Blog Home

«
»

Creating the Virtual Integrated Delivery System



October 5th, 2009

Preventing and more effectively managing chronic illness are critical national health priorities. Patients with chronic disease currently account for three-quarters of overall health spending. Multiple morbidities are common: More than half of Medicare beneficiaries are treated for five or more chronic conditions yearly. Nine chronic ailments account for nearly 60% of the recent rise in Medicare spending — conditions typically treated with prescription drug therapy and managed chiefly by patients at home and in outpatient settings. Despite significant health care outlays, though, chronically ill patients receive just 55% of clinically recommended services, and that gap in care may explain a nontrivial portion of preventable morbidity and premature mortality.

In response to these trends, chronic disease management programs have proliferated in the private sector and are common in the Medicaid and Medicare Advantage programs. They are notably absent in traditional fee-for-service (FFS) Medicare — a crucial lacuna, given that 81% of Medicare beneficiaries are enrolled in FFS Medicare and account for about 79% of the program’s overall health care spending.

Reports in the peer-reviewed research and practice literatures, along with a growing body of anecdotal evidence, demonstrate the potential for medical homes to deliver patient-centered, coordinated care; improve quality, outcomes, and efficiency; and, concomitantly, reduce health care spending. Large integrated delivery systems such as Geisinger and the Marshfield Clinic have been at the forefront of innovation and implementation. But systems like these remain a limited component of the nation’s overall health care delivery structure. The majority of all U.S. primary care medical practices are composed of just one or two physicians, and these small practices cannot meet the core elements of the patient-centered medical home developed by the National Committee for Quality Assurance (NCQA). Another approach is needed to make care coordination widely available.

One emerging model is to link provider practices with community-based care coordination teams (community health teams, or CHTs). CHTs apply key clinical functions and processes used by larger successful physician group practices and integrated plans and replicate them in less resourced and organized settings. Teams include care coordinators; nutritionists; behavioral and mental health specialists; nurses and nurse practitioners; and social, public health, and community health workers. These trained resources already exist in many communities, working for home health agencies, hospitals, health plans, and community-based health organizations. To better leverage their systemic impact, CHTs are needed to work seamlessly with small provider practices. In combination, CHTs and provider practices could meet the requirements of a medical home as defined by the NCQA.

In addition to large systems like Intermountain, Group Health, and Geisinger, states including Colorado, New Mexico, North Carolina, Oregon, Pennsylvania, Rhode Island, and Vermont have incorporated community-based prevention and care management into FFS programs. Evaluations show improvements in clinical services, reductions in emergency department (ED) use and preventable hospitalizations, lower rates of readmission, and cost savings. North Carolina’s Medicaid program saved an estimated $535 million in two coordinated care programs. Intermountain reduced hospital admissions for patients with complex illnesses by nearly 8.7% and mortality by 3.4% just two years into its care coordination program; reported savings per patient range from $640 to $1,650 per year. At Group Health, a 29% reduction in urgent care and ED use in the first year was sufficient to offset initial program investment. Geisinger’s ProvenHealth Navigator program reduced total medical costs by 7% for its first 11,000 members and posted an estimated 2:1 return on investment.

Best-practice lessons from the early adopters show that the fundamental components of CHT care coordination include:

  • Targeting the right patients and ensuring close integration between CHTs, patients, and the coordinating care provider.
  • Interacting frequently with patients (and families) to provide education, reminders, coaching, and self-management support, along with real-time evaluation and information on clinical markers with feedback. Medication and testing adherence are essential foci.
  • Providing transitional and palliative care.
  • Linking with and referring to effective community-based interventions.
  • Using health information technology in both the CHTs and the provider practice to monitor and track patients’ progress and compliance with the care plan.

A critical focus and function of the CHTs must be transitional care. MedPAC has estimated that 18% of all hospital stays result in a readmission within 30 days, costing $15 billion annually, approximately $12 billion of which is spent on potentially avoidable readmissions. Evidence shows that effective transitional care produces better outcomes and cost savings. A series of published studies with elders in Colorado demonstrated that patients in the transitions program had a 44% lower readmission rate at 30 days than did control patients, and a 35% lower rate at 90 days. In another series, over nearly 25 years, senior researchers at the University of Pennsylvania have conducted large, randomized controlled trials with high-risk elders. Their studies have demonstrated that comprehensive models focused on transitional care produce better health outcomes and significant cost savings. The most recent University of Pennsylvania research showed a 56% reduction in readmissions and 65% fewer hospital days for patients in transitional care. At the 12-month mark, average costs were $4,845 lower for these patients.

