Health care reform is back. By the time this article appears, President Barack Obama may have introduced his health care plan to Congress. The justification that we need comprehensive reform is compelling: in 2005 we spent nearly $6700 per capita—which is nearly twice as much as most other developed nations—for a system that fails to cover 44 million US citizens.1 Despite our expenditures, the United States has adult and infant mortality rates that rank it at the bottom among the world's high-income developed countries and that are about 50% higher than the median mortality rates of these countries.2

It is tempting to connect these problems and conclude that improvements in our health care system will raise our standards of health to those of the high-income countries of Europe or Asia. But a closer look suggests that reforming our health care system, although long overdue, is unlikely to greatly improve our health. For that, we will need to reform not just health care, but our entire system of health.

Scientific journals are filled with studies of the effect of drugs and procedures on specific diseases, but few studies address perhaps the biggest question of all: How important is medical care taken as a whole, to health? The studies that have been published seem to indicate that the effect is fairly modest. Among high-income developed nations, there is no correlation between health care expenditures and adult or infant mortality rates.2 This may seem an unfair test of medical care, when so many causes of death are incurable—for example, lung cancer and homicide. Some would argue that a fairer assessment of the medical care system would be to study its relationship to mortality rates from diseases that are amenable to effective treatment by medical care, such as breast cancer and pneumonia. Beginning in the late 1970s, several groups of researchers have done just that.3,4 These “amenable” causes account for about 25% of deaths.5 Across a series of studies, mortality from “amenable” deaths among developed countries or regions of countries was only infrequently associated with measures of medical care access or utilization, and when it was, the relationships were usually the opposite of what one would predict, with higher rates of “amenable” mortality where more physicians and hospitals were available.6 By contrast, some more recent studies have shown that areas with a greater number of physicians practicing primary care have lower rates of all-cause mortality, but the size of the effect is small compared with the variation in mortality between areas or the effect of socioeconomic factors.7,8 For example, in one study, US counties with the most primary care physicians per capita had a 2% to 3% lower all-cause mortality than did other counties, but counties with higher income inequality had mortality rates that were 11% to 13% higher than those in counties with low income inequality.7

Why is it so difficult to show a positive impact of so much medical care on population health? Probably all of us know people who would not be alive today were it not for medical care, and one model has predicted that medical care has been responsible for nearly half of the fall in mortality rates from coronary heart disease from 1980 to 2000.9 But those successes come at a high price. The Institute of Medicine has estimated that 44 000 to 98 000 deaths occur as the result of medical errors.10 Even that represents just a fraction of the estimated total iatrogenic deaths of 225 000 per year, with the difference being deaths caused by medical care applied according to accepted standards.11 The tools of modern medicine—drugs and surgical procedures—that are powerful enough to save lives are also powerful enough to end them, and in the loosely organized US medical care system, it is not possible to use these tools frequently without sometimes administering them to the wrong patient, in the wrong way, or in a “right” way that still ends tragically.

One could conclude from this that the only problem with our health care system is a shortage of primary care and an excess of specialty care. It certainly seems likely that more-accessible, higher-quality primary care could prolong life in people with the chronic conditions that are so common and so inadequately treated today, such as hypertension and diabetes, while also reducing iatrogenic deaths. If so, then medical care reform could contribute to closing the mortality gap between the United States and other developed countries. But there is reason to believe that wider population-based approaches would contribute far more.

The behaviors that are the most powerful determinants of health in the United States are smoking, unhealthy diet, physical inactivity, and alcohol use.12 Together these few behaviors are estimated to be responsible for about 40% of deaths in the US and have a similarly large impact on disability-adjusted life-years lost in high-income countries around the world.13 Although doctors can influence these problems, they do not arise from lack of medical care, cannot be solved by medical care, and often lead to fatal diseases and conditions (such as lung cancer and homicide) for which medical care can do very little. They can, however, be solved by changes to our everyday environment.

For example, the third leading cause of death in America is stroke.1 We could approach this problem by providing more specialized stroke units in hospitals, which compared with home care reduce mortality and increase symptom-free recovery from 10% to 12%.14 Treatment of the major risk factor of hypertension seems a better approach, however, because it reduces the incidence of stroke by approximately 30%, or nearly 3 times the rate of symptom-free recovery after intensive stroke treatment.15 But even this approach, which would be the goal of many medical care reform plans, benefits only people with hypertension and thus misses the people with “prehypertension” who have strokes and who could benefit from lower blood pressure.16 On the other hand, it has been estimated that a population-wide reduction in sodium intake of 50 mmol (1150 mg) per day, which could result from an achievable reduction in the sodium in processed food, would prevent approximately 50% more stroke deaths than successful treatment of every person with hypertension (even if that were possible, which it is not), because it would benefit people across the entire spectrum of blood pressure.17 Clearly to reduce stroke mortality we will need a combination of stroke units, hypertension treatment, and sodium reduction, but of these 3, the most effective and most cost-effective approach—sodium reduction—is the one most likely to be overlooked in health care reform plans.

The same principle can be seen in our experience with cancer mortality in the last few decades. Thirty-eight years after President Nixon declared a “war on cancer,” and many billions of dollars worth of cancer research later, cancer rates are falling in this country.18 But researchers at the National Cancer Institute have concluded that they are falling not primarily because of advances in screening or treatment but instead almost entirely because of declines in smoking.19

This is not simply an argument for more funding for public health agencies, although that is probably part of the solution. Unfortunately, our existing public health programs do not adequately address the leading causes of death either. For example, less than 2% of funding in big city health departments is used to prevent chronic diseases, even though chronic diseases are the cause of approximately 70% of deaths nationwide.20 Perhaps the greatest achievement in health status in the US in the past 30 years has been caused by the decline in smoking. Although many states and some cities now have strong tobacco control programs, the tools that we have learned are effective in smoking prevention—taxes, smoke-free air laws, counter-advertising—were developed primarily not by public health agencies but instead by outside activists.21 True health reform will require major changes in the structure and approach of public health agencies also. Those agencies should set priorities based on the leading underlying causes of death and focus on policy and environmental changes that will alter these behaviors.

