Increasing Stress on the U.S. Health Care System: Structural Crisis or Temporary?

Congressional Testimony
March 10, 2003
Len M. Nichols

Mr. Chairman, Senator Breaux and members of the Committee, my name is Len M. Nichols, and I am the vice president of the Center for Studying Health System Change (HSC). HSC is an independent nonpartisan policy research organization funded exclusively by The Robert Wood Johnson Foundation and affiliated with Mathematica Policy Research. We conduct recurrent nationally representative surveys of households and physicians, site visits to monitor ongoing changes in the local health systems of 12 U.S. communities, and monitor secondary data and general health system trends. Our goal is to provide members of Congress and other policy makers with unique insights on developments in health care markets and their impacts on people. Our various research and communication activities may be found at www.hschange.org.

The U.S. health care system is undergoing a period of growing stress. Rising health care costs and health insurance premiums, combined with the recent slowdown in economic activity, have forced many employers and workers to make tough choices about who will bear the brunt of cost increases and who will risk going without health insurance. The uninsured population did not shrink appreciably with the strong economy in the 1990s and has risen recently. Most coverage gains that did occur were due to the State Children’s Health Insurance Program (SCHIP)¹, but now with states in unprecedented fiscal distress, they are looking to reduce spending on public insurance programs, and some may actually reduce the number of people covered. Partly in response to long-term budgetary and demographic trends, the Bush Administration has proposed far-reaching Medicaid and Medicare reforms and benefit changes that could significantly alter not only the programs themselves but the nature of financing them as well.

So it is understandable that many are asking if our health care system is in crisis. The word "crisis" may be too strong, but the answer to the question may also depend on where you sit and what your goals are for the system as a whole. The system looks like it is in a crisis when, for fear of out-of-pocket cost, an uninsured woman does not seek care until symptoms have persisted and worsened, and then discovers it is too late for even our incredible delivery system and dedicated professionals to save her.

The system looks like it is in crisis for a small employer and his workers when insurance premiums jump beyond their reach. The system looks like it is in crisis when there are not enough nurses to staff hospital beds for patients who need them, and there are not enough people going into nursing school to meet our future needs. The system looks like it is in crisis when physicians refuse to heal out of frustration with our malpractice insurance system. The system looks like it is in crisis when a heart and lung, incompatible with her blood type, are nevertheless transplanted into a teen-aged girl at a first-rate academic medical center. To some policy makers, the system looks like it is in crisis when tens of millions of Americans go without health insurance for a year or more. To other policy markers, the system looks like it is in a crisis when they discover we now spend more than 14 percent of gross domestic product (GDP) on health care. Some policy makers may have concluded that the cost growth crisis is the reason we cannot solve the uninsured crisis, and they apparently prefer to avoid discussing both.

These valid concerns notwithstanding, our system also does marvelously or at least passably well for most of us most of the time. Most of us have health insurance all the time, most of us have a regular doctor whom we trust², and most of us have confidence that if we need serious secondary or tertiary care, we will have access to it without undue delay or financial ruin. Thus, to a majority of Americans at any given time, our system does not look like it is in crisis in relation to their personal concerns. Opinion polling does show cycles of unease about the system as a whole, and we are apparently in a complex period, wherein most are dissatisfied with the availability and affordability of health care, but other issues were much more important to voting patterns in the recent mid-term elections.³

But rather than debate the breadth and depth of our health system "crisis," I will focus on three fundamental problems-key sources of much stress that are wearing our system’s strengths and resilience down. These problems are deeply connected to each other, and thus attempts to solve them in isolation are likely to prove disappointing. It may indeed take system-wide reform to make real progress in any dimension. But system-wide reform is expensive and complex, as we learned painfully in our recent past. Thus, I will conclude with some key lessons from the Clinton reform era that seem relevant to your current deliberations.

In my view, the three fundamental problems in our health care system today are waste, uneven quality and uneven access to care.

