Revisiting Health Care’s Value Equation

Value is the buzzword in health care. Earlier this year, the US Department of Health and Human Services announced its goals of tying 30% of fee-for-service Medicare payments to quality or value through alternative payment models by the end of 2016, and tying 50% of payments to these models by the end of 2018.¹ Several of the nation’s largest health care systems and payers, joined by purchaser and patient stakeholders, followed suit by committing to place 75% of their business into value-based arrangements by 2020. Most recently, the Medicare and Children’s Health Insurance Program (CHIP) Reauthorization Act repealed the Sustainable Growth Rate formula and replaced it with a new system that eventually will tie payments to participation in value-based payment models.

But how should value be defined? Standard definitions describe it as the amount of benefit that can be received for something. Harvard Business School’s Michael Porter perhaps most eloquently defined value in health care as “health outcomes achieved per dollar spent.”²Porter’s definition can be viewed from both an individual patient’s perspective and a population health perspective. Importantly, his definition goes well beyond process measures of clinical quality and includes outcomes of care and total costs of care.

As the health care environment moves incrementally away from volume-based to value-based reimbursement, it is important to continually reflect on the meaning of the value equation and how best to maximize what we are receiving from health care based on what we are investing in it. Two approaches, one geared toward the numerator and the second toward the denominator, could help accelerate our progress toward a greater quotient.

The Numerator: Increasing Health Care’s Responsibility for Health

Clinical process measures of quality continue to account for a majority of the surrogates for the numerator in the value equation. For example, the Physician Quality Reporting System, the Electronic Health Record Incentive Program, and the Value-Based Modifier Program all look disproportionately at process measures. Measure sets for Medicare’s accountable care organizations (ACOs) and Medicare Advantage Star Ratings also focus largely on process measures, though they both include measures of the patient experience.^3,4 This reliance on process measures should not be surprising, as advances in modern medicine have led to a plethora of evidence-based clinical practice guidelines.

The Affordable Care Act (ACA) has helped push outcome measures of quality, such as hospital readmissions and hospital-acquired conditions, to the forefront. Preliminary estimates by the US Department of Health and Human Services show significant reductions in both of these outcomes over the last several years, due to an array of public- and private-sector initiatives.⁵

These measures notwithstanding, it is the presence of disease and unhealthy behaviors that profoundly affects health and functional status, and addressing these will ultimately lead to the improved health of populations, one leg of the Triple Aim.⁶ Behavioral and lifestyle patterns represent the largest share (40%) of the determinants of health.⁷ Chronic conditions, for which obesity (including poor nutrition and lack of physical activity) and tobacco use are the overwhelming risk factors, are the cause of 7 out of 10 deaths in the United States and the cause of the vast majority of morbidity and health care utilization. In addition, much of the growth in spending on Medicare beneficiaries over the last several decades can be attributed to rising spending on chronic conditions.⁸

Consequently, increasing health care’s accountability for reducing chronic diseases and their associated risk factors could significantly improve health outcomes. While it is true that agreement on the specific population-based activities that fall within health care providers’ scope of practice has been lacking,⁹ addressing lifestyle factors associated with chronic disease is more likely acceptable and realistic to providers than is addressing more upstream and trenchant determinants of health such as social circumstances (eg, education, employment, poverty). Note that several measures of health status are a part of the Institute of Medicine’s Core Metrics for Health and Health Care Progress, an initiative to streamline a set of measures that could provide consistent benchmarks for health progress across the nation and improve system performance in the highest-priority areas.¹⁰

Expanding the value equation’s numerator to include health status measures presents two immediate challenges. The first pertains to measurement. Claims data are often limited in how well they can capture risk factors such as tobacco use and obesity. In addition, the presence of chronic conditions often are underreported if they are not the subject of immediate evaluation and management. More than likely, assessing the ability of health care entities to reduce chronic diseases and their associated risk factors will require a mixture of data sources, including health risk assessments, electronic health records, and administrative claims data.

The second challenge may come from the health care industry, which might balk at being held accountable for measures that are not traditionally under its purview or areas in which it may not have immediate expertise. There is a reason, however, that we do not call the health care sector a sick care sector: we expect health care entities to be in the business of keeping patients healthy, not just treating sickness. Furthermore, from a budgetary perspective, most of the $3.1 trillion in national health expenditures in 2014 was spent on health care, not public health or research.¹¹ At some point, society must consider holding accountable for health the entities in our health sector receiving the bulk of dollars.

