Keeney, Ralph L.
As a society and as individuals, we Americans are preoccupied with risks, particularly risks to life. From AIDS to cancer to heart disease, from Alar to asbestos to benzene, from eating to drinking to smoking -- we worry about all the risks of living. We allocate substantial time, effort, and money to reducing risks, yet most of us believe that our world is riskier now than it was a generation ago. This simply is not so. Life expectancy in the United States has increased despite the scourges of cancer, AIDS, and violent crime. Still, major national and personal agendas to deal with life-threatening risks are taking on a new urgency. We cannot banish these risks, but we can and should learn better ways to deal with them. Dealing better with risks means worrying less and thinking more so that we make better-informed decisions about them.
With any class of problems, certain realities must be accepted as premises. With problems involving risk, those realities are what might be called facts of life rather than facts of science. For example, it is a fact of life that you cannot reduce all risks to zero. Laws and regulations that require the complete elimination of risk are simply not realistic. Any claim that some course of action reduces risk to zero is inaccurate and misleading, resulting in lack of respect for and trust in the communicator.
This essay presents seven realities to guide constructive thinking about medical and health care decisions concerning life-threatening risks. The ideas are relevant for thinking about any problem involving risk, whether the alternatives are appraised intuitively or by systematic analysis and whether they pertain to public or personal risks. The intent is to help people better understand, appraise, and communicate about specific problems involving risk.
There is no activity, process, or product that is free of risk. Going to the hospital or to work entails the risk of a mugging or an accident. Using electricity, drugs, food, and vehicles makes one susceptible to the risks associated with malfunction or misuse. And natural catastrophes, such as earthquakes, tornadoes, lightning, and floods, can strike anyone. Risks are clearly a part of life. Indeed, it is living that poses risks.
The logic is simple when viewed from an individual perspective. Everyone must eventually die. Hence, if one type of risk is reduced, other risks increase, although the timing and cause of one's eventual death are likely to change. For instance, seat belts reduce the risk of dying in an automobile accident. As a result, for those who wear seat belts, the risk of dying from cancer or heart disease increases.
This fact in no way suggests that an individual or society should not manage life-threatening risks. Most of us would probably rather die of so-called natural causes when we are 90 years old than in an automobile accident a year from now. The point is that when we speak of saving a life, it is more accurate to say that we are prolonging a life. When we speak of reducing risks, we mean reducing risks from particular causes, which in turn increases risks from other causes in the future.
For society, too, there are no risk-free alternatives. The decision to eliminate an option may sometimes create the illusion of zero risk. For example, consider a new drug that has been developed to cure a disease that takes many lives annually. Suppose clinical trials indicate that the drug reduces the risks associated with the disease but carries other risks due to detrimental side effects. Some people tend to think that not licensing the drug would result in zero risk. They reason that a risk-free drug to cure the same disease will soon be developed. Until then, however, many people with the disease will die, so the risk is not zero. Also, it may not prove possible to develop a similarly effective drug without side effects. Even if a drug with minimal side effects were developed, resources would have been spent developing this drug rather than others. Who might have been saved by those other drugs? The elimination of one alternative means the selection of another, and any selected alternative will always have risks.
A course of action taken to reduce a particular risk indirectly entails other risks. Surgery for heart disease, for example, may decrease the future risk of a heart attack, but surgery itself carries risks. Blood transfusions to fight one disease pose risks from other sources. Inoculations to prevent life-threatening diseases such as diphtheria or swine influenza can themselves result in loss of life.
Regulatory policies designed to reduce specific risks to the public also indirectly create other risks. For instance, whatever national health care policies are implemented will potentially increase many risks. Certain drugs or medical procedures may be discouraged. Resources for developing new drugs or techniques may be reduced. The relationship between doctors and patients will change, which can affect the care patients receive and their attitudes. Some medical professionals may be discouraged and give up the practice of medicine, and others may be discouraged from pursuing careers in medicine. And higher economic costs take a toll in lives through numerous pathways, as discussed below.
Public problems often involve a transfer of risks from one group of people to another. If the proposed ban on saccharin had been implemented in the late 1970s, millions of people might have had a slightly reduced risk of cancer, but those with diabetes would have had an increased risk of problems associated with that disease. The transfer of a risk from one group to another is often subtle. Using more resources for the treatment of patients with cancer may mean using fewer resources for the treatment of other patients -- say, those with heart disease. Consequently, the risk for the latter group may increase. Even with the best of intentions, risks may be inadvertently transferred. Cutting curbs at crosswalks makes it safer for people in wheelchairs to cross streets but may make it more dangerous for blind people.