Waste. In 2001, we spent 14.1 percent of our GDP on health care, substantially more than any other nation (Germany and Switzerland both spend a little over 10 percent). Yet, among the nations of the world, we rank 28th in infant mortality, right below Cuba, Ireland and Portugal, and 29th in life expectancy at birth. Assuming that some of our poor performance on these measures is due to the fact that most other nations have universal health insurance coverage and access for all their citizens, it is sobering to note that we also rank 26th in life expectancy at age 60, which means our almost universal Medicare program helps our elderly catch up a little-to be tied with Cuba, Korea and Slovenia-but we still live much shorter lives on average than our wealth and our system resource expenditure should buy, if we lived our lives and bought health care as wisely as other nations do.⁴

There are two other related ways to think about wasted resources in our system. One is care that is costly but has no strong clinical justification and, therefore, could be reduced without harm to patients. Dr. John Wennberg, perhaps our nation’s leading authority on geographic medical practice variation, recently and usefully described three categories of health care.⁵ One is effective care, whose use is supported by well-articulated medical theory and strong evidence for efficacy, as determined by clinical trials or valid cohort studies. Another is preference-sensitive care, wherein at least two valid alternative treatment strategies are available. In these cases, the choice of treatment involves trade-offs between risks and benefits that patients and caregivers should explore before choosing one.

The final category of care is what Wennberg and colleagues call "supply sensitive." In contrast to effective and preference-sensitive care, supply-sensitive care has much less well-developed medical theory governing decisions, especially around certain kinds of hospitalizations and the frequency of follow-up visits to specialists. Variation in practice patterns is particularly pronounced during the last six months of life, when 20 percent of Medicare spending occurs. Frequency of visits to specialists varies by 12-fold (two per decedent beneficiary in Mason City, Iowa vs. 25 in Miami, Fla.), and the number of hospital days per decedent varies by a factor of five (4.6 days in Ogden, Utah vs. 21.4 days in Newark, N.J.).

Variation in Medicare spending per capita is highly correlated with medical specialist visits, diagnostic tests, and the use of hospitalizations and intensive care for medical (non-surgical) conditions. There is a growing body of research that strongly suggests that areas with higher rates of supply-sensitive care do not have better health outcomes.⁶ Unproductive medical practice variation, long established through Medicare program data, is likely to extend to treatment patterns in the privately insured under-65 population as well and could at least partially explain the tremendous geographic variation in premium rates observed around the nation.

One final way to think about the burden of waste in our health care system is to consider how much care we are paying for that we do not need, and how much a typical insurance premium costs as a proportion of typical family income. Jon Gabel and colleagues estimate that a typical group family policy cost $7,954 (single $3,060) in 2002.⁷ That is 15 percent of our median family income of about $52,000.⁸ In 1987, a typical group family premium was about 8 percent of median family money income. Only about 35 percent of families today make enough to prevent more than 10 percent of their gross income from going toward the purchase of health insurance for their family (either in foregone wages or out-of-pocket payments). ⁹ Thus, as health care costs have continued to grow faster than wages for decades, an increasing share of our work force is finding it harder and harder to purchase comprehensive health insurance as we have come to know it.

Uneven quality. It is true that the world’s wealthy elite come to the United States for much secondary and tertiary medical care. We certainly have many of the best physicians and treatment facilities in the world. At the same time, to quote from the Institute of Medicine’s (IOM) National Roundtable on Health Care Quality:

Serious and widespread quality problems exist throughout American medicine. [They] occur in small and large communities alike, in all parts of the country and with approximately equal frequency in managed care and fee-for-service systems of care. Very large numbers of Americans are harmed as a result. ¹⁰

Two path-breaking IOM reports followed this Roundtable. To Err is Human¹¹ reported how it is likely that medical errors-mostly from poor communication and coordination among health professionals-lead to as many as 98,000 unnecessary hospital deaths a year. Crossing the Quality Chasm identified the major system problem areas in quality measurement, attainment, and improvement, and remarkably also laid out a blueprint for how we as a nation could begin to address our stunning shortcomings. The most recent IOM report summed up our situation quite well:

