Brave New World: Ideas for Reinventing How Public Behavioral Health should be Organized and Financed

1. Brave New World: Ideas for Reinventing How Public Behavioral Health should be Organized and Financed Dale Jarvis, CPA NCCBH Consultant MCPP Healthcare Consulting dale@mcpp.net www.mcpphealthcare.com

2. Session Overview • The public behavioral health system in the United States is fragmented, disorganized and under-funded. This has led to funding inequities between states, a near-inability to leverage change at a national level, and unrealistic expectations at the provider organization and individual service provider level. In this session we will explore an emerging set of ideas for transforming the public behavioral health system at the region, state and federal level and the implications for the Michigan system. This will include discussion of: • The essential ingredients of change and what’s missing in public behavioral health • An Institute of Medicine-inspired change model for reinventing the system • How P4P (Pay for Performance) can support positive changes in the system • How the federal government and the State of Michigan can support these changes • How this material is relevant to your organization and what you can do now to prepare for a “kinder and gentler” future

5. An Institute of Medicine-Inspired Change Model

6. Institute of Medicine www.iom.org I’d like to explore a change model presented in two important books from the IOM. • Crossing the Quality Chasm: A New Health System for the 21st Century (2001), is an important work that put forth a strategy for improving health care overall. However, health care for mental and substance-use conditions has a number of distinctive characteristics, such as the greater use of coercion into treatment, separate care delivery systems, a less developed quality measurement infrastructure, and a differently structured marketplace. • Improving the Quality of Health Care for Mental and Substance-Use Conditions: Quality Chasm Series (2005), builds on the first book, examining the differences between general health care and mental and substance-use conditions. This book addresses strategies pertaining to health care for both mental and substance-use conditions and the essential role that mental and substance use care plays in improving overall health and health care including the actions required of clinicians; health care organizations; health plans; purchasers; and state, local, and federal governments.

7. IOM’s Quality Chasm Series • Arguably the weakest chapter of Crossing the Quality Chasm is chapter 8, Aligning Payment Policies with Quality Improvement. • The Mental and Substance-Use Conditions edition, Chapter 8, Using Marketplace Incentives to Leverage Needed Change provides an even weaker analysis of behavioral health financing and what to do about it. • This, combined with the “50 states, 50 sets of rules” reality for public behavioral health, creates a significant gap between current reality and where we need to go. • Let’s try and do something about that…

8. IOM Strategy for Reinventing the System The 2001 IOM report described the components of an effective care system, including a supportive payment and regulatory environment, supporting organizations whose main purpose in life is to, support high performing patient-centered teams that achieve the six IOM aims for improvement (e.g. Outcomes).

9. IOM Strategy for Reinventing the System The Six IOM aims for improvement can be described as providing care that is: • Safe: avoiding injuries to patients from the care that is intended to help them. • Effective: providing services based on scientific knowledge to all who could benefit, and refraining from providing services to those not likely to benefit. • Patient-centered: providing care that is respectful of and responsive to individualpatient preferences, needs, and values, and ensuring that patient values guide all clinical decisions. • Timely: reducing waits and sometimes harmful delays for both those who receive and those who give care. • Efficient: avoiding waste, including waste of equipment, supplies, ideas, and energy. • Equitable: providing care that does not vary in quality because of personal characteristics such as gender, ethnicity, geographic location, and socioeconomic status.

10. IOM Strategy for Reinventing the System • The IOM has identified six redesign imperatives that must be addressed inside the service delivery system in order to have care systems that are able to achieve the six aims.

11. IOM Strategy for Reinventing the System • Addressing these challenges requires a fundamental re-engineering of how most community mental health centers are organized and managed and an even more fundamental shift in how CMHCs are financed. • I’d like to use the our time today to focus on creating a supportive payment and regulatory environment, making the assumption that this is a prerequisite for supporting improvements in the overall behavioral healthcare system.

12. Chapter 3: Creating a Supportive Payment & Regulatory Environment

13. Talking with Actuaries • If you were it sit down with any of the “big 3” actuarial firms and have a conversation about how much of what type of general health care service needs to be provided to a population in a specific state, region or county, they would pull up all kinds of exciting data. This would include detail down to the subspecialty level, presented as Relative Value Units (RVU) per 1,000 enrollees along with payments per RVU for the different payor types. The following table illustrates what you might see.