CHTs are also a vital link to refer patients to community-based prevention programs that can deliver effective primary prevention to avert disease as well as programs to detect and mitigate existing conditions and avert complications (secondary and tertiary prevention). Randomized trials of the diabetes prevention program (DPP) protocol demonstrated that an intensive diet, exercise, and lifestyle protocol reduced diabetes by 58% relative to the control group and by 70% among adults older than age 60. But clinical interventions are expensive: The cost per enrollee over an average follow-up period of three years was nearly $54,000. Taking the lessons from these trials, group-based DPP protocols have been administered in community settings and have produced similar outcomes, reducing disease incidence at a fraction of the cost of the clinical intervention.

Special Consideration: Health Information Technology

The federal benchmark for electronic health record (EHR) adoption in small medical practices (those with five or fewer physicians) is just 12% by 2012. Cost is a significant factor in slow uptake: Acquisition of an EHR system in a small physician practice is estimated to run at least $40,000, including time and cost required for training and transition. The stimulus bill provides $2 billion to the Health and Human Services (HHS) Office of the National Coordinator for Health Information Technology to develop standards for system interoperability. It specifies that $300 million is to support regional efforts toward health information technology through technical assistance centers as well as matching grants for health IT planning and implementation and competitive grants to support loans to health care providers for EHR purchase and staff training. The administration announced the first health IT grants August 20, 2009.

The stimulus package uses both carrots and sticks in the Medicare and Medicaid programs to spur health IT use, offering payment incentives beginning in 2011 and financial penalties starting in 2016 for non-adopters. Congress authorized a 75% increase in Medicare Part B fees to adopt and use certified electronic health records (C-EHRs); those who have not adopted C-EHRs by 2016 will face a 1% reduction in fees, escalating to 3% in 2018. Medicaid providers are authorized 85% increases for C-EHR use.

Implementation: Funding and Financial Incentives

CHTs are explicitly included in the Senate Finance Committee Chairman’s Mark. The stimulus bill endows a national “Prevention and Wellness Fund” with $1 billion, including $650 million for “evidence-based clinical and community strategies that deliver specific, measureable health outcomes that address chronic disease.” The House Tri-Committee and Senate HELP draft legislation both emphasize prevention, care, and treatment for chronic illness, including readmission reduction efforts, medical home pilots, medication management programs, additional grants for community-based wellness programs, and workplace health promotion program evaluation and best-practices dissemination.

These policies are necessary but insufficient to drive the change we need. Because reimbursement for crucial elements of effective chronic disease management — education, patient counseling, care coordination, and patient monitoring — is limited in FFS Medicare, payment reforms will assume a powerful role in incenting the adoption of CHTs and the development of accountable health teams that also include hospitals and specialists.

A federal contribution of about $3 billion annually would be required to make CHTs available to all beneficiaries enrolled in traditional FFS Medicare. (Funding for needed health IT would come from funds appropriated under the American Recovery and Reinvestment Act: $19 billion provided over a five-year period). Based on published evidence, we expect that reductions in hospitalizations, clinic use, and ED use from the CHT approach use will produce savings substantially higher than this modest investment. In order to expedite funding, ensure integration of public health and health care systems, and best support existing prevention services, CHT members would become salaried (not-for-profit) employees hired through the states. At least three potential payment reforms would provide strong incentives to move toward these integrated approaches and reduce some of the well-publicized problems with Medicare’s current FFS payment system.

  1. Primary care reimbursements. Physician practices that establish a formal relationship with a CHT would receive a supplemental amount per person per month for each enrolled Medicare and Medicaid-Medicare dually eligible patient they manage. North Carolina and Vermont currently reimburse an additional amount of about $1.50 to $2.50 per patient per month, and these payment levels have generated widespread participation by physician practices in both states.
  2. Changing payments for high-readmission-rate hospitals. Coordinated, accountable care can also be incented through reduced payments for hospitals with high readmission rates, starting with the seven high-readmission MS-DRGs identified by the Medicare Payment Advisory Commission (MedPAC) and, within a 3-5 year period, extending to all Medicare readmissions. (The seven conditions are heart failure, chronic obstructive pulmonary disease, pneumonia, acute myocardial infarction, coronary artery bypass graft, percutaneous transluminal coronary angioplasty, and a general category of “other vascular” conditions.) Hospitals with above-average readmission rates would receive reduced payments for patients readmitted within a 15- or 30-day period. In parallel, bundled payments should be made to hospitals to cover inpatient and all post-acute care within a 60-day window.
  3. Bonus pools. Primary care practices and CHT staff should be eligible for additional payments if key performance measures are met. In addition to preventable readmissions, other quality measures should include improvement in clinically recommended services for common and costly chronic illnesses. To be eligible for bonus payments, health teams would have to meet a three-part test. First, Medicare per capita spending in the hospital service/referral area (as defined by the states in establishing the CHTs) would have to be lower than an established benchmark amount (lower than the average annual per capita growth for the prior 2-3 years). Second, readmissions for the seven MedPAC tracer hospital conditions would have to decline. Third, quality measures (starting with the HEDIS measures for managing and treating diabetes, hypertension, and other targeted conditions) would have to improve.

