Order 1257985: Ethical Issues With an Aging Population

cost and end-of-life care • summer 2011 121

The Ethical Implications of Health Spending: Death and Other Expensive Conditions Dan Crippen and Amber E. Barnato

Overview In this essay I ask the reader to consider the “end of life” as a life stage, rather than as a health state. At one end of the life course is childhood and at the other end is elderhood. The basic inter-generational social compact in most societies is that working adults take care of their children and their parents, and count on their children to do the same for them. In developed countries, these obligations are met in part through government programs, with taxpayers funding signifi- cant portions of education, health care, and income support.

The financing of these public programs, in addition to other public services, involves ethically charged trade-offs. In the United States, public outlays on behalf of children and the elderly span roughly the same number of years, but with very different levels of spending. Cross-sectionally, transfers from work- ers (via taxes) go more to the elderly in the form of Medicare and Social Security income than to children in the form of public education and means-tested health insurance (e.g., Medicaid, SCHIP). Longitu- dinally, delayed “transfers” to these same children are manifest as better or worse economic conditions once the children become workers. If current workers, in addition to providing for the young and old through taxpayer-funded social programs, manage to save as well by reducing their own consumption of goods and services, then future generations are likely to be better off since these current savings are invested in capital which will allow the economy to grow (faster). In con- trast, if individuals, institutions, or governments bor- row for consumption today, then future generations are likely to be worse off since current consumption may reduce economic growth in the future and, in the case of public borrowing, additionally obligate future taxpayers to fund the cost of expenditures we make today.

In the United States, health care spending is a criti- cal component of examining both these intra- and inter-generational transfers. At present we spend much more on health care, and in total, for the elderly,

Dan Crippen, Ph.D., is the newly appointed Executive Di- rector of the National Governors Association. He has held various posts in the public sector, including Chief Counsel to the Senate Majority Leader, Assistant to the President and Domestic Policy Advisor, and Director of the Congressional Budget Office. Over the last decade, he has worked primarily in the private sector with various organizations providing or financing health care. Amber E. Barnato, M.D., M.P.H., M.S., is an Associate Professor of Medicine and Health Policy and Management at the University of Pittsburgh, and was a Visiting Scholar at the Congressional Budget Office during Dan Crippen’s tenure as Director.

122 journal of law, medicine & ethics

S Y M P O S I U M

than we do for educating and ensuring safety-net health care for our youth, thus posing the question of whether we are spending “enough” on the young. Moreover, by incurring large budget deficits at the fed- eral level, in part to continue supporting health care for the elderly (which constitutes a large and growing part of the total budget), we are pushing off the pay- ment for some of current health care spending on the elderly to future generations.

The primary driver of increased health care costs for the elderly is not spending for those who died, but spending for all beneficiaries. Secular trends in Medicare spending reveal that spending on behalf of beneficiaries who die during the year has been a stable 27% for decades. Although those who die incur a dis- proportionate share of total spending, the drivers of cost growth affect survivors and decedents identically. Increasing health care spending on the elderly who are in the “end of life” life-stage (not just the “end of life” health state) will crowd out investments in current and future generations of children. The inter-generational impact of health care spending on those who are in the “end of life” life stage may have greater ethical impli- cations than the narrow “futility” debate surounding those who are in the “end of life” health state.

The Issue Giving, and accepting, the assistance of others is a com- mon human experience. Whether as children, when education and the essentials of life are provided by adults, or as retirees who receive financial and physi- cal aid in their later years, virtually all of us experience both the give and take of life. Historically, much of this support has been voluntarily provided by familial sup- port and bequests, or by charitable contributions and institutions, including those of the church. In more successful societies with higher income, the elderly may also have savings to consume in retirement.

One of the hallmarks of “economically developed” civilizations is a large component of public, govern- mental transfer of resources, from those who are working to those who are largely not. In most coun- tries both culture and government shape these trans- fers, subject to societal and legal compacts between and across generations.

Altogether, these public and private reallocations of resources can consist of real property (land, housing, public roads), financial resources (credit and cash), or transfers, including direct services (bequests, educa- tion, health care).