Opportunities exist to incorporate objective measures of health status into the value equation’s numerator. Medicare and commercial ACOs should begin to include these measures as part of their quality metrics. Recently, the Bipartisan Policy Center issued a report entitled “A Prevention Prescription for Improving Health and Health Care in America,” which recommended that CMS integrate population health care quality measures such as the prevalence of risk factors and the incidence of disease into the next iteration of ACOs¹² to drive system change that supports health. Similar strategies should be considered for Medicare Advantage’s Star Rating System, for the Core Set of Adult Health Care Quality Measures for Medicaid, and for commercial payers.

Specifically, one could imagine focusing on a parsimonious set of measures for health care entities with respect to their service population. For example, these entities could start with 5 measures, which might include the prevalence of tobacco use and obesity measured at appropriate intervals and the incidence of diabetes, ischemic heart disease, and depression (chronic conditions with substantial morbidity) over appropriate periods of time. To prevent adverse selection, risk-adjustment strategies could be used in addition to assessing the health care entities’ relative improvements over time.

As a second example, under the ACA, charitable hospitals are conducting, once every 3 years, community health needs assessments, and adopting implementation strategies addressing these needs. Because the IRS’s requirements for charitable hospitals are, to some extent, broad, perhaps even lax, it is important to make sure that hospitals see this requirement as more than just checking a box.¹³ If they are serious about this, they will have an enormous opportunity to partner with other hospitals in their same geographic region, with public health departments, and with community-based health organizations specifically to prevent chronic diseases and reduce their associated risk factors, which currently affect every community in the country. The Centers for Disease Control and Prevention (CDC) Community Health Improvement Navigator offers a comprehensive database of interventions for health care entities to use for many chronic diseases and their associated risk factors.¹⁴

The Denominator: Focusing on the Highest-Cost Patients

The Institute of Medicine has estimated that roughly $750 billion of annual health spending is unnecessary,¹⁵ and efforts are under way nationwide to address many of the causes of this waste, or non-value-added care. These include efforts to improve care delivery and care coordination, reduce fraud and abuse, and reduce overtreatment, to name only a few. Reforms to improve the quality of hospital care and more aggressive health care fraud recoveries are now saving several billions of taxpayers’ dollars annually, according to the US Department of Health and Human Services.¹⁶

Another approach to cutting costs is to find where a disproportionate amount of health care dollars is flowing. In 2012, the top 1% of the population, ranked by their health care expenses, accounted for 22.7% of total health care expenditures, with an annual mean expenditure of $97,956, and the top 5% accounted for 50.0% of total expenditures, with an annual mean expenditure of $43,058.¹⁷ The top 1% of Medicaid beneficiaries has been estimated to account for 25% of Medicaid expenditures,¹⁸ and the top 1% of Medicare fee-for-service beneficiaries accounts for 14% of the costs of Parts A and B.¹⁹

Given this concentration of health spending, an array of initiatives are being used to reduce preventable health care costs in this population. Most of these efforts have focused on persistent high utilizers, also known as superutilizers, as opposed to high-cost patients in their last months of life or to those who experience acute, episodic bouts of illness necessitating high-cost care. For example, on the Medicaid side, CMS’s Innovation Accelerator Program recently announced plans to provide technical assistance to up to 10 states trying to improve care coordination for Medicaid beneficiaries with complex needs and high costs. The goals of this initiative are to enhance the states’ capacity to use data analytics, refine payment reforms, and facilitate the replication of programs demonstrating promising results for this population.²⁰ In addition to CMS, the National Governor’s Association, the Robert Wood Johnson Foundation, and the Center for Health Care Strategies are all engaged with states in reducing the costs of Medicaid superutilizers.

With respect to Medicare, of the first 107 Health Care Innovation Awards funded by the Center for Medicare and Medicaid Innovation, 23 addressed the needs of high-risk patients who are high users of medical and long-term care services and supports.²¹ Undoubtedly, every Medicare ACO and Medicare Advantage plan is currently focused on risk stratification and identification and intensive care management of high-cost patients. In addition, one could make the case that Medicare’s bundled payment initiative might help reduce costs for this population, given that inpatient care has been shown to be the largest component of their health care costs.²² On the commercial side, there is also continued awareness of the importance of high-cost patients. The Health Care Transformation Task Force has created the High Cost Patient Work Group, which recently issued its first white paper on proactively identifying the high-cost population.²³

