Consider the American health care system as an automobile. Four dashboard meters provide information on overall vehicle performance. The meters tell us how the engine is functioning in terms of cost, quality, access, and equity. Currently, all 4 meters indicate an urgent need for repair. Our health care system uses up a higher proportion of gross domestic product than any other system in the world.1 Yet the quality of care it delivers is appallingly low.2,3 It fails to cover even basic services for over 40 million people.4 And it apportions care irrationally and inequitably, such that minority groups with the greatest burden of illness receive the fewest services.5
Repairing the vehicle can be tricky. A fix applied to tune up one aspect of the system may inadvertently reduce performance in another. Occasionally, policymakers apply innovative fixes that enhance multiple areas of performance at once. Such was the case when, in the 1970s, the state of New Jersey enacted a set of reforms regulating the payment of hospitals. The reforms established fixed payment rates for hospitalizations, based on diagnosis-related groups, which served as the template for Medicare's Prospective Payment System.6 All payers—Medicare, Medicaid, private insurers—were required to pay the same rates. These rates included a surcharge to maintain a fund to pay hospitals for uncompensated care.7
New Jersey's adoption of all-payer rate setting was motivated primarily by the need to stem rising costs, and capping rates did control hospital costs, at least initially, in a fee-for-service environment that had previously incentivized providers to rack up large bills.6 Equal payment for patients unable to afford care also reduced cost shifting and enhanced access and equity. And uniform payment rates created incentives for hospitals to compete for patients and payers based on quality.
Although all-payer rate setting was far from a perfect solution,7 a majority of states adopted it in one form or another.6 Eventually, though, the fix proved not to be durable. In New Jersey, Medicare stopped paying the state-established rates, cost shifting resumed to make up for low Medicare payments, and payers began, in the new era of managed care, to demand the right to negotiate rates lower than those set by the state. In 1992, New Jersey was forced to dissolve all-payer rate setting. In its place, they enacted the Health Care Reform Act (HCRA), which essentially allowed hospital rates to be dictated by free market mechanisms.7 To enable open (but not necessarily fair) competition, the uncompensated care surcharge on hospital payments was lifted.
The HCRA was one of those “fixes” that targeted one aspect of system performance—i.e., controlling costs—at the predictable expense of the others. Volpp et al.,8 in this issue of JGIM, address whether the transition from all-payer rate setting—with its explicit focus on reducing cost shifting and providing full payment for the care of poor patients—to market-driven payment, exacerbated racial inequalities in the outcomes of hospitalized patients. They analyzed inpatient mortality rates for African American and white patients in New Jersey with 7 common conditions, before and after the implementation of HCRA, and compared these rates with those in New York, which did not abandon all-payer rate setting until later. They found that, overall, inpatient mortality rates declined in both states for all groups except uninsured African Americans in New Jersey. In that group, mortality from the 5 conditions for which hospitalization is generally considered to be non-discretionary rose from 12.3% to 13.9% (see erratum in the forthcoming December 2006 issue of JGIM). In stratified analyses, racial disparities in New Jersey's mortality rates relative to New York's worsened after HCRA for acute myocardial infarction (AMI), but not for the 6 other conditions.8 Prior studies by this group of investigators demonstrated, similarly, that mortality rates for the uninsured in New Jersey compared with New York worsened after HCRA for AMI, but not for the other conditions examined.9,10
On the surface, these findings might seem somewhat reassuring, in that no systematic rise in mortality or exacerbation of disparities was observed. On closer consideration, though, the findings are disturbing. First, AMI was probably the one condition most “sensitive” to changes in payment policies. It is a condition with a relatively high case fatality rate that has declined dramatically over time with the advent of technologically sophisticated, lifesaving coronary interventions, and effective but expensive new pharmaceuticals. In their previous work, Volpp et al.9 demonstrated that the use of coronary revascularization in New Jersey was lower than in New York after HCRA, offering one potential explanation for the rise in relative AMI mortality in New Jersey. For the other conditions examined—hip fracture, stroke, pulmonary embolism, gastrointestinal bleeding, congestive heart failure, and pneumonia—survival to hospital discharge is not as dependent on the use of expensive technology and drugs and is therefore likely not to be as cost-sensitive. Second, mortality is a blunt indicator of quality. If the change in policy resulted in higher relative mortality rates and worsening disparities at all, for any condition, it represents an alarming system failure. If the implementation of HCRA was associated with excess AMI mortality, one can presume that it was also associated with other, less drastic, more “proximal” adverse outcomes, such as the development of cardiomyopathy, slower recovery and return to work, and lower quality of life. In other words, the observed consequences of HCRA on AMI mortality rates may represent only the tip of the iceberg of its negative impact on quality and equity.
It makes sense that price competition—with its resulting reductions in hospital margins, fewer available funds to cover uncompensated care, and incentives to limit the use of expensive services—might adversely affect uninsured patients with AMI. And the simplest explanation for the worsening racial disparities in relative AMI mortality observed in this study, as suggested by the authors, is that African Americans are more likely to be uninsured or publicly insured, and that in a competitive environment, hospitals might limit services most for these patients.8 It is not readily apparent, from the results presented in the paper, whether the exacerbation of racial disparities in AMI mortality can be attributed solely to higher rates of uninsurance among African Americans. An interesting alternative hypothesis is that the time pressures and bottom-line focus engendered in a competitive market environment may exacerbate well-documented, existing disparities in the care of AMI.8,11 Numerous studies have documented wide and persisting disparities in the use of coronary reperfusion and revascularization between African Americans and whites with AMI.12,13 The root causes of these disparities are unclear, but there is some evidence that physicians may consider African Americans less appropriate candidates for coronary interventions14 and may be less aggressive in recommending these procedures to African Americans.15 African-American patients, in turn, may be more passive in discussions of invasive coronary procedures.16 In an environment where time pressures and financial incentives discourage procedure use, African-American patients may therefore be more likely to “slip through the cracks” and receive lower quality care for AMI.
American policymakers seem intent on keeping our current, ailing health care vehicle running, using incremental, market-oriented fixes whenever the engine starts to sputter. They argue, as did Adam Smith,17 that the Invisible Hand of the free market will reduce costs and improve quality for all. Health care, however, is not like other “commodities.” The quality of health care is difficult for even sophisticated consumers and payers to judge, making it easy for providers to compete on cost alone, at the expense of quality and equity.11 Furthermore, for many conditions, patients do not have the luxury of shopping for the best deal. A patient with crushing chest pain cannot tell her ambulance driver to take her to the hospital of her choice. Finally, because health care is so expensive, even the most powerful market forces are unlikely to make it affordable without insurance coverage. Insurance will continue to shield consumers from the true cost of their care, making them less-than-thrifty buyers who are likely to accept services offered by the very providers who stand to profit from them. Meanwhile, those without insurance will continue to access care late in the course of their illnesses,18 incurring high-cost services, which will go unreimbursed, resulting in shifting of costs to insured consumers, higher premiums, less affordable insurance, more uninsured persons, and so on. In health care, the Invisible Hand does not push us all forward; it merely widens the divide between the haves and the have-nots.
By applying serial, incremental, market-based fixes to our failing health care system, we are effectively replacing expensive parts on a gas-guzzling jalopy. We seem to love this old car, those of us who get to ride in it. But each fix seems to be less and less effective. And in the meantime, driving this vehicle flies in the face of the principles of efficiency and justice upon which our economy and nation are built. It's time to junk this old heap and get a new one.