EVENTS

Date:	Monday, March 22, 2004
Time:	9:00 AM -- 12:00 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

March 2004
The Promise and Pitfalls of Disease Management in Medicaid and Medicare

The cost of health care is skyrocketing. There is growing dissatisfaction with both health care quality and health outcomes. What can be done to reduce costs and obtain better value for our health care dollars?

Disease management (DM) is increasingly touted as the key to both improving care and reducing the costs of treating the chronically ill-the population with the highest health care costs. State Medicaid programs have begun to implement different approaches to DM, and the new Medicare legislation mandates several initiatives intended to improve the care of those with chronic illnesses. Does DM work in Medicaid, and can it work in Medicare? Two panels of experts, including the former secretary of Florida's Agency for Health Care Administration, grappled with those issues at a March 22 AEI conference.

Panel I: Does Disease Management Work in Medicaid?

Rhonda Medows
Agency for Health Care Administration, Florida

Florida's Medicaid program spent $12.7 billion in 2003 to serve 2.3 million beneficiaries.  It is the sixth largest spender and serves the fourth largest Medicaid population in the country. As in most Medicaid programs, inpatient care, nursing home care, and prescription drugs account for the highest percentage of the state's Medicaid budget. Half of Florida's Medicaid population is enrolled in MediPass-Florida's primary care case management program. This is where Florida Medicaid has focused its efforts on disease management.
 
A very small percentage of beneficiaries accounts for the majority of Medicaid spending. In fact, 52 percent of the entire program budget is attributable to just 5 percent of the population.  Included in this population are seniors; adults with congestive heart failure (CHF), HIV/AIDS, and others; and children with illnesses like cystic fibrosis or asthma-all patients with chronic illness. This segment of the Medicaid population has been the focus of cost-control efforts, both through improving outcomes to reduce utilization and through increasing the efficiency of the care delivered to them.
 
Florida began to test disease management models in Medicaid in 1997. The program was intended to serve as an additional resource for providers, as it partnered with physicians to educate patients and coordinate support services. DM aimed to improve care and health outcomes, reduce hospitalizations and ER visits, and lower overall program costs. While performance measures for diseases managed have evolved over time, the program has continually refined its objectives. The first step in launching DM was identifying high-risk populations such as patients with diabetes or HIV/AIDS, and high cost categories such as hospitalizations and prescription drugs. The original contracted vendors for certain diseases did not yield immediate cost-savings, and it was necessary to give pilot programs time to adequately ramp up before expanding the diseases targeted under them. In 2001, Florida Medicaid adopted a new approach to disease management that involved contracts with pharmaceutical companies for value-added programs. Instead of targeting a single disease, the new model recognized that patients often deal with several chronic illnesses at once. The focus shifted from short-term savings to long-term health outcomes, which would affect future utilization and costs.
 
The financing structure of DM programs has also evolved considerably. The first phase was an asthma program in which the vendor bore no risk. The second phase introduced an early generation of contracts in which vendors assumed risk for the administration of the program if its costs were not recouped in savings. Phase three was more aggressive. Vendors were at risk for program administration costs and additional savings targets. The fourth phase marked a shift not in what Medicaid was purchasing but in "in-house" initiatives that were undertaken in the areas of physician and community outreach. Florida Medicaid increased efforts to engage physicians and launched health literacy campaigns. In the latest phase, Medicaid entered into contracts with pharmaceutical companies for value-added programs. Pfizer and Bristol-Myers Squibb provided disease management programs in conjunction with their prescription drug rebate negotiations with Medicaid.
 
Three traditional DM programs have achieved net savings and endured. The HIV/AIDS program serves six thousand individuals and reaps approximately $3 million in savings each year. The CHF program has enrolled more than three thousand beneficiaries, and a pilot program for autoimmune disorders currently serves approximately three hundred.
 