Despite the numerous initiatives under way, one should consider whether the response to the challenge of high-cost patients has been commensurate with the need or urgency. While there is universal awareness of the challenge, there is limited evidence to suggest that public- and private-sector efforts are coordinated, aligned, transparent, or necessarily strategic. Even though broad-based payment and delivery system reform is important to reduce the fragmentation of our health care system, a complementary targeted focus on the highest-cost patients (not just groups of high-cost patients, such as dual beneficiaries) is warranted. Whether we consider the top 5% of the population, accounting for roughly $1.5 trillion of health care costs, or the top 1%, accounting for roughly $600 billion, one could argue that concentrating on these patients should be the top priority of health care policymakers. In many ways, this is health care’s greatest challenge; reducing preventable costs of high-cost patients is the most focused way of moderating health care costs in this country. The United States has successfully developed and implemented national plans on a variety of health topics; accordingly, we should have a national plan, with input from the public and private sectors on how to systematically reduce the preventable costs of the highest-cost patients.

Doing this has several challenges. The first is that high-cost patients may not necessarily be the same, year after year, often making cost savings elusive and targeting these patients difficult. Researchers at the Agency for Healthcare Research & Quality have shown that only 45% of those in the top 10% of spending in the population and 20% of those in the top 1% remained in that group in the next year.²⁴ The majority (nearly 60%) of Medicaid beneficiaries, however, who were among the top 10% in one year were still in the top 10% in the next two years.²⁵ In addition, evidence from Medicare‘s Care Management for High Cost Beneficiaries Demonstration at Massachusetts General Hospital showed that despite regression to the mean, managing the care of high-cost patients achieved substantial, statistically significant savings.²⁶ Nevertheless, a national plan would need to take into account that a portion of the highest-cost patients will be constantly changing and that the cohort needs continuous evaluation in order to identify cost-savings opportunities.

A second challenge is acknowledging the difficulty of the task. Reducing preventable costs of high-cost patients is predicated on public and private payers’ timely release of health care utilization data to providers. Continuous risk-stratification to identify the high-cost population amenable to reductions in preventable costs is critical. Finally, Medicare’s previous high-risk care management demonstrations have shown that matching the acuity of a patient to the intensity of a care management intervention is essential to ensure that cost savings outweigh intervention costs. As a recent longitudinal study suggested, to maximize cost savings, much more research is needed to identify subgroups of high-cost patients and to answer questions about program design and effectiveness.²⁷ One promising recent development to support these tasks is a new, 3-stage grant initiative launched by the Peterson Center on Healthcare to identify and validate high-performance health care solutions for high-need patients. This 18-month research project is an innovative collaboration among the Harvard T.H. Chan School of Public Health, the Institute of Medicine, and the Bipartisan Policy Center.

A third challenge relates to the proprietary nature of the predictive models and tools that the health care industry uses to identify and stratify high-cost individuals for care management. In pursuit of population health, health care entities are developing sophisticated methods to predict health care utilization but typically are not revealing them publicly. Although this is to be expected in competitive markets, without sharing best practices and peer-reviewed publications of successful methods, the pace of learning will continue to be slow. The challenge of high-cost patients will require collaboration on a broad scale if we are to realize the enormous cost savings potential. In the same way that Health Datapalooza is a national conference focused on liberating health data and bringing together health care stakeholders with the most innovative and effective uses of health data to improve patient outcomes, similar high-profile events should focus on liberating predictive modeling and risk-stratification approaches to high-cost patients, with prizes and awards for the most promising efforts.

Additional opportunities include new delivery and payment models targeted to high-cost patients. CMS’s Independence at Home Demonstration is an example of a model in which the acuity of patients was matched correctly with the intensity of the care management intervention, with first-year results showing significant cost savings in addition to improvements in quality of care.²⁸Models should be tested specifically for the top 1% or 5% of high-cost patients. In 2013, a bipartisan group of US senators led by Senator Michael Bennet (R-CO) introduced an amendment to require the secretary of health and human services to conduct a pilot program in 3 regions to improve care and lower costs for the top 5% of high-cost Medicare fee-for-service beneficiaries. Health insurance plans would receive a risk-adjusted capitation payment from CMS for all Medicare-covered services, at 95% of the project baseline costs for the enrolled population.²⁹ One idea would be to use this proposal to broaden eligibility for ACOs or other providers willing to engage in risk-sharing arrangements, in addition to traditional health insurance plans.