In the United States, children are provided a combi- nation of public and private resources. The lion’s share of funding for education, especially through high

school, comes from state and local taxes. The provi- sion of food, clothing, housing, and health care are mostly provided by parents, but federal programs for food stamps, school lunch programs, housing vouch- ers, Medicaid, and SCHIP contribute to lower-income families.

The elderly are typically supported with public pro- grams of cash transfers, primarily social security, and reimbursement for health care through Medicare and Medicaid. By increasing amounts, governments of all types are providing income and health care for their former employees, now retirees. Altogether, public programs for health care provide nearly half of the annual costs for all health care in the United States. The funding for these programs, including for retir- ees, comes largely from taxes on current workers, or from borrowing.

The moral or ethical implications of these realloca- tions, might be examined across both current popu- lations (living citizens of differing ages, for example), as well as across generations over longer periods of time. As there are always a finite number of resources at a given point in time, current spending reflects the choices we make among those in the current popula- tion who cannot fend for themselves — for example, the more we spend on our parents after they retire, the less we have for our children.

The basic compact in the United States between the current working population and their dependents has been that workers pay or provide for their children and help support their parents. As workers retire, they depend on their working children to pay taxes to sup- port them through public programs, whatever private transfers they receive, and a ready market to sell the non-cash assets they have accumulated. As support moves from private and familial to public and govern- mental, the association between workers and depen- dents becomes less direct since the reallocations occur within the larger society.

Figure 1 illustrates the typical pattern in the United States of consumption of resources, either directly or as provided by others, and the creation of resources as reflected by labor income. The total available for consumption is determined by the total production of goods and services of those working, i.e., the size of the economy. Across the current population the amount consumed is approximately a zero-sum game: the more consumption for any one group — children, workers, retirees — means less for the others.

The basic compact can be altered in many ways, the most important of which affects future generations. In basic terms, if the current working population not only pays for their children and supports their parents, but

cost and end-of-life care • summer 2011 123

Crippen and Barnato

also saves a portion of their income, then economic growth can be accelerated and future economies will be larger, thus making it easier to assist future depen- dent populations. Savings add to the capital available, making it easier (and cheaper) to provide investment that increases national output. It is important to note that it is national savings that counts — the total of individuals, business, government, non-profits — and borrowing by any sector reduces the positive effect of savings by others.

Similarly, if borrowing is used to fund today’s consump- tion of goods and services, future spending and saving will be reduced. In the case of gov- ernment borrowing, especially by the federal government, it is possible to push the costs of debt well into the future.

For example, current inter- est costs on the federal debt are nearly $200 billion. With the expected deficits over the next decade, interest is pro- jected to grow to $800 billion — nearly as much as the total for Medicare in 2020. That means federal spending will need to be cut, taxes increased, or additional debt incurred to cover just the interest on our debt.

As we continue to increase debt, future generations will

have less to spend and invest. To the extent the spend- ing and borrowing of government is needed/used to fund health care, the country is pushing the cost of today’s health care onto future citizens. Under cur- rent policies, the present cohort of children will be expected to pay not only for their own children and parents, but also for past generations as well, breaking the basic intergenerational compact.

Figure 1 Life-Cycle Income and Consumption in the United States

Figure 2 Per Capita Spending on Children and the Elderly in the United States

Source: Issacs, 2009

124 journal of law, medicine & ethics

S Y M P O S I U M

Health Spending across Populations A substantial part of cross-generational transfers, especially in developed countries, are attributable to spending for education and health care. Figure 2 shows estimates of resources used in the United States in a typi- cal year by or for a “child” (up to age 18) and “elder” (over 65).

Although there is a large disparity between the resources expended on the two groups, there is nothing inherently wrong, unethical, or amoral about the greater spending for our elders. Rather, it is a reflection of our collec- tive decisions on the amount of resources we chose to reallocate largely from the working population to those who are on both ends of life.

As a society, we provide education for our children, and income and health care for the elderly. Unfortunately, the tradeoff between the two is not often analyzed.1 With fixed resources, at any point in time the more we spend on one, the less we have to spend on the other — or the less workers have to consume. Currently in the U.S., most of the public transfers to the elderly occur at the national level, through the federal budget, while many of the decisions on education spending are made at the state and local levels.