The Florida: A Healthy State Program is a partnership between Florida Medicaid and Pfizer. This program is innovative for its community-based approach and its collaboration with the community, hospitals, physicians, and others. Patient outreach in this program is particularly effective because it incorporates literacy levels as well as linguistic and cultural considerations. Moreover, it contracts directly with ten safety net hospitals across the state and has care managers on-site in the hospitals. Pfizer has also contributed its own DM software to the program. The diseases targeted by the program are asthma in children; and CHF, diabetes, and hypertension in adults. McKesson Health Solutions provides a 24/7 nurse advice line. Of note is that the focus of care is the whole patient, not isolated illnesses. Nurse care managers are extremely accessible and make house calls to visit patients when needed. Fourteen thousand beneficiaries are enrolled in the Florida: A Healthy State Program, and many more benefit from the educational services it offers.
 
The Healthy Communities program provides community-based care for diabetics and people with depression. Care managers work on-site in Federally Qualified Health Centers serving Medicaid patients, and the unique Diabetes Promotora initiative trains members of the community to educate their peers. This program, funded by Bristol-Myers Squibb, has enrolled close to five hundred diabetics and almost one hundred patients with depression.
 
The success of disease-management programs truly hinges on physician, provider, and community involvement. The program should coordinate the efforts of the patient, physician, and other sources of care with the ultimate objective of educating the patient on his or her condition and appropriate care. Performance measures, once realistically outlined, should be regularly monitored and continually used to improve the system. A realistic barometer for success must maintain the goal of improving health outcomes, and not of immediate dollar savings. Long-term savings are anticipated from reduced utilization of costly acute and crisis care after patients have achieved change in health behaviors, their chronic illnesses are stabilized, and their health outcomes have improved. DM initiatives are increasingly moving away from fragmented program administration and multiple vendors and toward coordinated partnerships. Medicaid should also pursue the integration of disease management with its other interventions and program components.

Louis Rossiter
College of William and Mary

The Medicare Modernization Act of 2003 not only outlines prospects for disease management in Medicare but will also guide Medicaid's efforts in DM. A recent study I conducted with colleagues for the health policy journal Inquiry tracked asthma patients in a Medicaid disease management program in Virginia. The findings highlight the potential of DM to both improve care and contain costs.

Within the typical state Medicaid population, the aged, blind, and disabled consistently account for a disproportionate share of spending. Beneficiaries with asthma, diabetes, and CHF alone comprise just over eight percent of the patient population, but over twenty percent of Medicaid payments. This distortion makes these three diseases natural targets for DM initiatives. Comorbidities, particularly mental illness, dramatically ramp up costs.

States can take three different approaches to DM-disease management organizations, (DMO), enhanced primary care case management (E-PCCM), and programs that pay individual providers (PIP).

Virginia was the first state to introduce DM to fee-for-service Medicaid in 1996. Florida was first to issue a request for proposal (RFP) for Medicaid DM in 1998, and West Virginia was first to directly pay individual providers for improving diabetes outcomes in 2001. While most programs contract with physicians, Mississippi's PIP program pays pharmacists for DM counseling. North Carolina's E-PCCM is unique for its locally based efforts. After issuing an RFP for disease management in 2003, Texas is looking to partner with vendors who will guarantee savings for groups of target diseases. In 2002, Washington State issued two contracts under a DMO model and has already reported substantial savings. Missouri pays its providers standard, per-encounter rates to ensure that they follow up with their patients. This model may be particularly attractive to Medicaid DM initiatives. In Colorado, pharmaceutical companies fund programs centered on specific diseases. Indiana has relied heavily on state-sponsored outreach in collaboration with other public health programs. New Hampshire has issued a two-part DM contract, in which contractors guarantee savings and share the cost of an outside evaluator.

Because of recent legislation and budget resolutions, Delaware, Iowa, and New Jersey are the most topical and interesting cases in the evolution of DM programs. Twenty-five states have named staff responsible for Medicaid DM, a solid indicator of its progress. Disease management represents performance improvement in Medicaid, and returns the focus to quality and the consumer. Where the tort system has largely failed to promote high standards of care, DM has the potential to improve quality. A final incentive to pursue DM programs is that states can obtain federal matching funds for disease management as a quality improvement initiative.