14. Talking with Actuaries These figures come out of vast healthcare databases that have been compiled over several decades and have been refined to the point that they can be used to develop accurate predictions of healthcare utilization and cost. Note the differences in demand across the three payor categories.

15. Talking with Actuaries • If you change the conversation and start asking similar questions about publicbehavioral health utilization and cost you will have a very different experience. • The actuary may begin a similar conversation about things like “penetration rates of 3% and 2.5 visits per user per year”, which you will quickly realize is about their commercial mental health database and they have little or no data on public sector mental health. • If the actuary does have public sector numbers to pull up, they will contain large, unexplainable variation across states for similar populations (Medicaid and indigent), and the actuary will either throw up their hands in frustration or note that the variation simply reflects reality – utilization and cost in different systems are driven by dramatic differences in policy and funding. • This has contributed to the huge variation in spending that is illustrated in the following slide.

16. Status of a Supportive Payment & Regulatory Environment (per capital mental health spending in the U.S.) Source: NASMHPD 2004 Report

17. Why the Variation? • This experience with the actuary can be traced to the fact that the public mental health system is really 50 state level systems funded by state and local general funds, federal block grant funds and the federal/state Medicaid program (historically driven by state level policy within federal parameters). States have very different financial and structural arrangements, and service modalities and definitions, for the purchase and delivery of public mental health services. • Within many states are regional authorities that have significant variation in the coverage rules, services provided, payment methods, and more, even though this is generally not looked upon kindly by CMS. On the West Coast alone California has 57 plans, Oregon 10 and Washington 13. • Roughly half of the states have federal waivers for managed care of their Medicaid mental health programs. Where the state has structured the Medicaid program into a carve-out, much of the focus has been on reduction in cost, particularly for inpatient services. Assumptions regarding utilization and cost have often been built from commercial models or the general Medicaid population rather than based on serving the public mental health target population, generally identified as adults who are seriously and persistently mentally ill (SPMI) or children/youth who are seriously emotionally disturbed (SED).

18. So What? • The problem is that, with all of this variability, the public mental health system does not have the kind of cost and utilization information that has been collected over the years by health plans regarding the delivery of services to a defined population. And the available mental health data represent a historical snapshot of a system that often lacks access to appropriate services, has services that may not be as effective as newly emerging evidence-based practices (EBPs), and services that have been fragmented, with cost shifting to and from other payors. • This makes it extremely difficult for anyone to answer the question:What would it cost to provide quality services in the public mental health system? • To do this we need to return to our actuary friends and their vast healthcare databases, because this issue has already been addressed in the general healthcare system.

19. How Healthcare Does it… • When a new medication, medical device or clinical approach comes onto the scene, it generally starts out as an experimental approach and isn’t approved for coverage by health plans. Generally led by Medicare, there is a process through which the research is reviewed and health plans adopt new practices for coverage and payment. While some practices are spelled out in benefit packages, specification of evolving improvements may also only occur at the level of the claims adjudication process. This cycle of formal approval for specification in the claims coding and adjudication process automatically ensures that the next cycle of RVUs reflect the evolving practice. • For example, in the last fifteen years there have been many improvements in cardiac care. The angiogram (enabling pictures of blockages in coronary arteries) was succeeded by the balloon angioplasty (threading a balloon into the artery and expanding it to open a blockage), which was improved by placements of stents (a scaffolding to hold the artery open at the point of blockage) to the current use of drug-eluting stents (stents with drugs that discourage plaque from sticking). As these practices emerged, they have been approved, incorporated into practice, and their utilization and cost incorporated into RVUs and DRGs.

20. How Healthcare Does it… • However, what the health care system has also discovered is that coverage of evidence-based practices alone does not ensure that they are always consistently used. For example, Medicare is tracking whether hospitals are implementing what research shows makes a big difference in outcomes for patients arriving with a heart attack.--provision of aspirin and a beta blocker. • The first report in the New England Journal of Medicine a year ago showed variability among states in delivering the appropriate therapy (from 97% of the time to 85% of the time) and among hospitals (from 100% of the time to 50% of the time). While assuring payment is a necessary condition to the delivery of research based care, it may not be sufficient. • We also know from the health care system that some procedures, while covered, may be subject to over-utilization. We should be mindful from these examples of what infrastructure will be required over time to grapple with these same issues in mental health care.