. we have extraordinary knowledge and capacity to deliver the best care in the world, but we repeatedly fail to translate that knowledge and capacity into clinical practice. ¹²

Our quality gaps range from often not doing things that should be done (e.g., cancer screenings, certain blood and eye tests for diabetics, beta-blockers for post-heart attack patients, childhood immunizations), to too often doing things that should not be done if care were being managed appropriately (e.g., hospitalizations for ambulatory sensitive conditions like asthma and diabetes, hospitalizations for medical conditions generally, overuse of ICUs in the last six months of life, unnecessary follow-up visits to specialists, contradictory information from sequentially treating health professionals, and more than one simultaneous prescription, from different doctors, that might interact in harmful ways).

Spurred by the Quality Chasm and preceding research, the IOM¹³, the Agency for Healthcare Research and Quality (AHRQ)¹⁴, and the Centers for Medicare and Medicaid Services (CMS)¹⁵ have all now developed measures of quality clinical care that should (but cannot now) be tracked in a systemic manner, as they try to lead our health care system toward a higher and more uniform level of quality performance. AHRQ will release a much-anticipated report on the state of quality in our health care system later this year. Early research on quality indicators for Medicare patients, based on random samples of medical records, indicates: "Care for Medicare fee-for-service plan beneficiaries improved substantially between 1998-99 and 2000-01, but a much larger opportunity remains for further improvement."¹⁶ The National Committee for Quality Assurance (NCQA), a private non-profit organization, has developed and published HEDIS measures on important dimensions of care and service at the health plan level for the last five years, and these too have shown improvement recently, but also indicate just how far we have to go.

For example, the HEDIS measures that we have today mostly apply only to health maintenance organizations (HMOs) that cover about one quarter of insured Americans. The sad truth is at the moment very little systematic quality information is recorded for most care that is delivered in the United States today, especially in the preferred provider organization (PPO) and fee-for-service arrangements that still dominate provider-insurer contracts.

The NCQA publishes conservative estimates of avoidable deaths in the US from inappropriate treatment:

These estimates are derived by assuming everyone in America received care as good as that dispensed in the top 10 percent of reporting health plans, and they assume that the baseline is the care delivered by the median reporting plan. (Some would argue this is an optimistic baseline assumption, but it is impossible to say because we have no national data collection effort beyond the private efforts led by NCQA and specific research projects). And despite the improvements that five years of data reporting have brought to health plan performance and perhaps system-wide performance as well, the norm remains astounding variability across health plans. For example, childhood immunization rates that vary from 48 percent to 81 percent, and eye exams for diabetics are performed in only 70 percent of cases in the 90th percentile managed care plan, and a mere 35 percent of cases in the 10th percentile plan. There is great regional variation as well, with managed care plans in the south performing 20 percent to 40 percent worse on key measures like those mentioned above.¹⁸

Perhaps our basic problem here stems from this fact, noted by the NCQA: "No widely used system of reimbursement rewards high quality based on performance measurements, and some [reimbursement] systems are actually detrimental to it.¹⁹" Any economist will tell you, if we do not set incentives appropriately, we cannot expect optimal behavior, either at a personal, professional or systemic level.

Uneven access to care. First of all, 41 million people did not have health insurance for the entire year in 2001. The absence of insurance reduces access to needed care, reduces the chances of early detection and good health outcomes from serious conditions, and generally places a number of risks onto many of our most vulnerable fellow citizens. It is fairly well known that health insurance coverage rates vary substantially by income level, race and ethnicity. For example, Latinos are about three times as likely to be uninsured as whites, and have been since 1987, when detailed statistical record keeping began. African Americans, by contrast, are roughly twice as likely as whites to be uninsured.