Richard Vance
CorSolutions

Disease management programs are always evolving. Early DM initiatives partnered with vendors for a single disease or a single prescription drug, a narrow approach that largely ignored comorbidities. One of the first DM firms, CorSolutions at that time assumed full risk for its fees. That arrangement (turning a service company into a reinsurer) forced intense pressure to perform and achieve promised results. However, 100 percent risk is incompatible with flexibility and customization, and innovation. As the industry has changed, results are taken as a given by many customers. Risk arrangements, while still significant, have been modified so that programs could be adapted to specific clinical and business objectives.

For health plans, DM is a still perceived as a component of outpatient medical management; it is not a comprehensive medical management strategy. DM programs typically do not integrate inpatient management and provider network strategies, or control pharmacy benefit design and operations. These are separate silos within health plans. In contrast, when CorSolutions partners with employers for disease management, we are more likely to coordinate patient care on both an inpatient and an outpatient basis. Similarly in employer contracts, pharmacy and medical claims data are also typically integrated for more comprehensive care management.

Furthermore, in a typical employer population, 25 to 30 percent of employees never access their health benefits in a given year, making claims data insufficient to profile a population's health. In this environment, disease management firms undertake global population management-those with and without claims-so the assessment and improvement of population health is more comprehensive.

Disease management strategies for Medicaid need to reflect the fact that it is a very large and complex health plan-often with managed care options as subcontracts. In addition, the Medicaid population has significantly different characteristics from typical commercial populations. For example, children are a factor in employer-sponsored health plans primarily because of work-related productivity (absenteeism due to physician visits for a sick child). However, managing childhood asthma in Medicaid, while effective in improving health of the child, may not yield high financial returns. Institutionalized adults make up a significant portion of the Medicaid population, so strategies to improve care for this group both within the institutional setting and within the community may prove extremely valuable.

We are entering a new and innovative phase of Medicaid disease management. The experience of DM initiatives in health plans and employers can provide some new ideas that should guide Medicaid's efforts. Effective programs should integrate with community resources and partner closely with providers, and above all operate transparently by reporting regularly their own performance.

Panel II: Can Disease Management Work in Medicare?

Sandra Foote
George Washington University

The Medicare Modernization Act of 2003 mandates that Medicare intensify efforts in disease management. However, recalling the managed care backlash, Congress has made these parameters much less restrictive. In fact, "disease management" may not be the right term for the strategy, because it suggests that somebody is in control. DM is intended to function as a customer support program for both beneficiaries and providers. Moreover, disease management is not a well-defined program. It operates differently across populations, and we are only now beginning to identify effective strategies.

MMA sets up a new structure for disease management. Instead of outlining specific tactics, it establishes a structure that rewards improvements in performance. The Department of Health and Human Services (HHS) will contract with chronic care improvement programs (CCI) and prospectively assign them target populations. In an important departure from HHS's usual business practice, CCI payments will be tied to performance standards. The first such programs must operate by December of this year.

DM support functions in the private sector have largely been centered on beneficiary self-care. They have worked less in communication and coordination of aspects of care, but CCI organizations in Medicare have the opportunity to expand this function because they will serve much larger populations.

Evaluation of CCI programs will be critical to their evolution and success. Among many ways to measure their performance are: health outcomes, costs, adherence to evidence-based treatment guidelines, and hospitalization rates.

The biggest challenge facing CCI programs is expanding regional pilot programs to operate nationally. Medicare will undoubtedly face other hurdles in targeting viable populations, selecting CCI contractors, evaluating the programs, communicating with stakeholders, and managing expectations.

Charles Boult
Johns Hopkins University Bloomberg School of Public Health

As a physician, I have spent most of my career caring for chronically ill seniors, but in the last fifteen years, I have applied that experience to designing new ways to care for this population.