21. Public Mental Health’s Dilemma • Public mental health system stakeholders are caught in the conundrum of being unable to use historical actuarial data because, unlike general healthcare, the information rarely describes a system that has been using modern practices to meet the true needs of the population. Indeed, somewhere in the 50 states table shown above is a yet-to-be-drawn line that separates states into two groups: those that may have enough money to meet basic mental healthcare needs, and those that do not. • For those states that may have sufficient funding, there hasn’t been adequate analysis to determine whether the states are utilizing their resources in the most effective manner, nor an analysis of what services should be ramped down as recovery-oriented, evidence-based services are ramped up. The following charts illustrate the concern.

22. Public Mental Health’s Dilemma The pie charts describe expenditures for four urban Medicaid mental health plans in Washington State. Acute care includes crisis services, state hospital and community inpatient. Non-acute care includes outpatient, residential and other direct services. Even though the Medicaid capitation rates are somewhat similar in the four counties, Pierce and Spokane are having an extremely difficult time maintaining persons with serious mental disorders in the community because the amount of funding going to acute care results in an under-funding the non-acute care system. Do any of us think that we should we use the data from those two counties to set the RVUs for psychiatric inpatient and outpatient care?

23. Stop Whining and Do Something This diagram lists eight steps in a process for transforming the public behavioral health system at the region, state and federal level. Note: The last icon represents the fact that this process will require multiple change cycles.

24. Population-Based Planning • A data-driven approach to behavioral health planning is quite consistent with Actuarial Science and includes answering the following questions: • What are the target populations, what is the projected prevalence of mental illness in those groups, and how many people “ought to be” receiving service in a given year? How do these projections compare with the numbers of people served in the last three years? • Based on research and experience in this and other states, what services do clients need and how much of each type of service do they need? How do these projections compare with the service utilization over the course of the last three years? • What service capacity does the mental health system have in place to meet these needs? How well does capacity match with demand? What excesses and gaps exist by clinician type and service type? • What is the cost of meeting the projected demand (e.g. how much money “should” the state be spending on public mental health)? How do these projections compare with actual expenditures over the last three years? • These question add detail to the first three steps in the process. Some examples follow.

25. How Many People… Mental Health prevalence for a 3-County Region in rural Washington State.

26. How Many People… Substance Abuse Prevalence Forecast for a mid-sized California County.

27. How Many People… Key Point: We must adopt and begin using a standardized method for projecting how many persons require publicly funded behavioral health services in a given year. Gap Analysis for a 5-County Region in Northwest Washington State.

28. How Much Service… • More and more systems are using the LOCUS to develop Level of Care Systems. It is a national tool, with adult and child/youth versions, developed by the American Association of Community Psychiatrists to guide assessment, level of care placement decisions, continued stay criteria, and quality monitoring. It also allows system planners to understand how many low, medium and high need clients are in the system. • Until the public behavioral health system adopts methods for identifying the acuity of persons with serious mental disorders, There will be no basis for projecting how much service should be provided. • In many ways this is not unlike the DRG system used in hospitals and the Johns Hopkins University ACG Case-Mix System now being use in the medical outpatient world. Key Point: We must adopt and begin using a standardized Level of Care System with every individual that is served in the public behavioral health system.

29. How Much Service… The mental health system in Portland Oregon completed a clinical design, identified what services should be available to persons with SMI/SED as well as other Medicaid enrollees needing mental health treatment, and projected demand based on historical use, research, and projected utilization at each level. This slide projects use for non-SMI/SED persons.

30. How Much Service… This slide projects need for persons with SMI/SED in Portland Oregon.

31. How Much Service… This slide projects substance abuse service need for the mid-sized California County.

32. Chapter 4: Financial Design Considerations

33. Financing the New System • The current set of options for behavioral healthcare financing in the U.S. are stuck in one of the alternative universes know as managed-care or pre-managed care. • Although there continue to be debates about the pros and cons of sub-capitation, case rates and fee-for-service: • CMS has weighed in, telling us that, for all intents and purposes, they now set Medicaid capitation rates by counting and pricing the number of Medicaid services to Medicaid enrollees. • The rest of the healthcare world has all but abandoned non-fee-for-service payment mechanisms at the provider level. • And budget-based approaches to financing care have too few accountability mechanisms – unless they are made to look and feel like fee-for service. • What is missing from the behavioral health financing discussion are well-articulated new approaches.