Access to care is higher among the insured than the uninsured. Partly due to coverage disparities then, Latinos and African Americans, when compared with non-Latino whites, are less likely to have a regular health care provider, to have had a doctor visit in the last 12 months, to have seen a specialist, and more likely to visit an emergency room. These gaps have persisted for years, and would likely be reduced—but not eliminated—if coverage were expanded to all groups equally.²⁰ And these access gaps are the proverbial tip of the iceberg, since lack of timely access often leads to significantly worse health outcomes, including higher risk of mortality.²¹ Perhaps the most surprising and troubling inequity is that even if you control for income, insurance and age, minority death rates are still higher than whites by quite a margin for a large number of diseases.²² We are a long way from colorblind equality in our health care system.

Recently, these key problems of waste, uneven quality and inequities have been exacerbated by a re-acceleration of health care cost growth, made possible by the wholesale retreat from the most effective but even more unpopular tools of managed care. This cost growth is partly price recovery for providers who are asserting their renewed bargaining power vs. health plans, but the cost growth is actually mostly driven by increases in utilization.²³ Many analysts consider this double-digit rate of health care cost growth to be unsustainable, and, therefore, it is likely to abate fairly soon, leaving us with less immediate cost growth pressure but the continued problems of waste, uneven quality and inequities at even higher levels of resource claims.

Each of these three major problems with our health care system—waste, uneven quality, and inequity—is daunting enough to discourage many who seek solutions. Incremental efforts to solve one are often frustrated by spillover effects flowing back from the other two. Waste and poor quality exacerbate our cost problem, which in turn creates more uninsured, especially among low-income people, who are disproportionately racial and ethnic minorities. It is impossible to solve the quality problem without coordinated care and real-time communication among different providers, but communication infrastructures do not exist because no one has been willing to spend more than they already do to construct them.

The market return from investing in quality—both for providers and for health plans—has been low because most patients do not perceive there is a quality problem and trust their own doctor and the local hospital he recommends, regardless of what elites and the information advocates think. Uninsured and low-income patients present particular communication challenges with the overworked health professionals who see them. They are in general less compliant with treatment regimens than middle-class, privately insured patients, and since so much of quality care requires self-management and health literacy, coverage and access gaps maintain the quality chasm for vulnerable populations even more so than for the general population. Finally, profound fear of malpractice and economic loss generally has retarded provider engagement on the quality and error-reporting fronts, all of which serves to keep costs high, and so the dysfunctional set of interactions continues to stress our system.

Thus, all three problems need to be addressed simultaneously with system reforms and federal leadership on all fronts. This will require substantial new resources to be committed, even though waste reductions and general quality enhancements can be expected to offset some of these new costs in the long run.

It is still early, but not the first day, of a new national conversation about system reforms that could ultimately relieve the stresses on our health care system. Much detail remains to be worked out, understood and digested, and the range of possible forms an improved system might take is quite broad.²⁴ Senator Breaux has already laid out a vision for system reform that can serve as a cohesive springboard for ongoing discussions and proposal refinement, as have Democratic presidential candidates, Health and Human Services Secretary Thompson and the President, himself.²⁵

This might be a particularly opportune moment to look back at the experience of the Clinton reform effort in 1993-94 for overarching lessons that may inform current efforts to reduce the key and related problems of our health care system.

1. Bipartisanship. Our health care system is too large, complex and deeply personal to reform by narrow legislative margins of victory. If we cannot build a consensus from the middle of the political spectrum that can attract wide support, we are not ready to reform our system, as we apparently were not in 1993-94.

2. Federal leadership, not micromanagement. Federal leadership is indispensable, for the interests involved are just too powerful for any other actor to lead the way for systemic reform. Federal legislation must define responsibilities for all, including state governments, employers, health care providers, insurers, and most importantly individuals, but important choices regarding implementation must also be left with states and local market participants who know our heterogeneous health care markets best. Federal resource commitments must also be substantial, both in terms of creating an information infrastructure to support quality improvement and in terms of subsidies for coverage expansion, or this entire effort will prove disappointing.

