Health reform and public health: will good policies but bad politics combine to produce bad policy?

• Author: Pollack, Harold A.

1. PUBLIC HEALTH GAINS IN HEALTH REFORM II. THE POLITICAL CHALLENGE III. SUCCESSFUL MODELS OF POLITICALLY SUSTAINABLE PUBLIC HEALTH MEASURES IV. DESIGNING MORE SUSTAINABLE PUBLIC HEALTH POLICIES The enactment of the Patient Protection and Affordable Care Act (PPACA) (1) was an incomplete victory and will remain so even if the new Republican congressional majority does not curtail its provisions. The legislation has many shortcomings and compromises, Most importantly, it could have brought help sooner to millions of uninsured or under-insured Americans.

   Despite these compromises, public health researchers and practitioners have reason to celebrate. (2) Simply put, PPACA fundamentally altered and improved the public health infrastructure of the United States. Fully implemented, PPACA promises to markedly improve clinical preventive services and transform our nation's response to traditional centerpiece public health concerns, including HIV/AIDS, substance abuse, mental health disorders, and other conditions.

   In this Article, I note several ways in which provisions of PPACA promise to improve population health. (3) But this Article is less sanguine regarding the politics of public health. I submitted the first draft in early January 2011, the very week PPACA's opponents assumed the majority in the House of Representatives. (4) The Speaker of the House has vowed to repeal PPACA, (5) and twenty-six states support constitutional challenges to the individual mandate, a central pillar of the new law. (6) PPACA's ultimate success depends on the executive branch's ability to implement successfully one of the most complex policy reforms in American history against a backdrop of fiscal constraint and partisan acrimony. (7)

   As a sometime participant and advocate in this process, (8) I am struck in hindsight that the bill's fervent defenders did not focus more explicitly or effectively on these political matters in crafting the final bill. Many commentators noted that key provisions of PPACA would likely become politically impregnable once they became part of the fabric of American life. (9) On many of these provisions, PPACA proponents had less success in structuring the Act to ensure that these provisions would achieve such an embedded status. (10)

   Perceptive PPACA supporters certainly understood from the start that back-loading implementation was a fundamental political problem. (11) Executive branch and congressional advocates, along with outside activists, worked hard to win passage of the best possible bill consistent with political constraints, However, supporters appear to have been less effective in considering how PPACA might be optimally structured to withstand likely attack should its critics win an emboldening majority. As a result, the reformers missed opportunities to enact and defend effective public health policies.

1. PUBLIC HEALTH GAINS IN HEALTH REFORM (12)

   The contributions of PPACA to population health are easily overlooked. PPACA, like most health policy legislation, was not primarily designed to address public health services or public health concerns. (13) In this, the bill reflected the unbalanced political economy of American health care, which devotes a disproportionate share of public and private resources and political attention to personal medical services and tertiary care. (14) Although elected officials across the ideological spectrum cite the importance of public health, (15) public health investments often fare poorly in legislative bargaining over scarce resources.

   Moreover, many analysts who recognize the centrality of social determinants of health argue that expanded access to medical services is likely to have minimal impact on population health. Distinguished health economists Dana Goldman and Darius Lakdawalla argue with particular force that PPACA may be overpromising health improvement:

   Advocates of universal coverage often get confused on this point. They equate good health with having health insurance, and cite myriad academic studies. The problem is that these studies don't account for all the other differences between the insured and uninsured--what they eat, where they live, whether they smoke or drink, the amount of stress in their lives, and even their genetic predisposition to disease. No healthcare system is good enough to fully compensate for bad behavior and poor environmental factors. Perhaps the strongest and earliest such evidence came from the RAND Health Insurance Experiment (HIE), which randomly assigned families to health insurance plans of varying generosity. One of the main findings of this experiment was that families in the least generous plan (95 percent coinsurance) spent nearly 30 percent less on medical care--with little or no difference in their health. (16) These authors do not identify specific advocates who equate good health with having health insurance, or who deny the importance of behavioral or environmental risk factors. As a result, it is unclear how many advocates of universal coverage are actually confused or naive on these basic points.