One exception to this general division of responsi- bility is Medicaid, which provides health care to parts of the lower-income population, including the dis- abled and elderly, and is funded by both the federal (55%) and state (45%) governments. Medicaid allows states some discretion as to whom they cover and how they deliver services, but dictates a minimum level of benefits. An expansion of the Medicaid program, by an additional 15 million people, is one of the ways last year’s health reform legislation provides coverage to the uninsured.

This program, along with health spending for pub- lic employees and retirees, is putting increasing pres- sure on other state spending, particularly funding for education, which the states in turn are reducing and pushing down to the local level. While localities do not have to fund Medicaid, they do have health costs for employees and retirees putting similar pressure on education and other local responsibilities.

Before reaching any inherently subjective conclu- sions, such as how much is “enough,” it is important to understand more about what health spending buys, and for whom.

Figure 3 illustrates a perhaps obvious point that hospitals and physicians make up the majority of costs, and when lab work and other related services

are added, nearly 60% of the total. What may be more surprising is how relatively little is spent on some sec- tors, such as prescriptions and public health.

Over time there has been a dramatic shift away from private financing of health care (insurance and out- of-pocket) to financing by and through government. Last year private insurance covered only about 1/3 of all spending, and out-of-pocket costs to patients amounted to 12% of the total.

Most of this spending in any given year is incurred on behalf of a relatively small proportion of the popu- lation. Standard rules-of-thumb — 5% of the popula- tion spend 50% of the costs; 20% of the population make up 80% of the costs — suggest how concentrated health care needs are. In Figure 4, these “expensive” patients appear in the right-hand side of the graph. Included here are episodic expenses for maternity, a case of pneumonia, or trauma from an accident. Yet the vast majority, 75%, of spending in this right-hand side of the graph is for patients with chronic disease.

High Cost of Health Care Is Due to Chronic Disease, Not the End of Life Advances in the prevention and treatment of infec- tious disease, the widespread adoption of water treat- ment in developed countries, the ability to diagnose and treat chronic illness, and the decline in smoking, have all contributed to an increase in the number of years we can expect to live.

As we live longer, however, we tend to develop chronic disease, and often several chronic illnesses concurrently. For example, the “average” Medicare

Figure 3 The Composition of Health Care Spending in 2009

cost and end-of-life care • summer 2011 125

Crippen and Barnato

patient in the top quarter of spending has five chronic conditions, sees 12 separate physicians over the course of a year, fills 50 prescriptions for various drugs, and is hospitalized at least once. For some, this treatment occurs at the end of their lives. About 27% of all Medi- care spending is used each year for people who die, a proportion that has been stable over several decades.2 The stability of this concentration suggests that what-

ever is affecting overall cost increases is affecting dece- dents and survivors alike. It is not increases in end- of-life spending that is driving up the cost of health care, but overall health care costs propelled by the increase in chronic disease, technology to detect and treat it, and in the costs of treating patients with mul- tiple chronic conditions.3 In Medicare, most expensive patients live on for several years, and incur high costs such as repeat hospitalizations in subsequent years.

It is not just the elderly who develop chronic dis- ease. Increasingly, children are developing diabetes and associated conditions, often due to obesity. More asthma is being diagnosed. Patients with addiction and behavioral health issues, including depression,

often have chronic illnesses and are expensive to treat.

In fact, behavioral health issues are prevalent enough throughout all age groups that it will be dif- ficult to control costs without adequately addressing behavioral health as a co-morbidity. Figure 5 attempts to separate spending for typical chronic conditions by behavioral health diagnoses. This Medicaid popula-

tion (which is undoubtedly more prone to issues of mental health and addiction), suggests the number of hospitalizations in a year may be 3-6 times more for patients with behavioral health complications, result- ing in as much as a $30,000 difference in the annual cost of treatment.