Patients with chronic illnesses experience occasional flare-ups of their conditions and call upon physicians to restore them to their baseline. As Medicare faces the aging of the baby boomers, or the "silver tsunami," the program's expenditures are projected to double. Because the chronically ill account for a disproportionate share of the Medicare budget, it is imperative that we improve the ways in which we treat that segment of the population.

Care in the twenty-first century should be proactive, managing chronic illnesses every day and not just during acute exacerbations. Beneficiaries, their families, physicians, and all other members of the system must be engaged in comprehensive care coordination. In fact, "coordinators" themselves may emerge as important providers with the function of maintaining these partnerships. In addition, more accessible and reliable information will be vital to caring for the chronically ill.

Four parallel paths will lead us to twenty-first century care. First, physicians, nurses, and other health care professionals must be educated in the ongoing care of chronic disease. Beneficiaries and their families must be motivated and empowered to manage their own care. Patients and their various providers must all partner in coordinated care, and they must have the technology to safely and efficiently share medical information.

Disease management currently focuses on a small subgroup of people afflicted with only the target disease, when, in reality, the vast majority of chronically ill patients live with several conditions. Less than 5 percent of people with congestive heart failure have only CHF. For most chronic illnesses, comorbidities are the norm. One study showed that 61 percent of people with CHF were admitted to the hospital in the previous year, but that only 14 percent of them were admitted for CHF. Because associated illnesses are so prevalent, DM programs must expand beyond a single illness to be effective.

For beneficiaries and their families, DM provides education, monitoring, and motivation. For providers, it promotes the use of clinical practice guidelines and recognition of alerts. As such, DM is not a panacea for the chronically ill, but it represents a good start.

While disease management may have potential, a few caveats deserve consideration. A dangerous misinterpretation of hospitalization rates among chronically ill patients results from equating "avoidable" and "ambulatory care-sensitive" admissions. Studies using historical controls, or "before and after" studies, can be extremely misleading. For example, a patient who dies suddenly costs Medicare nothing, but may not result from better or more cost-effective care. Furthermore, the conclusions drawn from one successful DM program cannot necessarily be extended to a new initiative. The populations can differ substantially, as can the interventions, particularly as disease management has no standard definition or protocol. As new modes of care are added to the Medicare system, we must integrate the layers and coordinate the players. A final consideration is dealing with increased exposure to legal liability. As more information becomes available to the physician, his or her legal responsibility to the patient may increase.

Disease management programs in the future should be designed to span the top fifteen chronic conditions, and it must collaborate with primary care. Their evaluation should be based on sound research designs and have valid comparison groups to evaluate their performance.

DM could improve the quality of care and health outcomes, but we must be realistic about its scope and impact. In addition to disease management, we must strive for better care coordination, improved information systems, and better professional education; and from all of these, we must demand the best quality and value.

David Kreiss
Centers for Medicare and Medicaid Services

CMS understands the risks and challenges of disease management, but the Medicare Modernization Act of 2003 (MMA) has mandated its introduction into Medicare. MMA outlines a chronic care improvement program (CCI)-a new and even more expansive tactic than disease management. CCI is an opportunity for the government to buy better performance and improve clinical outcomes, which over time will yield cost savings. While some may regard CCI as a superficial fix for fragmented health care, CMS strongly believes that it will be a building block for a better-integrated system.

Even while CMS is undertaking several major reforms, the CCI program is a definite priority among senior leadership at the agency. We are essentially creating a new line of business, one that requires that we invest in information management architecture. Accessible data is not only imperative to integrate the various components of care but also crucial to monitoring the program's performance and continually improving our strategies.

One great challenge to these new business partnerships will be adjusting the risk that CCI vendors are required to bear. The government has to be willing to pay a risk premium for asking that contractors put their fees at risk from the outset. While this risk will encourage high performance from CCI vendors, the sustainability of better outcomes in the long run will largely depend on physician engagement. It is imperative that providers participate in CCI and not perceive it as an external program.

Because of the great scope and potential of the coordinated care improvement program, this provision in MMA enjoyed overwhelming bipartisan support.

AEI research assistant Ximena Pinell prepared this summary.