   Still, the authors' point is well-taken. Health insurance coverage does not address many threats to individual and population health. Balanced policies must attend to social determinants of health-education, housing, accidents and violence, public safety, workplace safety, environmental protection--as well as to reducing barriers to personal medical services. Social policy is health policy in each of these areas. (17)

   Health policy researchers are increasingly aware of the dangers in overstating the link between insurance and health. At the same time, it is tempting to misapply social epidemiology to wrongly dismiss the importance of expanded health coverage. There is substantial evidence, much of it compiled in the Rand Health Insurance Experiment (HIE), that expanded access to care improves health outcomes, particularly in patients with cardiovascular risk-factors and hypertension. (18)

   The experiment was too short to directly investigate mortality differences across the different treatment plans. (19) However, study investigators were able to explore statistical models of human mortality attributable to basic measurable risk factors. (20) Within these analyses, low-income HIE participants enrolled in high-deductible plans displayed notably higher predicted mortality than did their otherwise comparable peers enrolled in a free care plan. (21)

   Almost all of the predicted mortality reduction reflected improved hypertension detection and treatment. (22) Subsequent studies support these findings. (23) Goldman and Lakdawalla themselves note that "poor people with high blood pressure had slightly higher [predicted] mortality rates" when assigned to high-deductible rather than free health care plans. (24)

   As these studies demonstrate, health insurance coverage facilitates the receipt of basic health services, which in turn enables primary care providers or others to detect hypertension and to provide accompanying treatment through inexpensive medications. Access to such care rarely produces large changes in health behavior or in social determinants of health, (25) but it does give individuals access to basic, effective treatments that improve health and prolong life.

   Using nationally representative longitudinal data from the Health and Retirement Study, J. Michael McWilliams and colleagues found substantially higher eight-year mortality rates among uninsured individuals ages fifty-five to sixty-four than among facially similar insured peers. (26) Mortality differences were especially stark within the lowest quartile of household income as well as among adults with diabetes, hypertension, or heart disease. (27) Mortality hazard rates among the uninsured were more than fifty percent higher than those observed among the insured in both the low-income and the cardiovascular-illness groups. (28)

   Given the non-experimental nature of these analyses, it is possible that unobserved factors account for the observed mortality differences. Yet the authors conducted sensitivity analyses which suggest that complete confounding is unlikely:

   [A]n unobserved factor similar to smoking in prevalence (approximately 25 percent of the study cohort) and its association with insurance status (relative risk of being uninsured equal to 1.66) would have to be associated with a relative eight-year mortality risk of 2.65 for the association between insurance status and mortality to become non-significant when further adjusted for this unobserved factor. In comparison, smoking was associated with a relative eight-year mortality risk of 2.48. (29) Medicare studies provide further support for the causal importance of health coverage. In a series of other studies, the same researchers documented improvement in many health measures when individuals reached the age of Medicare eligibility, along with surprisingly large reductions in racial, ethnic and educational disparities,s[degrees] Racial disparities in systolic blood pressure decreased by approximately sixty percent] Educational, racial, and ethnic disparities in blood glucose control decreased by more than seventy-five percent. (32) Educational disparities in total cholesterol levels became negligible. (33) Receipt of key preventive services, such as mammography and prostate cancer screening, also sharply increased at age sixty-five among the previously uninsured. (34)

   Expanded Medicaid eligibility is an especially important institutional shift. (35) Although expanded Medicaid is usually discussed as a vehicle to reduce the number of uninsured, it has particular implications for the patchwork of providers and services in safety-net care.

   Medicaid is currently a means-tested categorical program which only covers certain types of low-income individuals--for example, children under the federal poverty level, or those who are eligible for specific programs such as Temporary...