Ultimately it is the complexity of treatment of multi- ple diseases, and our inability to treat complex patients satisfactorily with current practices, that is driving

health spending on chronic disease. Dr. Gerard Anderson, updating some of

his previous work on the nature of chronic disease in America, recently concluded:

Unfortunately, while our health care needs have evolved, often the health care system has not. It remains an amalgam of past efforts to treat infec- tious diseases and acute illnesses. It does not focus on today’s current and growing problem — increasing numbers of people with chronic condi- tions, especially those with multiple chronic conditions…. Many people with chronic conditions have multiple chronic conditions and this neces- sitates multiple caregivers. The cur- rent system provides few incentives for physicians and other caregivers to coordinate care across providers and service settings. We know that many people with chronic conditions report receiving conflicting advice from differ- ent physicians and differing diagnoses for the same set of symptoms. Drug-to- drug interactions are common, some-

Figure 4 Per Capita Health Spending as a Percent of the Population

It is not increases in end-of-life spending that is driving up the cost of health care, but overall health care costs propelled by the increase

in chronic disease, technology to detect and treat it, and in the costs of treating patients with multiple chronic conditions.

126 journal of law, medicine & ethics

S Y M P O S I U M

times resulting in unnecessary hospi- talizations and even death. People with chronic conditions are getting services, but those services are not necessarily coordinated with one another, and they are not always the services needed to maintain health and functioning.4

As exemplified by the expensive Medicare patients cited above, the dozen doctors delivering care are an assortment of largely specialists who are trained to treat specific diseases. The coordination of care, where it occurs often falls to general practitioners and inter- nists, who are paid less and who often do not have even the basic information from all the providers. Bet- ter health information systems will help this task, but it is far from sufficient to ensure success.

More important is the development of more knowl- edge about how to treat multiple conditions in one patient. Dr. Cynthia Boyd of Johns Hopkins Univer- sity5 compared the recommended treatment for an elderly woman with five typical conditions for a female of her age: diabetes, hypertension, arthritis, osteopo- rosis, and difficulty in breathing (COPD). Applying the best-evidence medicine for each of the five conditions resulted in a large pharmacy load (12 medications, 20 pills a day), contradictory exercise regimes, and con- flicting dietary advice. Dr. Boyd concluded:

Although CPGs (Clinical Practice Guidelines) provide detailed guidance for managing single diseases, they fail to address the needs of older patients with complex co-morbid illness…. It is evi- dent that CPGs, designed largely by specialty-dom- inated committees for managing single diseases, provide clinicians little guidance about caring for older patients with multiple chronic diseases.6

The Department of Health and Human Services has begun to recognize this challenge and issued a report in December 2010, entitled, New Strategic Framework on Multiple Chronic Conditions.7

Health Spending across Generations Thus far we have considered how current spending for health care could crowd-out spending for other needs across the country, looking particularly at allo- cations for children. Again, it is not self-evident that the current allocation between the young, workers, and retirees is somehow wrong or unethical. Rather, it is demonstrative that the high cost of health care, if it could be reduced, could free up resources to be used for the young, to reduce public deficits, or returned to taxpayers.

It is possible to examine how the current distribu- tion of resources will be affected in the future. Current deficits, to the extent they continue, will push more of the funding for today’s federal programs into the

Figure 5 Hospitalizations for a Medicaid Population, by Disease and Behavioral Health Status

Source: Center for Health Care Strategies

cost and end-of-life care • summer 2011 127

Crippen and Barnato

future, leaving interest and principal payments for generations to come.

Figure 6 depicts the finances of the federal gov- ernment over a long period, wherein total revenues amounted to about 18% of the economy (GDP) and spending around 20% of GDP. In the last two years, revenues have fallen and spending has increased by historic amounts, resulting in deficits of 9-10% of GDP.

Perhaps more important, the outlook for the future, while improved, is the continuation of large deficits. Revenues do increase over time (this graph assumes the Bush-era tax cuts are not made permanent) and spending declines somewhat before the impending retirement of the baby boom generation begins to push retirement and health spending up rapidly.

In perhaps more understandable terms, the federal government currently spends the equivalent of roughly $30,000 per household in the United States, collects $20,000, and borrows $10,000. The $10,000 in new debt will be added to the $90,000 per household in previous borrowing. Interest payments on the debt amount to about $2,000 of the $30,000 in spend- ing. With the continued deficits in the current budget, interest payments will rise to $8,000 per household by 2020, just a few years from now. These numbers all pale in comparison to the approximately $750,000 per household in unfunded future promises — prom- ises made mostly for pensions and health care.