We all know that people who live in wealthier countries tend to live longer than those who live in poorer countries. What is less well known is that the wealthier residents of a country live longer, on average, than the poorer residents [1,2]. Wealthier nations have the economic resources to take preventive actions to reduce risks as well as ameliorative actions to address medical and national disasters. Wealthier nations have the infrastructure, knowledge, and funds to reduce risks substantially.
On an individual level, too, richer is safer, [3,4] and in fact the economic costs of risk-reduction policies can induce substantial risks. All the money required to implement risk-reduction policies, including those paid for by government, must come from individuals, leaving them with less money to spend on all other needs, including personal activities to reduce risk. There is strong evidence that a higher income in the long term leads to lower individual risks . A quantification of this principle suggests that a cost in the range of $5 million to $12 million (in 1990 dollars) that is borne collectively by many people may induce one death, because those people are slightly poorer and therefore cannot make many life-sustaining choices,  such as obtaining better health care services, buying better tires, eating more healthful foods, working fewer hours, living in safer communities, and in general leading less stressful lives.
Suppose a rare disease is diagnosed in one teenager, Janet. Without treatment, she is expected to survive for only three months. A clinical appraisal indicates that, at this stage of the disease, it would cost $5 million to try to save Janet, an effort that has only a 50 percent chance of success. The decision is made to proceed with treatment. Most people would concur that it is worth the $5 million for a 50 percent chance of saving Janet's life, since $5 million is certainly of less value than Janet's life. Just before treatment begins, a researcher familiar with the disease voices a dissenting opinion: ``This disease results in the deaths of 12 young people annually. An educational program, with even a modest budget of $5 million, could reduce the occurrence of this disease. The statistics suggest that four lives would be saved by this program. Should we spend $5 million for a 50 percent chance to save one teenager's life when we could use the same amount to save the lives of four teenagers''?
Perhaps $5 million should be spent for each purpose, but if only one of them can be pursued, many people will choose to save Janet's life. Saving Janet's life would avert an identifiable death, since everyone would know who was saved. Saving four teenagers who would otherwise have had the disease would avert four statistical deaths. In this case, it would never be known whose lives were saved. Because of this distinction, many people believe it is appropriate to assign a higher economic value to averting an identifiable death than to averting a statistical death. There is, of course, no right or wrong answer to this problem.
Most problems of public risk concern statistical deaths, whereas problems of personal risk concern identifiable deaths. It is also interesting to note that the deaths averted by programs focused on cure are often identifiable, whereas those averted by prevention programs are more often statistical. Perhaps this partially explains the tendency to emphasize cure rather than prevention in considering various risks.
Values are essential for evaluating risks. To some people, this is obvious. How could one evaluate anything without values? But others apparently would like to believe that it is not necessary to account for values in analyzing risks. They wish to rely only on facts in evaluating alternative courses of action and making choices. Yet even when the facts are abundant and unambiguous, which is rarely the case, someone with a choice to make wishes to choose the best (or a good) alternative. The notions of ``best'' and ``good'' are based on values.
Personal medical decisions, especially for serious problems, indicate the essential nature of values. Suppose you are considering whether or not to undergo a sigmoidoscopy to detect possible polyps in the lower intestine. The current data suggest that there is 1 chance in 20,000 that you will die in the next five years from colon cancer that may be recognized by sigmoidoscopy. This test, and the resulting procedures to eliminate any polyps, can reduce your five-year risk of death to 1 chance in 100,000. Suppose the sigmoidoscopy costs $600. Is it worth it to you? You have all the data. But to make this decision, you need to think about whether you value a reduction in risk from 5 chances to 1 in 100,000 over five years more than you value the $600. You can avoid thinking hard about these value judgments or perhaps thinking about them at all, but you cannot avoid making them. To benefit from decision making based on a review and appraisal of such value judgments, they should be made systematically, consistently, and explicitly .
The relative value of individual lives is an essential component in all problems of public risk. Over the past several years, I have asked many people whether they think that the evaluation of problems of public risk should include value judgments about the relative importance of different people's lives. The response is almost always no. In a follow-up question, I ask whether the lives of all citizens should be valued equally, and the answer is invariably yes. I then point out that this preference for equality is a value judgment.