34. Healthcare Financing • “Current payment methods do not adequately encourage or support the provision of quality health care… All payments methods affect behavior and quality. For example, fee-for-service payment methods … raise questions about potential overuse of services. On the other hand, capitation and per case payment methods … raise questions about potential underuse… Too little attention has been paid to the careful analysis and alignment of payment incentives with quality improvement.” (Crossing the Quality Chasm, page 17.)

35. Mental Health/Substance Use Treatment Financing • Recommendation 8-2. State government procurement processes should be reoriented to give the most weight to the quality of care to be provided by vendors. • A substantial proportion of public M/SU treatment services are purchased through government grants to local providers. These providers are frequently private nonprofit organizations that serve the population of a particular geographically defined catchment area, and are typically well-established organizations having long-standing relationships with state and local governments. Services are most commonly purchased through a system of grants. The grants are awarded subject to the provider’s meeting licensing standards and achieving specified service levels. Funding is frequently set at levels that result in patient queues - indicating excess demand for services. There are few quality-of-care standards forming a basis for accountability for these organizations. Moreover, pressures from excess demand create incentives for local providers to expand the volume of treatment even if doing so results in reduced quality. • Recommendation 8-3. Government and private purchasers should use M/SU health care quality measures (including measures of the coordination of health care for mental, substance-use, and general health conditions) in procurement and accountability processes. (Improving the Quality of Health Care for Mental and Substance-Use Conditions: Quality Chasm Series pages 316-317.)

36. Healthcare Financing • In light of these concerns, several important projects have been underway in healthcare that have direct applicability to behavioral healthcare financing. • Two fall under the rubric of Pay for Performance (P4P) – one in the California commercial insurance system; the other in the Medicare system. • Under these systems, providers are paid through the existing payment system and are offered bonuses for performance. • P4P bonuses are often related to quality and efficiency. • Quality is currently being measured through a set of “process measures”, which will be described shortly. • Efficiency (a euphemism for cost) is measured through comparisons with peers. • Both sets of measures require risk adjustment so that Provider X, who has very ill patients, is properly compared with Provider Y, who has very healthy patients. • This approach has the potential in behavioral health to combine a fee-for-service reimbursement system with a P4P system where bonuses are paid for reducing cost and improving quality. • Lets look at they two systems and then talk about a BH financing model.

37. California’s P4P Project • The California P4P project hit the streets in 2002, which was a testing year. Plans included: • Aetna (California) • Blue Cross of California • Blue Shield of California • CIGNA HealthCare of California • Health Net (California) • PacifiCare (California) • Totaling ~ 7 million HMO enrollees, 215 medical groups and 45,000 doctors

38. California’s P4P Project • Vision • The achievement of breakthrough improvement in healthcare performance. • Central Goal • The overall goal of Pay for Performance (P4P) is to significantly improve physician group performance in quality of health care and patient experience through public recognition and financial reward. • Core Principles • Collaboration (with purchasers, health plans, physician groups and consumers) • Measurement (clinical quality, patient experience, and infrastructure to support patient care) • Reward (financial incentives tied to performance results; significant and sustained to justify investment in system reengineering) • Accountability (including a public scorecard of physician group performance for consumers and providers in order to make informed choices. (Source: IHA’s February 2006 Five Year Report)

39. California’s P4P Project • P4P First Year - Measurement Set • Clinical Quality (50% weighting) • Preventive care: breast cancer screening, cervical cancer screening, childhood immunizations • Chronic care: asthma (medication), diabetes (testing), heart disease (cholesterol management) • Patient Experience (40% weighting) • Communication with doctor; timely access to care; specialty care and overall ratings of care • Investment and Adoption of IT to support patient care (10% weighting) • Point of care and population management (disease registries, electronic medical records, physician and provider reminders)

40. California’s P4P Project Clinical Measures can be captured from health plan information systems. About ½ are preventive measures and the other ½ are chronic care management measures. These are “process measures” versus “outcome measures”; e.g. cervical and breast cancer screening has been proven to reduce cost and mortality related to these diseases; they are not being scored on actually reducing cervical or breast cancer.

