Policy options for making the Medicare program sustainable over the long run will have to identify approaches that reduce costs and improve the quality of care delivered. Effective interventions will be ones that target the key cost drivers in the system, chronic diseases. Since 1987, 10 percent of the growth in Medicare spending is associated with a doubling of obesity among seniors. Moreover, over half the Medicare population receives treatment for five or more chronic conditions during the year, accounting for nearly 80 percent of spending.
My earlier paper, “The Medicare Advantage Experience: Lessons for Reform to Original Medicare,” identified some best-practice approaches for prevention and care coordination derived from Medicare Advantage plans and other private-sector delivery models. Clearly the original Medicare program needs comprehensive care coordination and more effective approaches for reducing the rise in chronic disease incidence and prevalence.
This post outlines a plan to add a new Medicare option featuring evidence-based care coordination and prevention, with the goal of improving the health care outcomes for Medicare beneficiaries while reducing costs for the federal government. Others have set forth related plans, indicating that there is a momentum building behind the goal of more effectively addressing chronic disease in the Medicare program.
II. A New Option for Medicare Beneficiaries: “Medicare Integrate”
Medicare beneficiaries could choose among the original Medicare program; Medicare Advantage; or a new option, Medicare Integrate. The key elements of original Medicare and Medicare Advantage are well known. The key elements of Medicare Integrate would include: .
Access to Evidence-Based Programs, like the diabetes prevention program (DPP),that reduce the incidence of diabetes and related chronic health care conditions. Today, 50 percent of seniors are pre-diabetic and the DPP intervention has been shown to reduce the incidence of diabetes among these seniors by 71 percent.
Access to Care Coordination Services to engage chronically ill patients to keep them healthier and out of the hospital, clinic and emergency room. These services could include comprehensive medication therapy management, transitional care, health coaching, and use of a 24/7 nurse care coordinator. These services and evidence of their effectiveness are detailed in the Appendix.
Engaging Beneficiaries. Medicare beneficiaries that select the Integrate option would face no cost sharing for clinically recommended services to manage their chronic health conditions (annual eye exams, HgA1c tests, medications, etc). In addition, overweight and obese seniors who have at least one cardiovascular risk factor (pre-diabetes) would be encouraged to enroll in intensive lifestyle programs like the diabetes prevention program at no charge.
Team-Based Care. A clinical strategy designed to provide comprehensive “whole person” care. Care coordination would involve a care plan for all chronic and other medical and behavioral health conditions facing the patient.
Provider Choice. Medicare Integrate would remain similar to the original Medicare program in allowing free choice of where to receive health care services. As part of the effort to improve quality, plan vendors could also create networks of high quality providers for beneficiaries to use.
Smart Data Application. An important part of effective care coordination is timely collection of utilization data allowing ongoing feedback to providers. Comprehensive utilization data facilitate the development of risk-stratification models for targeting high-use patients. The data also allow real-time information on patients’ health status and the development of health care problems, allowing more effective ongoing medical management.
Private-Sector Partnership. The Centers for Medicare and Medicaid Services (CMS) would release a request for proposals for vendors (to include health plans, insurance companies, providers, third party administrators) to provide the full range of prevention and care coordination services. To generate savings and to ensure successful implementation at scale in the marketplace, it is important to allow contractors with past performance and capabilities to participate. CMS could use the current 15 (with future consolidation to 10) Medicare Administrative Contractor (MAC) regions for the contracting process. Contractors would be paid per member per month fee for serving patients throughout each region.
III. Financing the Integrate Option — Shared Savings
Similar to the financial incentives built into the accountable care organization model, Medicare Integrate plan vendors would split savings with Medicare only once a pre-determined savings threshold is achieved for the Medicare program over a defined time period and quality metrics are met or exceeded. Savings beyond the pre-determined threshold would accrue to the vendor, while savings short of the pre-determined threshold would be paid back by the vendor to CMS.
Quality metrics used in the Medicare Integrate option should be similar to those used by Medicare Advantage and should be extended to include original Medicare. To the extent possible, these metrics should focus on clinical outcomes, be adaptable to new technologies, and act as a counterbalance to short-term cost-saving activities that do not improve health long term.
Adding Medicare Integrate as a third choice for seniors could be accomplished in short order. Having both prevention and care coordination as an integral part of care delivery is what baby boomers entering into the Medicare program expect; indeed it is the type of care they are receiving through private plans prior to Medicare. Adding the option would provide better continuity of care, as those turning age 65 with chronic health care conditions could continue to have their care coordinated.
