Publc Health Briefs
7. Horsburgh CR, Jarvis JQ, McArthur T, et al. Seroconversion to human immunodeficiencyvirus in prison inmates.AmJPublic Healt 1990;80:209-210. 8. Morse DL, Truman BI, HanrahanJP, et al. AIDS behind bars: epidemiology of New YorkState prison inmate cases, 1980-1988. NYState JMed 1990;90:133-138. 9. Burke DS, Brundage JF, Herbold JR, et al. Human immunodeficiency virus infections among cvilian applicants for United States military service, October 1985 to March 1986: demographic factors associated with seropositivity. N Engi J MeL 1987;317: 131-136. 10. Burke DS, Brundage JF, Goldenbaum M, et al. Human immunodeficiency virus infections in teenagers: seroprevalence among applicants for U.S. military service. JAMA 1990;263:2074-2077. 11. Brundage JF, Burke DS, Gardner LI, et al. Tracking the spread of the HIV infection epidemic amongyoung adults in the United States: results of the first four years of screening among civilian applicants for military service. JAIDS. 1990;3:1168-1180. 12. Garland FC, Douglas ML, Hickey TM, et al. Incidence of human immunodeficiency
.-----.... i ....
.....
..
virus seroconversion in US Navy and Marine Corps personnel, 1986 through 1988. JAMA 1989;262:3161-3165. 13. McNeil JG, Brundage JF, Wann ZF, et al. Direct measurement of human immunodeficiency virus seroconversion in a serially tested population of young adults in the United States Army, October 1985 to October 1987. NEnglJMed 1989;320:15811585. 14. Kramer MS. Clinical Epidemiology and Biostatistics. New York, NY: SpringerVerlag; 1988:32-35. 15. Lilienfeld AM, Lilienfeld DE. Foundations of epidemiology. 2nd ed. New York, NY: Oxford University Press; 1980:335-337. 16. Langmuir A. AIDS projections are too high. Pan Am Health Qrg BulL 1989;23: 121-126. 17. Bregman DJ, LangmuirAD. Farr's law applied to AIDS projections. JAMA. 1990;263:1522-1525. 18. McNeil JG, Brundage JF, Gardner LI, et al. Trends of HIV seroconversion among young adults in the US Army, 1985 to 1989. JAMA. 1991;265:1709-1714. 19. McGrane WL, Toth FJ. The use of inter-
active media for HIV/AIDS prevention in the military community. Milit Med. 1990;6:235-240. 20. Hahn RA, Onorato IM, Jones TS, et al. Prevalence of HIV infection among intravenous drug users in the United States. JAMA 1989;261:2677-2684. 21. Schoenbaum EE, Hartel D, Selwyn PA, et al. Risk factors for human immunodeficiency virus infection in intravenous drug users. NEngIJMedL 1989;321:874-879. 22. Chaisson RE, Moss AR, Onishi R, et al. Human immunodeficiency virus infection in heterosexual drug users in San Francisco. Am J Public Health. 1987;77:169172. 23. Herbold JR. AIDS policy development within the Department of Defense. Milit Med. 1986;151:623-627. 24. Hoffman KJ. 1989 Sir Henry Welicome medal and prize winner: mandatory HIV testing and the duty to screen. Milt Med. 1990;155:195-201. 25. Tramont EC, Redfield R, Burke D, et al. HTLV-EIILAV infections in the military. Milit Med. 1987;152:105-106. 26. MayerW. Whatwe are doing about AIDS. Defense. 1986;1:30-32.