Another idea, proposed by The Commonwealth Fund Commission on a High Performance Health System, is to identify 50 to 100 geographic areas or health improvement communities across the country with a substantial concentration of high-cost patients with multiple chronic conditions. These areas would receive financial and technical support and regulatory accommodations by CMS and private-sector partners to engage in community-based accountable care arrangements. The overall goals would be to curb excessive health care spending and improve the quality of care.³⁰

Finally, as risk-stratification methods improve and are validated and as payers transmit more timely patient cost and utilization data to providers, individual frontline providers should be expected to regularly perform basic risk stratification of their panel of patients in order to match the most appropriate care management intervention to each patient based on the level of evidence at that time. Even though the National Committee on Quality Assurance’s Patient-Centered Medical Home standards include identifying high-cost patients for care management,³¹ this element could be weighted much higher and made mandatory for all practices seeking certification.

The Role of the Community

In summary, the combination of greater attention to reducing chronic diseases and associated risk factors and to reducing preventable costs of the highest-cost patients offers the best chance for our health care system to substantially improve value for the American public. One common element in both these approaches is the importance of community support. Health care entities will not be able to prevent chronic disease solely within the 4 walls of the clinical setting. Similarly, the costs for many persistently high-cost individuals are most likely to decrease through health care partnerships with entities such as community nonprofit organizations, social service providers, public health departments, and the Aging Network. Therein lies the paradox for increasing value in health care; though there will always be opportunities to continue improving value inside the clinical setting, many of the greatest opportunities for health care lie outside it. The sooner we embrace this reality, the better our chance for increasing value from our health care system to the public.

