Richard “Buz” Cooper, M.D., who died on January 15, 2016, had two different stays at the University of Pennsylvania. He was a former chief of hematology in the Department of Medicine and a pioneering cancer researcher. Back then, he wrote the grant proposal to create what would become the University of Pennsylvania Cancer Center (now the Abramson Cancer Center) and served as its first director until 1985. He then became dean and vice president of the Medical College of Wisconsin. More recently, he returned to Penn as a senior fellow in the Leonard Davis Institute of Health Economics. His focus had changed by that time, as his passionately but meticulously argued book, Poverty and the Myths of Health Care Reform, makes clear. In it, Cooper contends that the leading cause of the nation’s high health care costs is not physician-induced demand, waste, and fraud, as many influential organizations assert. Instead, he argues, it is poverty.
With the permission of the Johns Hopkins University Press, here is an excerpt from Cooper’s Introduction.
It seems fair to ask, how did I come to write this book? After all, I am a physician, a specialist in hematology, the study of blood diseases. I practiced medicine, led an academic hematology group, and conducted research for 25 years. In 1985, I returned to Milwaukee, the city of my birth, to become dean of the Medical College of Wisconsin, and that started the detour that led directly to this book.
As my deanship was drawing to an end, Bill and Hillary Clinton were developing their health care plan, and it was in this context that I was confronted by two questions that would redirect my career. The first concerned physician supply, and the second concerned poverty, which is the focus of this book. One led to the other, and understanding something about the first will help in understanding why I set out to answer the second. At issue was whether the United States would soon have too many physicians, as was projected by the federal Bureau of Health Professions (BHPr) and accepted by most policymakers. If so, action was necessary to avert a physician surplus.
The question confronting me was whether the projected physician surplus was valid, and the answer proved to be no. The BHPr’s projections were wrong in two ways. First, they overestimated the future per capita supply of physicians because they had underestimated future population growth. Second, they underestimated demand because future projections were based on care as it existed at the time, never considering that new therapies and procedures would require more physicians in the future. When these shortcomings were appreciated, it became apparent that, rather than a surplus of physicians, there were likely to be shortages soon after the turn of the century, only a decade ahead.
While forecasting the future demand for physicians was not difficult, it did prove difficult to change the minds of those who believed in surpluses. But the dominant view at the time was that there would be as many as 100,000 too many physicians in the year 2000.
Despite certainty within the policy community that surpluses would soon appear, the year 2000 was greeted by none, and the years that followed saw deepening shortages. In 2002, I published a paper in Academic Medicine entitled “There’s a Shortage of Specialists: Is Anyone Listening?” Those who were listening included not only a vast array of state medical and hospital associations but also the same professional organizations that only a few years earlier had called for caps on residency training. Even the Council on Graduate Medical Education, which had been a prime mover in popularizing the BHPr’s notion of surpluses, changed its long-standing position that there would be 100,000 too many physicians in 2020 to one stating that there would be almost 100,000 too few.
Most policymakers were not listening, however, which is why the “caps” have held firm, creating the shortages we have today. But the basis for holding firm ceased to be the BHPr’s erroneous projections of physician surpluses. Rather, it was a belief that physician practices are wasteful and inefficient and driven by supplier-induced demand – Roemer’s law – and therefore more physicians would be undesirable. It was supported by a growing body of data from the Dartmouth Institute that attributed geographic differences in health care spending among regions of the country to the unwarranted overuse of supply-sensitive services. This view was shared by a broad coalition of agencies, foundations, and academics.
Health Care Spending: Why Is It So High?
Initially, this second question was not associated with any national policy issue. It arose in the context of a pragmatic local concern. Why were health care costs much higher in Milwaukee than elsewhere in the upper Midwest? My colleagues and I explored many possible reasons, but it was only when we examined the distribution of costs, neighborhood by neighborhood, that the answer emerged. In the 30 years I had been away from Milwaukee, its black population had burgeoned and the city had become the most segregated in the North, more segregated than Detroit. Social problems were legion. We found that patients who resided in Milwaukee’s highly segregated “poverty corridor” had hospitalization rates much higher than among those living elsewhere, so much higher that they accounted for the entire excess utilization of care in the Milwaukee region as a whole.
The critical observation is that Milwaukee’s poorest were its sickest and used the most care. This proved to be the rule in other communities, as well. Chapter 1 takes a journey along two subway routes in New York City, where incomes swing from poverty to wealth and back to poverty over the course of only a few stops and where rates of disability and hospital utilization track poverty all the way. Chapter 3 provides a detailed view of Los Angeles, which has more poor people than most cities have people and where low-income patients lift health care costs to among the highest in the nation. In each case, poverty distinguishes areas where health care spending is high from others where it is low. But a word of caution. Don’t blame the victim! Poor patients do not use more health care because they wish to. They do so because their health is poorer and their social circumstances are weaker. The basis for their high health care spending is embedded in the fabric of their lives.