3. Transparency. The Clinton Administration, elected in 1992 with 43 percent of the vote, believed that the people would not support payroll or income tax increases to pay for universal health insurance and fundamental system reform. They then devised financing arrangements-Medicare provider payment cuts and statewide global budgets-that turned out to be technically feasible but politically unsustainable. Many analysts and observers doubted that the nation could actually do something on this scale with no new taxes. The primary risk of the Clinton plan was that technological advances might have been curtailed due to the global budgets on expenditures that made the financing plans technically feasible. Those risks were not made clear, and other risks were imagined and exaggerated. The key point is, the loss of credibility on financing was partially self-inflicted and due to a lack of transparency. A successful system reform campaign will elicit a straightforward willingness to pay for a better health care system for all Americans, or not, in which case we are not ready, again.

4. Do not make enemies of key players in the health care system. Almost all feel justified in blaming someone else for our current system’s failures. There is plenty of blame to share, and blame is not the point; developing a simultaneous approach to waste, uneven quality and uneven access is the point. We cannot get out of our morass without the enthusiastic participation of most providers, health plans, employers, governments, and citizens/workers/patients. There is plenty of work for all major players to perform, and most are willing. We just need to reform information and payment systems to encourage, facilitate and reward better health and health care choices.

5. The key health system unit is not the health plan. All managed competition proposals, like the Clinton plan, put the health plan at the vortex of health system reform. In essence, since 1994, employers hired managed care plans to take care of their cost growth problem, and they did, until patient and provider backlash made the employers tell their agents-the health plans-to back off on the techniques that worked. This backlash has revealed the difficulty of depending on a deus ex machina to solve our health care cost, quality and access problems for us, magically and painlessly. Health plans cannot be the silver bullet we may wish for. But they are very likely to be part of the solution, for they remain potentially useful devices for organizing and disseminating information to providers and coordinating care for patients among networks of collegial providers. Professional specialty societies and government agencies are alternative communication devices that will be necessary as well.

But the key health care transactions will always be between caregivers and patients. Incentives must be correct for this set of transactions, or we have no hope. If we are able to get the incentives right here, much else will take care of itself.

What kind of system is most likely to get these key incentives right?

A system that pays for quality care and good health outcomes, and does not pay for failure to provide quality care, is the kind of system most likely to have sustainable incentives. Building this system will require public investment in an information infrastructure so that providers and patients will find it easier to jointly produce good quality health care and the best health outcomes possible. This kind of system will also require organized group purchasing, with serious leadership from our public insurance programs, to create an environment in which widespread provider participation in quality improvement efforts is not an option dependent on the balance of local market power but a reality.²⁶

In the long run, improvements in technology drive health care cost growth. We certainly value these technological improvements, but heretofore our financing and delivery system has been biased in favor of paying for everything, whether truly effective in promoting better health outcomes or not. By learning now about how to reduce unhelpful services, the supply sensitive sort that Wennberg and colleagues have identified, and to improve quality in the Chasm sense, we would not only create savings we can then spend more wisely, we might also create an evaluation infrastructure that can help us decide which future health technologies should be financed with pooled resources and which should be financed outside our insurance system, if at all.

This seems to me to be the wisest way to prepare for the future; learn how to decide what health care we really want to buy. Otherwise, I see two unhappy fates. We could spend 25 percent of GDP on healthcare by 2030 and still have uneven quality and mountains of wasted care. Or we could render our inefficient care even further out of the reach of ever-larger fractions of our population, and keep our GDP share down by making access to health care even more exclusive than it is today. I would much rather see our nation invest now in learning how to buy and deliver health care more effectively so that we can always afford good quality care for all.

¹Strunk, Bradley C. and James D. Reschovsky, Working Families’ Health Insurance Coverage, 1997-2001, Tracking Report No. 4, Center for Studying Health System Change, Washington, D.C. (August 2002); Cunningham, Peter J., James D. Reschovsky and Jack Hadley, SCHIP, Medicaid Expansions Lead to Shifts in Children’s Coverage, Issue Brief No. 59, Center for Studying Health System Change, Washington, D.C. (December 2002).

²Marie C. Reed and Sally Trude, Who do you Trust? American’s Perspectives on Health Care, 1997-2001,Tracking Report No. 3, Center for Studying Health System Change, Washington, D.C. (August 2002).