Both the diabetes prevention program and care coordination functions are readily scalable and replicable. The capacity for care coordination already exists in the marketplace in the form of home health firms, health plans, and specialty population management vendors, among others. Moreover, the functions performed by the care coordination teams are based on both published randomized trials and actual best-practice experiences from the private sector. These data show that effective care coordination and the use of preventive lifestyle models like the diabetes prevention program reduce spending and improve the quality of care delivered.
However, implementation of these models needs to be particularly sensitive to beneficiaries who are the sickest of the sick — those dually eligible for Medicare and Medicaid. For these patients a slower implementation schedule may be wise to assure that any changes result in better outcomes. In this case, we have much to learn from the best practices of special needs plans in the Medicare Advantage program of how best to design care coordination that is quality improving.
APPENDIX: KEY IMPLEMENTATION STEPS
I. Transition Away from Fee-for-Service (FFS) in Medicare
A key to introducing care coordination into traditional Medicare is to transition away from fee-for-service payments and, as a start, replace them with more bundled payments for Part A and B covered services. The incentives to increase the volume of services in FFS run completely counter to the incentives to provide clinically effective care coordination. As FFS is phased out over time, it would be replaced by bundled payments for (most) hospital admissions that include all covered post-acute care services 30 days after discharge.
There is broad agreement that Medicare’s FFS payment model (notably the sustainable growth rate (SGR) physician payment formula) is outdated, drives up the volume of services, and must be replaced to improve health care delivery. Our entire health care system is built around FFS, and updating our current health care delivery structure will set the stage for an innovative, high-quality health care system. However, transitioning away from FFS will not be easy and will not happen overnight; reforming the Medicare system so that it pays for quality will require significant data collection and monitoring, updates to regulations, and testing and scaling of new and innovative payment models and incentives. Advancing these objectives and facilitating a gradual shift from FFS medicine will take time, and will therefore likely occur in stages and lead to a number of new payment model reforms.
The experience from the Medicare Participating Heart Bypass Center Demonstration, which made bundled payments to cover all inpatient hospital and physician services for coronary artery bypass graft surgeries, show the potential for substantial savings. Bundled payments in this case reduced Medicare expenditures on heart bypass surgeries by about 10 percent. Evaluations of the demonstration also found slight improvements in the quality of care provided using measures included in the study.
The Affordable Care Act includes some payment reforms that move in the right direction. The Act includes a bundled payment demonstration that started in January 2013. To date, CMS has outlined four different payment models that include a single payment for an episode of care. To start, CMS has identified 48 different episodes that would be included in a combined bundled payment to 500 different providers. Typically, these bundles include Medicare-covered services (acute inpatient hospital; physician services delivered in and outside the hospital; outpatient hospital services; emergency room services; and post-acute services such as physical therapy) for each of the 48 care episodes. (Part D drug spending around the episode is not included.)
Over the next five years Medicare should expand the bundled payments to virtually all episodes of care. During this time, CMS could quickly evaluate the quality and spending performance of providers for each of the four payment models and recommend adopting either one or two of the most effective ones (ideally the more comprehensive models that include more Part A and Part B services). In the interim, providers would be paid on a mix of FFS and bundled payments.
Building quality metrics into the payment bundles will be essential. These quality metrics should recognize the fact that over half of all Medicare patients are under medical treatment for five or more conditions, accounting for nearly 80 percent of Medicare spending. In addition, quality metrics that are developed should continue to move the Medicare program toward consistent measures throughout the program with a focus on clinical and patient-centered outcomes. Measures will also need to preserve the adaptability needed for the adoption of medical advances and new technologies. Finally, measures should serve as a counterbalance to encourage quality improvements that may be counter to near-term cost savings, such as activities that promote wellness and prevention that reap benefits over time.
Several integrated health centers (Marshfield Clinic, Geisinger), demonstrations, and health teams in Vermont and North Carolina have produced lower spending and improved quality. A common feature of all of the above is the use of team-based care. Team-based care occurs when health care professionals (physicians, nurses, pharmacists, social and mental health workers, and nutritionists, among others) work collaboratively to achieve shared health goals with patients and their families. The health team works in close collaboration with the physician practices, accountable care organizations, and hospitals. In some models a nurse care coordinator is embedded in the teams and in others the teams meet on a regular basis to track quality metrics and outcomes. The teams essentially provide “whole person” care and do not focus on disease specific care. Some of the functions that have proven to improve quality and reduce spending are outlined below.