Fear of Dying and HIV Infection vs Hepatitis B Infection Lawrence J. Schneidennan, MD, and Robert M. Kaplan, PhD
Intrducion Health care providers are at risk of infection resulting from accidental exposure to blood, such as needlestick injury. For many years before a vaccine became available, providers risked infection and death from hepatitis B. These risks were faced without the outcry and debate and the disruption of services that now attend the acquired immunodeficiency syndrome (AIDS) crisis. One of us, in a lecture to health care providers, pointed out this apparent irrationality of health providers' behavior in response to human immunodeficiencyvirus (HIV). As can be demonstrated by simple calculations, the probability ofdeath from accidental exposure to HIV is no worse than was the probability of death from accidental exposure to hepatitis B in the recent past. The chance of hepatitis B infection following accidental exposure is in the range of 25%; we conservatively assume that about 5% of those infected die either from fulminant or chronic progressive disease.1'2 The chance of HIV infection following acci-
dental exposure is much lower, apparently no greater than 1%, and it is presently assumed that 100% of patients who are HIV infected will develop AIDS and its complications and die.3 Multiplying the probabilities ofrisk of infection following exposure by the risk of death following infection yields approximately equal outcomes. That is, in both circumstances there is an approximately 1% chance of dying following accidental exposure. Why then are health providers, who took relatively few precautions in the hepatitis B era, now taking exceptional precautions, some of them even refusing The authors are with the Division of Health Care Sciences, Department of Community & Family Medicine, University of California, San Diego. Requests for reprints should be sent to Lawrence J. Schneiderman, MD, Division of Health Care Sciences, 0622, Department of Community & Family Medicine, University of California, San Diego, La Jolla, CA 920930622. This paper was submitted to the Journal March 25, 1991, and accepted with revisions July 9, 1991.
April 1992, Vol. 82, No. 4
Public Health Briefs
to care for patients who are IRV infected? Are they sharing in society's "AIDS hysteria," much of which is known to be fanned by prejudice and ignorance? A member of the audience provided a thought-provoldng answer: "The problem is that once you have HIV infection you have no chance. You know you're going to die." The answer suggested that even though the probability of infection is very low, the certainty of death might be so undesirable that it might account for the greater concern about HIV than about hepatitis B-a hypothesis we chose to test in an empirical study. Expected utility theory"5 has traditionally served to explain decision-making processes that involve risk. According to this theory, individuals should favor the alternative that maxinizes gains and mininizes losses. In the case of HIV and hepatitis B exposure, the expected utility of either alternative is nearly equal. Thus, it would be expected that the motivation to avoid each condition is approximately equal. An alternative to expected utility theory was introduced more recently. This theory, known as prospect theory,6 places greater emphasis on context effects. The theory suggests that a choice involving risk is strongly influenced by other contextual information. Unlike expected utility theory, prospect theory suggests that individuals do not take a comprehensive view of all possible outcomes that might be associated with a particular choice. Instead, they view the consequences of a particular choice as a change, which is judged as a gain or loss rather than the summation or final state of all possible outcomes. According to this theory, high probabilities of loss, such as certain death, might overcome low probabilities of the situation developing. Thus, prospect theory might suggest that fear of HIV should exceed fear of hepatitis B.
MetdW4o Subjects The subjects were 264 undergraduate students at the University of California, San Diego, and San Diego State University. The students were enrolled in lower division courses in psychology and health sciences. These courses had not, at the time of the study, considered the AIDS epidemic. Sixty-one percent of the subjects were female and 39%o were male. The mean age of the participants was 20.30 years (SD = 3.87 years).
April 1992, Vol. 82, No. 4
Questionnaire Each subject responded to one of three versions of a questionnaire. The first version descnibed a situation explicitly specifying hypothetical hepatitis B and HV exposures. Group 1 received a questionnaire that stated: Two disesewhich can be acquired by accidental needlestick exposure to an infected patient are hepatitis B and AIDS. The risks of contracting hepatitis B from a needlestick are approximately 25%, and if one gets the disease there is approximately a 5% chance of dying. The risks of contracting AIDS from a needlestick exposure to an infected patient are approximately 1%, but 100%o of those infected will die. Assume that a preventive vaccine is available for both diseases. You are a health care worker in a foreign country in which the prevalence of both diseases appears to be equal, but which can afford only one or the other vaccine for you. Which vaccine would you choose?