References

1. Burwell SM. Setting value-based payment goals—HHS efforts to improve U.S. health care. N Engl J Med. 2015;372:897-899.
2. Porter, ME. What is value in health care? N Engl J Med. 2010;363:2477-2481.
3. Centers for Medicare and Medicaid Services. Shared Savings Program quality measures and performance standards. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Quality_Measures_Standards.html. Accessed September 1, 2015.
4. Centers for Medicare and Medicaid Services. Prescription drug coverage—general information—Parts C & D performance data. https://www.cms.gov/medicare/prescription-drug-coverage/prescriptiondrugcovgenin/performancedata.html. Accessed September 8, 2015.
5. US Department of Health and Human Services. Partnership for patients: methods to estimate the baseline 2010 PFP national hospital-acquired condition rate. Washington, DC: US Department of Health and Human Services; 2014. http://www.ahrq.gov/professionals/quality-patient-safety/pfp/index.html. Accessed September 15, 2015.
6. Whittington JW, Nolan K, Lewis N, Torres T. Pursuing the Triple Aim: the first 7 years. Milbank Q. 2015;93(2):263-300. doi:10.1111/1468-0009.12122.
7. McGinnis JM, Russo-Williams P, Knickman JR. The case for more active policy attention to health promotion. Health Aff. 2002;21(2);78-93.
8. Thorpe KE, Ogden LL, Galactionova K. Chronic conditions account for rise in Medicare spending from 1987 to 2006. Health Aff. 2010;29(4);718-724.
9. Kassler WJ, Tomoyasu N, Conway PH. Beyond a traditional payer—CMS’s role in improving population health. N Engl J Med. 2015;372:109-111.
10. Institute of Medicine. Vital Signs—Core Metrics for Health and Health Care Progress. Washington, DC: National Academies Press; 2015. https://iom.nationalacademies.org/~/media/Files/Report%20Files/2015/Vital_Signs/VitalSigns_RB.pdf. Accessed September 8, 2015.
11. Keehan SP, Cuckler GA, Sisko AM, et al. National health expenditure projections, 2014–24: spending growth faster than recent trends. Health Aff. 2015;34(8):1407-1417.
12. Bipartisan Policy Center. A Prevention Prescription for Improving Health and Health Care in America. Washington, DC: Bipartisan Policy Center; 2015. http://bipartisanpolicy.org/library/a-prevention-prescription-for-improving-health-and-health-care-in-america/. Accessed September 1, 2015.
13. Fox DM. Policy commercializing nonprofits in health: the history of a paradox from the 19th century to the ACA. Milbank Q. 2015;93(1):179-210. doi:10.1111/1468-0009.12109.
14. Centers for Disease Control and Prevention. CDC Community Health Improvement Navigator. Atlanta, GA: Centers for Disease Control and Prevention; 2015. http://www.cdc.gov/chinav/database/index.html. Accessed September 8, 2015.
15. Institute of Medicine. Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. Washington, DC: National Academies Press; 2013. http://iom.nationalacademies.org/Reports/2012/Best-Care-at-Lower-Cost-The-Path-to-Continuously-Learning-Health-Care-in-America.aspx. Accessed October 12, 2015.
16. US Department of Health and Human Services. The health care fraud and abuse control program protects consumers and taxpayers by combating health care fraud. Washington, DC: US Department of Health and Human Services; 2014. http://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2015-Fact-sheets-items/2015-03-19.html. Accessed September 15, 2015.
17. Cohen SB. Differentials in the concentration of health expenditures across population subgroups in the U.S., 2012. Statistical Brief No. 448. Rockville, MD: Agency for Healthcare Research and Quality; 2014.
18. Center for Medicaid and CHIP Services Informational Bulletin. Targeting Medicaid Super-Utilizers to Decrease Costs and Improve Quality. Baltimore, MD: Centers for Medicare and Medicaid Services; 2013. http://www.medicaid.gov/federal-policy-guidance/downloads/cib-07-24-2013.pdf. Accessed October 15, 2015.
19. Medicare Payment Assessment Commission. A Data Book: Health Care Spending and the Medicare Program. Washington, DC: Medicare Payment Assessment Commission; 2011. http://www.medpac.gov/document_TOC.cfm?id=617. Accessed February 22, 2012.
20. Centers for Medicare and Medicaid Services. Improving care for Medicaid beneficiaries with complex needs and high costs. http://www.medicaid.gov/state-resource-center/innovation-accelerator-program/beneficiaries-with-complex-needs/beneficiaries-with-complex-needs.html. Accessed September 8, 2015.
21. Miller WD. HCIA Complex / High-Risk Patient Targeting: First Annual Report. Bethesda, MD: NORC at the University of Chicago; 2015. http://innovation.cms.gov/Files/reports/HCIA-CHSPT-FirstEvalRpt.pdf. Accessed September 1, 2015.
22. Joynt KE, Gawande AA, Orav EJ, Jha AK. Contribution of preventable acute care spending to total spending for high-cost Medicare patients. JAMA. 2013;309(24):2572-2578.
23. Health Care Transformation Task Force. Proactively identifying the high cost population—insights from the Health Care Transformation Task Force. Washington, DC: Health Care Transformation Task Force; 2015. http://www.hcttf.org/resouces-tools/2015/5/14/whitepaper. Accessed September 8, 2015.
24. Cohen SB, Yu W. The concentration and persistence in the level of health expenditures over time: estimates for the U.S. population, 2008–2009. Statistical Brief No. 354. Rockville, MD: Agency for Healthcare Research and Quality; 2012.
25. Coughlin T, Long S. Health care spending and service use among high-cost Medicaid beneficiaries, 2002–2004. Inquiry. 2009/2010;46(4):405-417.
26. McCall N, Cromwell J, Urato C. Evaluation of Medicare Care Management for High Cost Beneficiaries (CMHCB) Demonstration: Massachusetts General Hospital and Massachusetts General Physicians Organization (MGH). Research Triangle Park, NC: RTI International; 2010. http://www.massgeneral.org/News/assets/pdf/FullFTIreport.pdf. Accessed September 15, 2015.
27. Johnson TL, Rinehart DJ, Durfee J, et al. For many patients who use large amounts of health care services, the need is intense yet temporary. Health Aff. 2015;34(8):1312-1319.
28. Centers for Medicare and Medicaid Services. Independence at Home demonstration. http://innovation.cms.gov/initiatives/Independence-at-Home/. Accessed September 8, 2015.
29. Bennet-Toomey Fourth Amendment to the SGR Repeal and Medicare Beneficiary Access Improvement Act of 2013. US Senate. http://www.finance.senate.gov/imo/media/doc/SGR.Health.Extenders.Final%20Amendment%20List.pdf. Accessed September 1, 2015.
30. Blumenthal DB. Performance improvement in health care—seizing the moment. N Engl J Med. 2012;366(21):1953-1955.
31. National Committee for Quality Assurance. PCMH 2011–PCMH 2014 crosswalk. http://www.ncqa.org/Programs/Recognition/Practices/PatientCenteredMedicalHomePCMH/PCMH2011PCMH2014Crosswalk.aspx. Accessed September 8, 2015.

Acknowledgment: The author thanks Gerard Anderson of Johns Hopkins University for his helpful comments on an earlier draft of the manuscript.

About the Author