It is exciting that, after several presidents in the 20th century tried and failed to reform a medical care system that has run amok, we may finally fix its biggest flaws. It should be accessible to all, efficient, and oriented to maximize health through primary and preventive care. Americans, who pay $2 trillion a year for this system, deserve that. But if we want to make major improvements in health, we will need to reform not just how we pay for medical services, but our system of health. That means focusing on the behaviors and environments that make us sick. Public health professionals need to become activists, not only in pointing out the benefits of public health approaches but also in redirecting their own focus to our nation's greatest health problems.


I would like to thank Deborah A. Cohen, Thomas R. Frieden, and Mary T. Bassett for their comments on a draft of this article.


1. National Center for Health Statistics. Health, United States, 2007, With Chartbook on Trends in the Health of Americans. Hyattsville, MD: National Center for Health Statistics; 2007. Google Scholar
2. World Bank Group. WDI Online—World Development Indicators. Accessed December 8, 2008. Google Scholar
3. Charlton JR, Hartley RM, Silver R, Holland WW. Geographical variation in mortality from conditions amenable to medical intervention in England and Wales. Lancet. 1983;1(8326 Pt 1):691696. Crossref, MedlineGoogle Scholar
4. Rutstein DD, Berenberg W, Chalmers TC, Child CG, Fishman AP, Perrin EB. Measuring the quality of medical care. A clinical method. N Engl J Med. 1976;294(11):582588. Crossref, MedlineGoogle Scholar
5. Nolte E, McKee CM. Measuring the health of nations: updating an earlier analysis. Health Aff (Millwood). 2008;27(1):5871. Crossref, MedlineGoogle Scholar
6. Mackenbach JP, Bouvier-Colle MH, Jougla E. “Avoidable” mortality and health services: a review of aggregate data studies. J Epidemiol Community Health. 1990;44(2):106111. Crossref, MedlineGoogle Scholar
7. Shi L, Macinko J, Starfield B, Politzer R, Wulu J, Xu J. Primary care, social inequalities, all-cause, heart disease, and cancer mortality in US counties, 1990. Am J Public Health. 2005;95(4):674680. LinkGoogle Scholar
8. Shi L, Macinko J, Starfield B, Wulu J, Regan J, Politzer R. The relationship between primary care, income inequality, and mortality in US States, 1980-1995. J Am Board Fam Pract. 2003;16(5):412422. Crossref, MedlineGoogle Scholar
9. Ford ES, Ajani UA, Croft JB, et al.. Explaining the decrease in U.S. deaths from coronary disease, 1980-2000. N Engl J Med. 2007;356(23):23882398. Crossref, MedlineGoogle Scholar
10. Institute of Medicine. To Err is Human: Building a Safer Health System. Washington, DC: National Academy of Sciences; 1999. Google Scholar
11. Starfield B. Is US health really the best in the world? JAMA. 2000;284(4):483485. Crossref, MedlineGoogle Scholar
12. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. JAMA. 2004;291(10):12381245. Crossref, MedlineGoogle Scholar
13. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ. Selected major risk factors and global and regional burden of disease. Lancet. 2002;360(9343):13471360. Crossref, MedlineGoogle Scholar
14. Kalra L, Evans A, Perez I, Knapp M, Swift C, Donaldson N. A randomised controlled comparison of alternative strategies in stroke care. Health Technol Assess. 2005;9(18):iii–iv, 179. CrossrefGoogle Scholar
15. Neal B, MacMahon S, Chapman N. Effects of ACE inhibitors, calcium antagonists, and other blood-pressure-lowering drugs: results of prospectively designed overviews of randomised trials. Blood Pressure Lowering Treatment Trialists' Collaboration. Lancet. 2000;356(9246):19551964. Crossref, MedlineGoogle Scholar
16. Prospectives Studies Collaboration, Lewington S, Whitlock G, et al.. Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. Lancet. 2007;370(9602):18291839. Crossref, MedlineGoogle Scholar
17. Law MR, Frost CD, Wald NJ. By how much does dietary salt reduction lower blood pressure? III–Analysis of data from trials of salt reduction. BMJ. 1991;302(6780):819824. Crossref, MedlineGoogle Scholar
18. Jemal A, Thun MJ, Ries LA, et al.. Annual report to the nation on the status of cancer, 1975–2005, featuring trends in lung cancer, tobacco use, and tobacco control. J Natl Cancer Inst. 2008;100(23):16721694. Crossref, MedlineGoogle Scholar
19. Thun MJ, Jemal A. How much of the decrease in cancer death rates in the United States is attributable to reductions in tobacco smoking? Tob Control. 2006;15(5):345347. Crossref, MedlineGoogle Scholar
20. Georgeson M, Thorpe LE, Merlino M, Frieden TR, Fielding JE. Shortchanged? An assessment of chronic disease programming in major US city health departments. J Urban Health. 2005;82(2):183190. Crossref, MedlineGoogle Scholar
21. Glantz SA, Balbach ED. Tobacco war: inside the California battles. Berkeley: University of California Press; 2000. CrossrefGoogle Scholar


No related items




Thomas A. Farley, MD, MPHThomas A. Farley is with the Department of Community Health Sciences, School of Public Health and Tropical Medicine, Tulane University, New Orleans, Louisiana. “Reforming Health Care or Reforming Health?”, American Journal of Public Health 99, no. 4 (April 1, 2009): pp. 588-590.

PMID: 19197077