The subject could choose a hepatitis B or an AIDS vaccine. The second and third groups were presented scenarios with similar contrasts in risks as in the first group, but in descnbing the hypothetical situations the scenarios eliminated the wordsAIDS and hepatisRB. The first of these conditions stated: You have been shipwrecked on the shore of a tropical country. The only passage to safety is by one of two routes, both of which expose you to insects of varying population intensity and lethality. Route A sendsyou to avalleywhere your likelihood of being stung is 25%. The chance of dying if you are stung is 5%. Route B sends you through a valley in which the insects are sparser but more lethal.
Through this route you have a 1% chance of being stung, but if you are stung you are 1O00o certain to die. Which route would you choose?
The third set of scenarios (given to Group 3) was identical to the set given to the second group, except the probabilities of being stung and dying were reversed. Route A involved a 100%o chance of being stung but a 1% chance of dying, while Route B involved a 5% chance of being stung and a 25% chance of dying.
Statistical Methods Comparisons between conditions were achieved using x2 tests. T tests and X2 tests were used to evaluate the relationships between choice and demographic variables.
RemuS Table 1 summarzes the choices for the three groups. Each choice was conceptualized as either avoiding HIV or avoiding hepatitis B risk situations. Because of the wording of the scenarios, this risk component was systematically varied across groups. In the insect scenario, 64% chose the route that avoided the 1% chance exposure and 100% death; in the explicit HIV-hepatitis B scenario, 67% preferred an AIDS over a hepatitis B vaccine. The difference between these two groups was nonsignificant (X2 = 0.26). When the probabilities in the insect scenarios were reversed (100% chance of being infected, 1% chance of dying; 5% chance of being infected, 25% chance of dying), the results changed. Under these circumstances, only 31% preferred the route with the 100%o chance of being infected and 1% chance of dying. The difference between these two insect scenar-
American Journal of Public Health 585
Public Health Briefs
ios (i.e., reversing the chances of being stung and dying) was statistically significant(X2 = 19.31,P < .001). The subjects' response to the insect scenario with the reversed probabilities was also significandy different from the response to the explicit HIV-hepatitis B scenario (X2 = 22.85, P < .001). Age and sex did not significantly influence choice, as evaluated respectively by t tests (age) and x2 tests (sex). For example, in Group 2, 69% of the women and 66% of the men selected Choice 1. The mean age of those selecting Choice 1 was 21.61 years, while the mean age of those selecting Choice 2 was 20.34 years
(t8, = 1.33, NS). Discussion Expected utility theory suggests that decisions should be based on the product of outcome and expectancy.4'5 In this experiment we created scenarios in which the expected utilities of two outcomes were nearly equivalent. In all scenarios the subjects avoided outcomes that had a low probability ofcertain death. In the one condition that did not include certain death, subjects preferred the choice that offered the lowest probability of death. The effect of context observed in this study lends support to prospect theory.6 In comparison to expected utility theory, prospect theory is better able to account for the effects of certain outcomes upon judgment. Since responses to the insect scenarios and the identified disease scenario were not significantly different, data from this study may explain the differential attention to hepatitis B and HIV by health care providers. If fear of AIDS caused alarm independent of the probability information, we should have observed differential responses to the insect and identified disease scenarios with the same probabilities. Although the expected value of death due to exposure to HIV is about equivalent to the expected value of death from hepatitis B, the fear of HIV appears to exceed the fear of hepatitis B.
