Integrated Care Delivery Models Managing Comorbidities and Improving

Integrated Care Delivery Models: Managing Comorbidities and Improving Care in Medicaid June 6, 2008 Melanie Bella Center for Health Care Strategies 1

Total Per Capita Costs Per Capita Medicaid Spending Percent of Medicaid Population Source: Sommers A. and Cohen M. Medicaid’s High Cost Enrollees: How Much Do They Drive Program Spending? Kaiser Commission on Medicaid and the Uninsured, March 2006.

Impact of Chronic Illness on Medicaid • High Cost – Top 3. 6% of beneficiaries accounted for nearly half of total Medicaid spending • High Need – Nearly two-thirds (61%) of Medicaid beneficiaries have one or more chronic or disabling condition – Almost half (46%) of Medicaid beneficiaries with one chronic or disabling condition have another • “Really” High Need: Dual Eligibles – 7 million dual eligibles drive 42% of Medicaid spending and 24% of Medicare spending – 87% of dual eligibles have 1 or more chronic conditions

Cluster Data Analysis: Faces II Purpose – Describe clusters of comorbidities among Medicaid recipients and the utilization and expenditure patterns associated with the clusters – Provide a description that will be useful to purchasers, plans, and providers in figuring out how to improve the care for patients with multiple chronic conditions Project – Analysis of 2001 and 2002 national, person-level Medicaid utilization and cost data – Conducted by Rick Kronick, et al at UCSD 4

Medicaid-Only Disabled, by Number of CDPS Categories 5

Top 5% Disabled, by Number of CDPS Categories 6

Key Findings • Among high-cost beneficiaries: – Virtually all have multiple chronic conditions. Within the most expensive 1% of beneficiaries in acute care spending, almost 83% had three or more chronic conditions, and over 60% had five or more chronic conditions. – Almost all have many different types of problems. Average number of diagnostic groups among highcost patients is above 5; many of these patients have cardiovascular disease, psychiatric illnesses, pulmonary problems, and many other conditions.

Key Findings • For Medicaid-only persons with disability, each additional chronic condition is associated, on average, with an increase in costs of approximately $700/month, or approximately $8, 400 per year (“super-additivity”). • Some pairs of diagnoses demonstrate strong correlations. For example, 68% of Medicaid-only disabled beneficiaries diagnosed with diabetes also have cardiovascular disease. • Identifying the most prevalent diagnostic pairs/sets of diseases (“dyads” or “triads”) holds promise for prioritizing care and developing care pathways.

Top Five Diagnostic Dyads among the Most Expensive 5% of Patients Cardiovascular–Pulmonary 30. 5% Cardiovascular–Gastrointestinal 24. 8% Cardiovascular–Central Nervous System 24. 8% Central Nervous System–Pulmonary 23. 8% Pulmonary–Gastrointestinal 23. 8% 9

Managing Comorbidities: The Medicaid Value Program (MVP) • MVP was a 2 -year $2. 8 M national initiative funded by Kaiser Permanente, with additional funding from the Robert Wood Johnson Foundation. • Ten competitively selected teams designed and tested interventions targeted at a range of comorbid conditions. • Rigorous study designs, including randomized controlled trials.

Why was MVP Important? • Traditional disease management programs often fall short in Medicaid because: – Presence of comorbidities – Need for non-medical (wrap-around) social service supports – Fragmentation of physical and behavioral health care • Core elements of effective care models: – Service integration – Multi-disciplinary care teams led by a “go-to” person – Consumer and provider engagement

MVP Innovation Teams Team Clinical Focus Target Population Complex conditions High-utilizing adults Comprehensive Neuro. Science Schizophrenia and comorbidities Adults D. C. DOH, MAA Home-based medical/social services Frail elderly MH, SA and comorbidities Adults Predictive modeling vs. HRA SSI Memorial Healthcare System Multiple chronic conditions Adults Mc. Kesson Health Solutions Diabetes and comorbidities Aged, blind, disabled Partnership Health. Plan of CA Diabetes, CHF, depression Adults Univ. of California at San Diego Diabetes and depression Latino adults Washington DSHS/Molina PH, MH, SA, LTC, DM SSI Care. Oregon Johns Hopkins Health. Care Managed Health Services

MVP Evaluation • Independent evaluation conducted by Mathematica Policy Research • Mix of qualitative and quantitative analysis • Research Questions: – What interventions did grantees implement and what were they trying to achieve? – Were grantees successful in implementation and what factors facilitated or impeded this? – Did the interventions achieve the outcomes or impacts sought? What could have made the interventions more successful? – How generalizable is the MVP experience? What was learned about the various models as well as their replicability and utility?

MVP Evaluation: Critical Factors for Implementation • Leadership commitment • Favorable environmental conditions • Staff, patient, and provider buy-in • Medicaid support and leadership • Intervention standardization

MVP Evaluation: Analysis of Outcomes • Two grantees had a rigorous design to support assessment of their impacts: Washington State, Comprehensive Neuro. Science • Easier to implement interventions than rigorously test effects: – Issues with comparison group – Small numbers – Statistical tests • Design weaknesses and/or implementation problems limited the results, but all of the interventions generated important insights on changing care processes

Key Takeaways • Efforts to integrate care across services are promising • Multi-faceted, well-targeted interventions have greater potential to affect outcomes • Intervention intensity matters • Growing interest in focusing on high-need, high-cost patients • Building an empirical evidence base is challenging • There is a critical need for rigorous evaluation in Medicaid

Additional Resources @ www. chcs. org • The Faces of Medicaid II: Recognizing the Care Needs of People with Multiple Chronic Conditions • Medicaid Value Program Evaluation, Pilot Project Case Studies and Logic Models • Subscribe to CHCS e. Mail Updates for news about CHCS programs and resources www. chcs. org