The $10,000 for each household in new debt this year covers the deficit for the entire federal budget, not just spending for health care. In that sense, the borrowing could be attributed to any, or all, federal programs, including the wars in the Middle East. However, the ever-increasing health care costs in the budget, propelled by the retirement of the baby- boomers, will require very significant tax increases, cuts in other programs, or continued unprecedented borrowing.

Figure 7 suggests that these three federal programs for retirees and disabled will, in less than 20 years, require funding levels near the entirety of what we currently collect in federal revenues. Tax increases to cover these programs (while maintaining the rest of the budget) would amount to 4-5 times the size of the Bush-era tax cuts. These additional taxes would have to come largely from workers, rich and poor alike, to cover the gap. Borrowing to fill the hole would require yearly debt increases in the 100s of billions of dollars — amounts that our creditors around the world may, at some point, be reluctant to provide.

Without conjecturing how this might all work out, it is fairly clear that increasing health care spending, particularly for the elderly baby boomers, on a scale promised under current law will likely result in at least some increased borrowing (along with tax increases and spending cuts in other programs). To the extent it does, future generations will be paying for the boom-

Figure 6 Federal Revenues and Spending as a Percent of GDP

128 journal of law, medicine & ethics

S Y M P O S I U M

ers’ retirement, not just the boomers’ children as the basic social compact would suggest.

State and local governments, already under extreme financial pressures, will fare no better. As noted earlier, some states are approaching the time when spend- ing on health care will overtake that for education. In the future, sometime before mid-century, health care spending will exceed all other state spending. The current disparity between young and old will be greatly exacerbated and for those states who manage to borrow, the onus on future generations will be even larger.

(A Word about) Health Reform With the passage of health reform legislation last year, there is the prospect of change in the patterns iden- tified earlier. The primary objective of the legislation is to expand financing and coverage of health care for previously uninsured citizens.

On that score, if the legislation is successfully imple- mented, as many as 30 million more Americans will have some form of health coverage starting in 2014. It should not be surprising that this coverage will come at a substantial cost since giving more people

more health care is the intent. The short-term debate is over whether tax increases and spending cuts in the leg- islation will be enough to cover these increased costs.

The longer-term issue, however, is whether other changes — in payment systems, additional research on what works, and more decision making outside normal congressional pro- cesses — will help hold down future cost increases. Skeptics abound and even proponents admit it will be years before we know how effective these changes can be. It is particularly uncertain that any of the changes will alter health care delivery, and there- fore costs, for the 75% of spending devoted to chronic care. Suffice it to say, rapidly growing health care costs will be a major concern and public issue for at least the next decade, and

likely much longer. One case study is the State of Massa-

chusetts. In 2006, Massachusetts enacted sweeping health care reform similar to that passed by the Congress earlier this year. In many fundamental ways, Massachusetts was a model for the federal legislation. The reforms are also similar in that Massachu- setts was successful in providing coverage

for many more of its uninsured citizens. Costs, how- ever, have gone up considerably — much more than anticipated when the legislation passed — to the point where total for health care spending is 40% of the entire budget, almost twice that for education.

Conclusion A large disparity exists between resources consumed by today’s children relative to their retired counter- parts. A considerable amount of the difference can be attributed to spending on health care. To the extent a question of ethics arises, it may come down to whether we are spending “enough” on our children to support them in their youth and prepare them for the future.

This disparity will grow over the foreseeable future and, based on current law and experience, cause more debt to be accumulated for upcoming generations to pay. We are financing some of today’s spending on health care by effectively borrowing from future gen- erations. We will likely do so in ever-larger amounts in the future. The ethics of foisting off the payment for our health consumption on unborn citizens provokes a sharper dilemma.

Figure 7 Spending for Social Security, Medicare and Medicaid as a Percent of GDP

Unless we change the practice of medicine and reduce future costs, and explicitly address the ethical dilemmas we face, there may come a time when our kids simply cannot afford us.

cost and end-of-life care • summer 2011 129

Crippen and Barnato

In both cases — allocation of resources across our current population and the practice of borrowing from future generations to fund current consumption — health care plays a major role. While end-of-life spending is significant, and any reduction would help ease both cases, it is the ongoing health care of citizens with multiple chronic conditions that poses the great- est challenge in resolving these dilemmas, for current and future generations.