II. Functions Performed by Health Teams
Health teams would provide the following evidence-based functions when coordinating care. .
Coordination of care for all covered Medicare services utilizing a team-based approach
Approaches that provide a “whole” person focus on preventing disease and managing acute, chronic, and mental health services
Medical advice from a care coordinator available 24/7
Assessment of patient risk and development of an individualized care plan
Comprehensive medication management
Transitional care and health coaching
Regular contact with enrollees
Close integration of the care coordinator nurse and primary care (and specialist) physicians
Evidence-based health coaching to train patient self-management skills and facilitate behavior change.
Each of the major functions outlined above (transitional care, medication adherence, health coaching) have several published randomized trials showing they individually result in improved health outcomes at lower levels of health care spending. Collectively they serve as a powerful, team-based approach to generate substantial proven savings and improved quality of care.
Team-based care has been introduced at the state level (Vermont and North Carolina) and increasingly plays a major role in most delivery systems. There is emerging evidence from these states and individual practices that team-based care can produce better quality at lower cost. Milliman estimates savings to North Carolina’s Medicaid program of up to 15 percent. Recent evaluations of the community health teams in Vermont also are showing promising reductions in costs. A brief summary of some of the randomized trials highlighting the clinical effectiveness and cost savings associated with these care coordination functions is presented below.
Two of the best known models of transitional care have been developed by Eric Coleman at the University of Colorado and Mary Naylor at the University of Pennsylvania. The team at Penn defines transitional care as providing “comprehensive in-hospital planning and home follow-up for chronically ill high-risk older adults hospitalized for common medical and surgical conditions.” The heart of the model is the Transitional Care Nurse (TCN), who follows patients from the hospital into their homes and provides services designed to streamline plans of care, interrupt patterns of frequent acute hospital and emergency department use, and prevent health status decline.
While TCN is nurse led, it is a multidisciplinary model that includes physicians, nurses, social workers, discharge planners, pharmacists, family caregivers, and other members of the health care team in the implementation of tested protocols with a unique focus on increasing patients’ and family caregivers’ ability to manage their care. For the millions of Americans who suffer from multiple chronic conditions and complex therapeutic regimens, TCM emphasizes coordination and continuity of care, prevention and avoidance of complications, and close clinical treatment and management — all accomplished with the active engagement of patients and their family and informal caregivers and in collaboration with the patient’s physicians. More information is available here.
A second model, developed by Eric Coleman, uses transition coaches to train patients and family caregivers how to manage their care. Transition coaches are generally not physicians but are nurse practitioners, nurses, or community health workers. To smooth transitions from hospital to home, the Care Transitions Intervention (CTI) uses coaching and home visits by trained care coordinators. The coach makes one home visit and several phone calls to the patient over a 30-day window. More information on this program is available here.
According to randomized trials, both programs reduce dramatically hospital readmission rates. Among Medicare patients, the CTI program reduced 30-day readmissions by 30 percent and hospital costs at 90 days by 25 percent. Randomized trials of the TCN model have demonstrated reductions in readmissions of 56 percent, with similar reductions in total Medicare spending after one year.
Under the Part D program, drug plans must offer medication therapy management program (MTM). However, the criteria for targeting Medicare beneficiaries enrolled in Part D plans are those with multiple chronic conditions (maximum of 3 required for targeting) and expected annual drug spending for 2013 of $3,144. However, the current MTM program would not include patients with high Part A and B medical costs who may not be appropriately taking medications (non-adherent, etc) and would not hit the $3,144 spending threshold.
As part of the new care coordination services in traditional Medicare, the current MTM program should be broadened and integrated into the overall set of care coordination services provided. A pharmacist working as part of the care coordination team would work with patients who have high prior-year total Medicare spending (not just those with high Part D spending) to resolve drug therapy issues (drug effectiveness, dosage, compliance, and adherence). This broader approach would, as part of the overall care coordination team, link medication management and resolving drug therapy problems to clinical improvements in seniors. Substantial work has already been completed on the design of such a benefit from the Patient-Centered Primary Care Collaborative and the Agency for Healthcare Research and Quality Innovation Exchange Quality Toolkit.
