites –Care Management Plus model (Dorr; OHSU) –Geriatric Resources for Assessment and Care of Elders (GRACE) model (Counsell) –Mass. General Hospital high cost program  Only 3 saved money The Best Evidence on Care Coordination in FFS 6

 Interventions only work for high-risk subsets  Different solutions, based on beneficiary’s needs  Need both managed care and fee-for-service models Targeting is the Key for High-Need Beneficiaries 7 Beneficiaries% of duals Managed Care Models Fee-for-service Models In nursing homes18%EvercareINTERACT II In community, using LTSS18%PACE, CCAGRACE Severe chronic illnesses, no LTSS 26%CareMoreMCCD, Mass. Gen. Less severe chronic illness38%??PGP

Care coordinators: 1.Have monthly face-to-face contact with patients 2.Build strong rapport with patients’ physicians through face-to-face contact at hospital or office 3.Use behavior-change techniques to help patients adhere to medication and self-care plans 4.Know when patients are hospitalized and provide support for the transition home 5.Act as a communications hub for providers and between patient and providers 6.Have reliable information about patients’ Rx and access to pharmacists or medical directors What Distinguishes Effective Care-Coordination? 8

 Fully integrated plans should offer the best opportunity for savings  ACA calls for reduced capitation rates to SNPs –But base capitation payment has to be “just right” –Are risk adjusters adequate? –Will plans drop out in some/many markets? –Will enhanced payments for meeting quality standards eliminate savings to Medicare? Can Medicare Reduce Costs for Chronically Ill? 9

 Models of care suggested in ACA that have promise of improved care coordination: –Patient-centered medical homes for high risk patients –Multi-payer Advanced Primary Care Practice ACOs –Comprehensive Primary Care Initiative –Medical homes –Dual eligible demonstrations to integrate care –Home health providers who offer multidisciplinary care teams (IAH)  But success will depend on how implemented –And evidence suggest net savings will be modest  Bundling models had better results, but create new silos  Structure of managed care plans offers greater potential Proposed Models in the ACA 10

 What program components are most effective for D- SNPs?  Does aligning Medicare and Medicaid coverage reduce hospitalization for dual eligibles in AZ?  Does integrating behavioral and physical health improve outcomes in PA?  Do “disability competent” managed care plans yield net savings to Medicare?  Do wait lists for HCBS lead to higher nursing home entry and hospitalizations for Medicaid beneficiaries?  Does the PACE program save money now? Evidence Coming in March from the Center of Excellence for Disability Research 11

Whether FFS, shared savings, or managed care solutions are tried:  Adopt key features of successful past programs  Focus effort on high risk patients  Steer beneficiaries to programs designed for their needs  Feed back information to programs and physicians  Build in studies of operational issues How Can We Increase Likelihood of Success? 12

1. Excessive attention to rapid cycle learning –Quick answers are often wrong answers –Takes time to learn, train, adapt, build rapport –So use intermediate outcomes and build in tests of program implementation issues (Mahoney) –Don’t sacrifice rigor of evidence for speed –Building on prior successes should shorten time to improvement 2. Lack of political will –Failure to withstand pressure from special interests will thwart attempt to save—fees/premiums have to be set low enough 3. Lack of information and incentives for providers –Physicians need data on quality and efficiency (own and others) –Payment to providers should be tied to both factors Potential Barriers to Success 13