The good news is that it is possible to give patients with chronic disease better health, and thereby reduce costs, by (among others): emphasizing, and paying for, coordination of care; recognizing the role behavioral health plays in physical health; and increasing patient compliance with treatment protocols. In addition, a national research agenda should be developed and funded to better understand the interactions of chronic conditions and their treatment, including pharmacy, and the role of patients in developing and implement- ing disease management. Unless we change the prac- tice of medicine and reduce future costs, and explicitly address the ethical dilemmas we face, there may come a time when our kids simply cannot afford us.

FIGURE LEGENDS Figure 1. Life-Cycle Income and Consumption for the United States Over the course of a lifetime, there are distinct periods where we consume more than we produce (as children and retirees) and periods where we produce more than we consume (as working-age adults). The patterns vary by country, economy, and culture, and are reflective of the public programs in place to redistribute income from current workers to those too young or old to work.

Figure 2. Per Capita Spending on Children and the Elderly (by Type of Spending in the United States) In the U.S., we provide much more support to our indi- vidual retirees than we do our children. The single most important reason for the difference is the dispar- ity of spending on health care.

Figure 3. The Composition of Health Care Spend- ing in 2009 National surveys allow estimates of the breakdown of health care dollars among the various providers of care, and other systemic costs such as public health.

Figure 4. Per Capita Health Spending as a Percent of the Population There is not much validity to citing an

“average” patient. While there are certainly people in the middle, there are essentially two groups of patients: those who don’t spend much in any given year, and those who spend a lot. The general rules of thumb are: 5% of the population drives 50% of total spending in a year; 20% of the population drive 80% of total costs. The mirror image is important for policy as well — 80% of the population spends only 20% of the total costs.

Figure 5. Hospitalization for a Medicaid Population, by Disease and Behavioral Health Status Health care costs are often characterized as the expenses for treating physical maladies. However, those physical health costs, especially those for chronic care, are often compounded by behavioral health problems, such as depression or addiction. For this population it is not the cost of dying that drives costs as much as the cost of living.

Figure 6. Federal Revenue and Spending as a Per- cent of GDP Figure 6 illustrates a number of points, but for these purposes perhaps the most important is that even after post-recession revenues recover (to nearly historic highs) and stimulus spending disap- pears, there is a substantial and growing gap between the two, resulting in large deficits that continue for decades.

Figure 7. Spending for Social Security, Medicare and Medicaid as a Percent of GDP These three programs, driven by retirement of the baby boomers and the ever-increasing cost of health care, will require nearly all the federal revenues collected in 2030.

References 1. J. B. Isaacs, Public Spending on Children and the Elderly From

a Life-Cycle Perspective, Brookings Institution, Washington, D.C., November 2009.

2. C. Hogan, J. Lunney, J. Gabel, and J. Lynn, “Medicare Ben- eficiaries’ Costs of Care in the Last Year of Life,” Health Affairs 20, no. 4 (July 2001): 188-195.

3. A. E. Barnato, M. B. Mcclellan, C. R. Kagay, and A. M. Garber, “Trends in Inpatient Treatment Intensity among Medicare Benefi- ciaries at the End of Life,” Health Services Research 39, no. 2 (April 2004): 363-376.

4. G. Anderson, Chronic Conditions: Making the Case for Ongo- ing Care, Johns Hopkins University, November 2007, at 5-6.

5. C. M. Boyd, J. Darer, C. Boult, L. P. Fried, L. Boult, and A. W. Wu, “Clinical Practice Guidelines and Quality of Care for Older Patients with Multiple Comorbid Diseases,” JAMA 294, no. 6 (2005): 716-724.

6. Id., at 720. 7. Department of Health and Human Services, New Strategic

Framework on Multiple Chronic Conditions, December 2010.

Copyright of Journal of Law, Medicine & Ethics is the property of Wiley-Blackwell and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.