Conclusions

A third of the growth in health care spending since 1990 is associated with the doubling of obesity and rise in associated chronic conditions. Reversing or at least slowing the rise in incidence and prevalence of chronic disease is critical to better health and reduced health spending over the long term. Better managing existing disease is as well. Episodic, uncoordinated care is ineffective and inefficient for patients like most Medicare beneficiaries who have multiple, chronic comorbidities.

A body of research-based evidence shows that practice changes, such as transitional care, patient coaching, and community-based lifestyle change interventions like the adapted DPP protocol, produce better health outcomes and can reduce health spending. A mounting body of practice-based evidence from large integrated delivery systems that incorporate these functions also points to lower costs and better outcomes. The challenge is scaling and replicating these functions in a fragmented delivery system in which smaller provider practices are typical. We believe a virtually integrated model linking CHTs and smaller physician practices is a promising approach.

Email This Post Email This Post Print This Post Print This Post

 to the #1 source of health policy research.

4 Trackbacks for “Creating the Virtual Integrated Delivery System”

  1. Matthew Yglesias » The Next Health Reform Debate
    November 17th, 2009 at 12:16 pm
  2. Health Reform Debate 2.0 | The Incidental Economist
    November 17th, 2009 at 8:54 am
  3. Health Care Costs: A Market-Theoretic View | The Incidental Economist
    November 5th, 2009 at 1:34 pm
  4. Emory University | Let's Talk Health Reform » Blog Archive » Health Reform…and Community Health Teams
    October 14th, 2009 at 6:26 pm

2 Responses to “Creating the Virtual Integrated Delivery System”

  1. acavale Says:

    It is good see recognition of the fact that at least 50% of clinical care takes place in practices of 3 or less physicians. So, far such practices have been almost completely ignored by CMS and other payers. The Geisingers, Group Healths, etc. of the world provide care to relatively small pockets of the population.

    For over 4 years, my practice consisting of a solo Endocrinologist along with well-trained CRNP/CDE and Medical Assistants, have been providing care to about 5000 diabetics in a three-county area of PA. We have developed a “continuous care model” for successful management of diabetes. We have effectively used IT to achieve superb results (much better than larger practices). Of course we don’t have the resources to conduct a formal study to prove our effectiveness. But the word from our patients and that of local ER docs that lament that they don’t see hardly any of our patients in their ERs, provide us with fairly strong anecdotal evidence of the effectiveness of our strategy. However, it is been impossible to convince payers to simply take a look at our success, let alone think of getting paid for all this.

    Having personally helped the ACP craft their PC-MH program, I am quite disappointed that there is no provision for sub-specialists with the knowledge and resource to be a medical home to participate. This rigidity of thought is truly sad because it impedes innovation in medical care. Further, unless we move away from a inflexible CPT-based reimbursement system, there will be no room for rewarding better care.

  2. Christopher Langston Says:

    This concept seems to already be gaining traction in the Medicare context.

    Two weeks ago Secretary Sebelius announced “Medicare to Join State-Based Healthcare Delivery System Reform Initiatives” http://www.hhs.gov/news/press/2009pres/09/20090916a.html in a joint press conference with Vermont Govenor Jim Douglas and White House Reform leader Nancy-Ann DeParle. Although there is a lot to be worked out in the concept . . .

    The transcript from the event makes clear that community health teams are part of the discussion. http://www.vermontbiz.com/news/september/douglas-white-house-over-medical-home-plan-modeled-vermont-blueprint

    The Secretary commented,
    “Patients just don’t [get] care from their doctors in this plan. They also get care from community health teams, staffed by nurses, social workers, and behavioral health counselors who check up on patients to make sure they’re managing their chronic conditions.”

    However, our experience at the John A. Hartford Foundation suggests two issues that need to be addressed for this to work.

    1. All of these health care providers need additional expertise in geriatric care. Caring for 85 year olds with multiple conditions is not the same as caring for a 55 year old with diabetes. Most health professionals are not familiar with many basic issues in geriatric care and are therefore not prepared for their roles.

    2. Forming teams is very difficult. It takes time, structure, and specific skills. Again, health care training provides very little preparation for this way of working.

    Great ideas take strong implementation.

Leave a Reply

Comment moderation is in use. Please do not submit your comment twice -- it will appear shortly.

Authors: Click here to submit a post.