Health Coaching and Patient Literacy
Coaching empowers patients with one or more chronic conditions to understand and follow through on their care plan, participate in shared decision making with their health care providers, and more effectively navigate the health care system. Understanding the care plan and working to consistently execute it are elements critical to reducing unnecessary utilization of health care services. Effective coaching empowers individuals with a wide range of conditions, including but not limited to chronic conditions, to participate in medical treatment decisions with their doctors.
Coaching would be another key component of care coordination services provided in Medicare Integrate.A large randomized trial conducted by Health Dialog and published in the New England Journal of Medicine utilized telephonic health coaching to work with a large population (more than 174,000—7,000 of whom were Medicare patients) of patients. This recent randomized trial showed that total health care spending was 3.6 percent lower in the treatment group (yielding about a 3 percent net savings after accounting for the cost of the intervention). This single component of care coordination alone reduced hospitalizations in the trial by 10 percent and total spending by more than 3 percent.
There is considerable evidence that these care coordination functions improve outcomes and reduce spending. Randomized trials evaluating transitional care models have found 50 to 60 percent reductions in 30-day hospital readmission rates among Medicare patients. A recent randomized trial found that comprehensive medication therapy management in Minnesota produced lower total healthcare costs (over 31 percent in one year) while significantly improving clinical outcomes with a 12:1 return on investment.
States could also use the health teams to manage Medicaid patients and those dually eligible for both Medicare and Medicaid. Section 2703 of the Affordable Care Act provides states a 90 percent match for eight quarters toward the cost of establishing community health teams that provide the care coordination services above.
Adding evidence-based lifestyle programs and health teams to provide care coordination services would target two of the major issues facing Medicare: rising rates of chronic disease prevalence and the need to coordinate care for chronically ill patients throughout the Medicare program. The DPP (discussed below) and community health teams could be replicated and scaled quickly throughout the country. Both changes would improve health outcomes and have tremendous potential to reduce health care spending.
Diabetes Prevention Program
The diabetes prevention program is an intensive lifestyle protocol that includes a 16-lesson curriculum covering diet, exercise, and behavior modification delivered over a 24-week period. The program targets overweight and obese pre-diabetic adults. The goal of the intervention is to achieve and maintain weight loss of 7 percent of starting body weight. Large randomized trials have shown that the intervention reduced the incidence of diabetes by 58 percent and among seniors by 71 percent. Even with some weight regained, recently published results of a 10-year follow-up study of the original trial found that adults in the intensive lifestyle program continued to maintain around five pounds of weight loss and those aged 60 and older maintained longer term weight loss of just under 11 pounds. Diabetes incidence in the 10 years since DPP started was reduced by 34 percent.
Community-based adaptions of the DPP protocol by the YMCAs and other groups have found, in initial studies, similar reductions in weight, approximately 6 percent. The community based adaptation of the model reduces the per enrollee costs substantially to around $350 for each 16-week class. As a result, tabulations show, on the conservative side, that adopting the DPP into the Medicare program would produce net budget savings totaling several billion dollars within the 10-year scoring window.
III. Key Design Issues in Incorporating Care Coordination into Medicare through Medicare Integrate
CMS would develop a request for proposals, and contractors that can provide the full range of care coordination functions would be eligible to bid. Bidders would be expected to complete a needs assessment of service gaps facing patients and primary care providers, as well as addressing how to coordinate existing services (public health nurses, etc) with the new care teams and how to provide the full range of care coordination services outlined above. Eligible entities would have to meet certain financial reserve/capacity tests and would carry medical malpractice liability insurance. The awards would be for three years to parallel the Medicare Shared Savings Program.
Bids would be on a per-member per-month basis based on initial estimates of Medicare patients receiving care coordination and reconciled at year end. As such, contractors would not be risk-bearing entities and would not face state insurance rules and regulations. Contracts would be renewed every three years based on meeting quality and spending targets similar to those outlined under the shared savings program for accountable care organizations.
Firms providing care coordination services could be the same administrative contractors that process claims or other firms (i.e. home health and population health management firms) that would partner with the Medicare Administrative Contractors.
Any Medicare provider could contract with the health teams at no fee or charge. Provider practices, accountable care organizations, hospitals, health systems, or other providers using the teams would be expected to report quality metrics similar to those expected of Medicare Advantage plans. Participating providers would receive a small $2-$3 per-member per-month payment for each patient using the health teams.