When my colleagues and I began to examine why health care costs were higher in Milwaukee, we thought we were addressing a local problem. However, it proved to be part of a national dialog that was unfolding from publication of the studies by John Wennberg and his colleagues, using their newly created Dartmouth Atlas of Health Care. The Atlas divided the United States into 306 hospital referral regions (HRRs), based on where most patients received most of their care. Dartmouth researchers documented marked differences in Medicare expenditures among these regions, and Milwaukee was among those with higher spending. However, the diagnosis made by the Dartmouth group was quite different from ours. Rather than attributing Milwaukee’s higher spending to poverty, they attributed it to the overuse of “supply-sensitive services,” reminiscent of the supplier-induced demand that Roemer had popularized.
Why did the Dartmouth group consistently fail to recognize the central role of poverty? After all, they acknowledged that low-income people are sicker and that sick people require more care. Yet they persistently claimed that “regional differences in poverty and income explain almost none of the variation.” Others concurred, including influential committees of the Institute of Medicine. However, the major methodological reason is that the Atlas aggregates all of the data from all of the people residing within each HRR. The approximately 1.6 million people in Manhattan and the 10 million in Los Angeles [County] are distilled into single numbers. Economic distinctions between places as different as Harlem and Park Avenue and South Los Angeles and Beverly Hills disappear. Indeed, it was only by disaggregating HRRs into their constituent zip codes that my colleagues and I were able to discern the enormous impact of poverty on health care utilization in these and other areas.
Silencing Poverty
The Dartmouth Atlas was not alone in ignoring poverty. Poverty was not on the political agenda in the years leading up to Clinton’s Health Security Act, or in the 15 years between that and ObamaCare, or throughout President Obama’s first term. Instead, the president repeatedly pointed to the lower health care spending in small towns, like Green Bay, Wisconsin, and Grand Junction, Colorado, which are devoid of concentrated poverty, never mentioning the dense poverty and high burden of disease in other areas, such as on the south side of Chicago where he had been a community organizer. Seattle and Salt Lake City were offered as models for the nation, while Los Angeles, which has more poor people than these two cities have people, was marked as a place of egregious waste. One could not avoid hearing about the wonders of the Mayo Clinic, located in Rochester, Minnesota, although it is the highest-cost facility in the otherwise low-cost upper Midwest, or about the poor performance of the University of California, Los Angeles, which borders LA’s dense urban poverty.
Throughout this period, the silence about poverty was deafening. Its relationship to health care spending is not mentioned once in any of the more than 20 books on health care reform that grace my library shelf. Yet every physician, nurse, and hospital administrator knows how poverty affects health care utilization. They live it every day.
Beginning in the mid-1990s, John Billings, at the United Hospital Fund in New York, reported that hospital admission rates for chronic conditions were four to five times higher among patients from poor zip codes in New York than among those from rich ones, and the same was true in other large metropolitan areas. And by the end of the 1990s, my colleagues and I had uncovered the enormous contribution of poverty to the high health care spending in Milwaukee and, later, in Los Angeles. Nonetheless, poverty was not on the radar screen of health care reform as ObamaCare was being crafted.
What policymakers did have on their radar screens was that deficiencies in clinical practice were the principal cause of excess spending and poor outcomes. The Institute of Medicine further popularized the notion that 30% of U.S. health care spending is wasted. Curiously, it ignored the possibility that poverty may be a contributory factor. Indeed, in more than 1,500 pages of the IOM’s several reports, “poverty” was not mentioned even once, while “waste” was mentioned more than 250 times.
This body of work, flowing as it has from multiple respected sources, has been taken as “evidence” that more specialists and more spending add no value and that 30% of health care spending is wasted. Yet when viewed through the lens of poverty, each line of “evidence” proves to be a manifestation of the increased care required by patients who are poor. Call it waste if you want. Treating a homeless man’s frostbitten toes is surely a waste, when a pair of shoes could have prevented it.
The Challenge Ahead
It may be difficult to believe that poverty has been blotted from view. After all, ObamaCare is meant to help poor people, and it does. It expands Medicaid eligibility for many who are poor, creates federally subsidized insurance exchanges for others who are near-poor, and funds an expansion of community health centers, which serve poor people. But it was blind to the socioeconomic factors that underlie high health care utilization. For example, it established penalties for hospitals with “excessive” numbers of hospital readmissions, ignoring the reality that most readmissions are of poor patients, and it imposed penalties for higher 30-day mortality rates, failing to recognize that it is the poorest who have the highest rates. Indeed, instead of strengthening the ability of providers to care for poor patients, they are trying to restructure the health care system into something it cannot be.
The inescapable conclusion is that the United States will not be able to constrain its spiraling health care spending without addressing the high costs of caring for patients at the bottom of the economic ladder. But how? Greater attention must be directed to activities that exist beyond traditional health care, such as housing, transportation, and social support, which have been shown to reduce costs and improve outcomes. At a broader level, what is needed is a reduction in income inequality and the creation of a social infrastructure that enables low-income families to exit from the cycle of poverty. Without it, no amount of health care spending will permit all Americans to lead the long and healthy lives they desire, and health care spending will continue its unsustainable upward spiral.