The preference for equality suggests that the death of any one person is as important to avert as the death of any other person. Thus, the death of a speeding drunk driver is as important to avert as that of a passenger in a car hit by the drunk driver. Also, saving the life of a 90-year-old, mentally impaired person in pain is just as important as saving the life of a healthy 10-year-old child. But the 10-year-old stands to lose an expected 70 years of healthful life, whereas the 90-year-old stands to lose only about 2 years. Should the preference for equality mean that we count the loss of each expected year of life the same, so that the death of the 10-year-old is 35 times as important as the death of the 90-year-old ? One way or another, value judgments about the relative importance of individual lives are inescapable in addressing problems of public risk.
In essentially every context of public decision making, spending more money could reduce the risk of loss of life. More safety features could be installed on automobiles, more testing could be done before licensing drugs, stricter controls could be required for sources of pollution. All these steps would presumably reduce the risks involved, but they would also increase costs. Hence, a critical question is how we should use economic resources most effectively to reduce risk. The essential trade-off is between different approaches to reducing risk.
In many situations, a specific decision about whether to spend funds to reduce a certain risk must be made independently of the consideration of other potential uses for the funds. In these circumstances, it is necessary to make a trade-off between the objective of controlling economic costs and the objective of preventing loss of life -- that is, to determine how much we are willing to spend to avert a statistical death. By using this trade-off between economic costs and statistical deaths in analyzing various problems, we can make decisions in an effective, consistent manner.
In making personal decisions, people must make trade-offs involving their own risks and their own funds. You can spend money in many ways to enhance your safety and health. You can buy an automobile with special safety features, join a health club, obtain preventive health care services, and install an alarm system in your home. But because your funds are limited, purchases made to lower one risk may preclude purchases made to lower other risks. Knowing the economic costs of reducing various risks may provide some guidance for making the trade-off between costs and risks.
There is a subtle interaction between the appropriate value trade-offs for problems of public risk and problems of personal risk. With problems of public risk, all the economic costs of purchasing additional health and safety measures are eventually paid for by individual people through a multitude of mechanisms, such as taxes and higher product costs. These people then have fewer discretionary funds to promote their personal health and safety. In reviewing a regulation of toxic substances proposed by the Occupational Safety and Health Administration, the Office of Management and Budget used this logic to suggest that the proposal might claim more lives than it would save . Consequently, the office suspended its review until the risks associated with economic costs could be addressed along with the risks associated with exposure to toxic substances. The question is whether the public should pay for very expensive risk-reduction programs that would avert a few statistical deaths but result in considerably more deaths associated with reduced income.
Some people consider it immoral to make a trade-off between lives and economic costs. However, as stated above, many problems of risk require such a trade-off. Moral theories clearly hold that an action is not immoral when there are no alternatives,  so making the trade-off in values between loss of lives and economic costs is not immoral.
In addition, many moral theories argue that refraining from actions that save lives is immoral. To the extent that an analysis can lead to better decisions that in turn will lead to the effective use of economic resources to save more lives, it is perhaps immoral not to address explicitly the crucial trade-off between costs and statistical deaths . The important questions are whether the trade-off is made explicitly or implicitly and what the trade-off should be.
Ralph L. Keeney
University of Southern California
Los Angeles, CA 90089-0021
Supported by grants from the National Science Foundation (SBR-9308660) and the Electric Power Research Institute (RP-2560-03).
1. Kitagawa EM, Hauser PM. Differential mortality in the United States: a study in socioeconomic epidemiology. Cambridge, Mass.: Harvard University Press, 1973. [Context Link]
2. Pappas G, Queen S, Hadden W, Fisher G. The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med 1993;329:103-9. Ovid Full Text Bibliographic Links [Context Link]
3. Wildavsky A. Richer is safer. Public Interest 1980;60:23-39. [Context Link]
4. Wildavsky A. Searching for safety. New Brunswick, N.J.: Transaction, 1988. [Context Link]
8. Graham JD, Raiffa H, Vaupel JW. Science and analysis: roles in risk and decision making. In: Covello VT, Menkes J, Mumpower J, eds. Risk evaluation and management. New York: Plenum Press, 1986:503-18. [Context Link]
9. MacRae JB Jr. Letter from Acting Administrator MacRae, Office of Information and Regulatory Affairs, Office of Management and Budget, to Honorable Nancy Risque-Rohrbach, Assistant Secretary for Policy, U.S. Department of Labor, March 10, 1992. [Context Link]
10. Beauchamp TL. Philosophical ethics. New York: McGraw-Hill, 1982. [Context Link]
11. Keeney RL. Ethics, decision analysis, and public risks. Risk Anal 1984;4:117-29. [Context Link]Accession Number: 00006024-199407210-00010