³Robert J. Blendon et al. "Where Was Health Care in the 2002 Election?" Health Affairs web exclusive (December 11, 2002).

⁴All international comparisons are from either Health United States, 2002, or the World Health Organization Web site, www3.who.int/en/.

⁵John Wennberg, Elliott S. Fisher and Jonathan S. Skinner. "Geography and the Debate Over Medicare Reform," Health Affairs, Web Exclusive (February 13, 2002).

⁶E.S. Fisher et al. "Associations Among Hospital Capacity, Utilization, and Mortality of U.S. Medicare Beneficiaries, Controlling for Sociodemographic Factors," Health Services Research, Vol. 34, No. 6 (2000); J.S. Skinner, E.S. Fisher and J.E. Wennberg. "The Efficiency of Medicare," National Bureau of Economic Research Working Paper No. 8395 (July 2001).

⁷Jon Gabel et al. "Job-Based Health Benefits in 2002: Some Important Trends," Health Affairs, Vol. 21, No. 5 (September/October 2002).

⁸The most recent family income data are for 2001. I inflated that median income level at last year’s growth rate, 1.3%, to estimate median family income level for 2002. Given the recession, that is probably optimistic.

⁹Author’s calculations based on 1988-2002 Kaiser Family Foundation/Health Research and Educational Trust, KPMG Peat Marwick, and Health Insurance Association of America employer surveys, as well as Bureau of the Census family income measures. The latter can be found at www.census.gov/hhes/income/histinc/f08.html.

¹⁰M.R. Chassin and R. W. Galvin, "The Urgent Need to Improve Health Care Quality. Institute of Medicine National Roundtable on Health Care Quality," Journal of the American Medical Association, Vol. 280, No. 11:1000-5. (1998).

¹¹Institute of Medicine. To Err is Human: Building a Safer Health Care System. L.T. Kohn, J. M. Corrigan, and M. S. Donaldson, eds., National Academy Press, Washington, D.C. (2000).

¹²Institute of Medicine. Priority Areas for National Action: Transforming Health Care Quality. Karen Adams and Janet M. Corrigan, eds., National Academy Press, Washington, D.C. (2003).

¹³Ibid;

¹⁴National Healthcare Quality Report: Update on Current Status. AHRQ Fact Sheet.

¹⁵S.F. Jencks et al. "Quality of Medical Care Delivered to Medicare Benficiaries," Journal of the American Medical Association, Vol. 284, No. 13 (October 4, 2000).

¹⁶S.F. Jencks et al. "Change in the Quality of Care Delivered to Medicare Beneficiaries, 1998-99 to 2000-01." Journal of the American Medical Association, Vol. 289, No. 3. (January 15, 2003).

¹⁷National Committee for Quality Assurance, The State of Health Care Quality. (2002).

¹⁸Ibid.

¹⁹Ibid p.7.

²⁰Hargraves, J. Lee, The Insurance Gap and Minority Health Care, 1997-2001, Tracking Report No. 2, Center for Studying Health System Change, Washington, D.C. (June 2002).

²¹Institute of Medicine. Care Without Coverage, Too Little, Too Late. National Academy Press, Washington, D.C.. (2002).

²²Institute of Medicine. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care, National Academy Press, Washington, D.C. (2002).

²³Strunk, Bradley C., Paul B. Ginsburg, and Jon R. Gabel, "Tracking Health Care Cost: Growth Accelerates Again in 2001," Health Affairs, Web exclusive, (September 25, 2002).

²⁴See Covering America for a representative sample of alternative but feasible models of our health delivery and financing systems.

²⁵http://breaux.senate.gov/~breaux/releases/2003123A14.html; http://www.os.dhhs.gov/budget/04budget/fy2004bib.pdf;

²⁶Christianson, Jon B. and Sally Trude, "Managing Costs, Managing Benefits: Employer Decisions in Local Health Care Markets," Health Services Research, Vol. 38, No. 1, Part II (February 2003).