41. California’s P4P Project Sample Report Card for Los Angeles County, http://iha.ncqa.org/reportcard/Frames.aspx?cid=38

42. California’s P4P Project Patient Experience Measures are captured through the Consumer Assessment Survey (CAS). This slide lists the measures and provides a snapshot of the doctor-patient communication items. Payment formulas vary by plan, but are generally based on scores (e.g. 100% for being at the 75th percentile or higher, 50% for being between 50th and 74th percentile).

43. California’s P4P Project • Sample Report Card for San Diego County, Communicating with Patients.

44. California’s P4P Project Provider Groups are “rewarded” if they invest in specific technologies that have been determined to improve the quality of care. This information is based on self-reporting.

45. California’s P4P Project “Adoption of IT systems for purposes such as building patient registries for at-risk or chronically ill patients and use of electronic decision support systems at the point of care offer potential improvements in the quality of care. The physician groups who received full credit on the IT measures had average clinical scores that were nine percentage points higher than physician groups who showed no evidence of IT adoption.” (Source: IHA’s February 2006 Five Year Report, Page 17)

46. California’s P4P Project Core Principle: Reward Financial incentives tied to performance results. Significant and sustained to justify investment in system reengineering.

47. Medicare Inpatient P4P • A Medicare pay-for-performance pilot program that rewards hospitals based on quality has "steadily improved the quality of patient care," according to the latest results from the three-year program, the New York Times reports (Abelson, New York Times, 1/25). The demonstration project, which launched in October 2003, includes about 260 hospitals in 38 states (Kaiser Daily Health Policy Report, 6/22/06). Under the program, hospitals can earn bonuses if they rank among the top 20% in providing specified treatments in at least one of five areas of patient care: joint replacement, coronary artery bypass graft, heart attack, heart failure and pneumonia. According to the data, participating hospitals overall experienced nearly 1,300 fewer deaths in treating heart attack patients, and they generally have scored higher on quality measures than other U.S. hospitals. CMS officials on Friday will announce performance bonuses of $8.7 million to 115 hospitals that were the top performers based on 30 quality measures in the second year of the project. Premier, a not-for-profit hospital alliance, is managing the program. CMS and Premier have begun discussions about whether to extend the project as "Congress has also asked Medicare to look into developing a new payment system that would put more emphasis on rewarding the best care," the Times reports. (Medical News Today, 1/29/07) • Medicare’s pilot program has used a design that rewards the top 20% regardless of how well the other 80% are doing. “Hospital payments should focus on encouraging hospitals to make significant strides in improving care, and should not be based on rankings. The hospitals say they are now clustered so closely at the top that it is increasingly hard to qualify for extra payments, even if their results continue to improve.” (NY Times, 1/25/07)

48. Medicare Inpatient P4P • Why is this important? The heart is a muscle that gets oxygen through blood vessels. Sometimes blood clots can block these blood vessels, and the heart can’t get enough oxygen. This can cause a heart attack. Chewing an aspirin as soon as symptoms of a heart attack begin may help reduce the severity of the attack. This chart shows the percent of heart attack patients who were given (or took) aspirin within 24 hours of arrival at the hospital.

49. Medicare Outpatient P4P • A new pay-for-performance experiment in four states may be a precursor of Congress's intent to implement such a quality-control program throughout the Medicare system. • The Medicare program will launch a three-year pilot pay-for-performance, or P4P, program next year to encourage physicians who treat chronically ill patients to adhere to specific quality-control guidelines. • The demonstration project will compensate physicians based on the quality of care they provide to Medicare beneficiaries with chronic conditions in 800 small- or medium-sized practices in Arkansas, California, Massachusetts, and Utah. • During the first year, physicians will be paid for reporting data on quality measures. In subsequent years, the program will offer physicians annual performance-based bonuses of $10,000 per clinician and up to $50,000 per medical practice. • The program will continue to pay physician groups on a fee-for-service basis, but participating physicians will submit annual data on up to 26 "quality measures" on the care of patients with diabetes, congestive heart failure, and coronary artery disease, as well as the provision of preventive health services, such as immunizations and cancer screenings, to high-risk patients with a range of chronic diseases. Mental illnesses are not among the conditions in the program, according to the Centers for Medicare and Medicaid Services (CMS), which administers Medicare.

50. Chapter 5: Putting the Pieces Together - A Proposal for a new Design for Financing Public Behavioral Healthcare