586 American Journal of Public Health
Perhaps the most important explanatory variable is the probability of certain death. The chances of being infected with HIV are small, yet the consequences are very serious. Fear of living for any length of time infected with a virus for which the outcomes are always fatal apparently overrides the concern that the initial chances of being infected are small. Conversely, accidental exposure to hepatitis B, although highly infectious, offers the hope that even if infected the victim has only about a 5% chance of ultimately dying of the illness. Thus, what at first appears to be irrational behavior in the face of alternative risks might actually be rational. Who would not prefer to live, no matter how long or short a time, in a state of hopeful uncertainty when it comes to death (as most of us do every day) than to be forced to alter our perception and live in a state of dismal certainty? The results might also explain fears associated with other low-probability events that are perceived to have certain mortal outcomes. For example, the fear of being involved in an airplane crash appears to greatly exceed the fear of becoming involved in an auto crash, even though the probability of injury and death from auto crashes exceeds that from airplane crashes, given that many more auto crashes occur. We hope these results also offer insights into public attitudes that might lead to practical approaches to education. For one thing, it seems apparent that the customary use of statistics in allaying anxieties and enforcing precautions must be improved by incorporating the "rationally irrational" fears experienced by most, if not all, human beings. This study has a number of limitations. The scenarios were hypothetical, and participation was restricted to college students. It might be argued that fears of accidental exposure to HIV and hepatitis B are related to factors other than those that we studied. For example, there is the issue of medical secrecy. Although information about hepatitis B is typically listed in the patient record, information about the HIV status of patients is typically ab-
sent. Thus, providers can take precautions against hepatitis B but do not have the information to alert them to take precautions against HIV. We cannot discount this explanation; however, we point out that in this study the greater fears of HIV exposure over hepatitis B exposure were demonstrated by subjects who were not health care providers and were not familiar with patient records. Furthermore, the island scenario produced results almost identical to those of the HIV-hepatitis B scenario and did not mention the names of the diseases. Another limitation ofthe study is that these scenarios are not entirely realistic. For example, the AIDS vs hepatitis B scenario asked which vaccine the respondent would prefer. Unfortunately, there is no real AIDS vaccine, making the generalization to the real-life situation difficult. However, we suggest that the apparently excessive fear of HIV exposure is not an irrational manifestation, but rather the rational reaction to the prospect of living with the certainty of proximate death. l
Ackmowledgments The authors thank Cathie Atkins, Bea Roppe, Liz Eakil, John Polich, and Amy Good for assistance in acquiring the data.
References 1. Centers for Disease Control. Guidelines for prevention of tansmission of human immunodeficiency virus and hepatitis B virus to health-care and public-safety workers.
MMWR 1989;38(suppl 6):5-6. 2. Centers for Disease Control. Recommendations for protection against viral hepatitis. MMWR. 1985;34:313-324, 329-335. 3. Marcus R. CDC cooperative needlestick group: Surveillance of health care workers exposed to blood from patients infected with the human immunodeficiency virus. NEngi JMed 1988;319:1118-1123. 4. Raiffa H, Schlaifer R. Applied Statistical Decision 7heory. Cambridge, Mass: Harvard Business School; 1961. 5. Von Neumann J, Morgenstern 0. Theor of Games and Economic Behavior. Princeton, NJ: Princeton University Press; 1944. 6. Kahneman D, Tversky A. Choices, values and frames. Am PsychoL 1983;39:341-350.
April 1992, Vol. 82, No. 4
7. Horsburgh CR, Jarvis JQ, McArthur T, et al. Seroconversion to human immunodeficiencyvirus in prison inmates.AmJPublic Healt 1990;80:209-210. 8. Morse DL, Truman BI, HanrahanJP, et al. AIDS behind bars: epidemiology of New YorkState prison inmate cases, 1980-1988. NYState JMed 1990;90:133-138. 9. Burke DS, Brundage JF, Herbold JR, et al. Human immunodeficiency virus infections among cvilian applicants for United States military service, October 1985 to March 1986: demographic factors associated with seropositivity. N Engi J MeL 1987;317: 131-136. 10. Burke DS, Brundage JF, Goldenbaum M, et al. Human immunodeficiency virus infections in teenagers: seroprevalence among applicants for U.S. military service. JAMA 1990;263:2074-2077. 11. Brundage JF, Burke DS, Gardner LI, et al. Tracking the spread of the HIV infection epidemic amongyoung adults in the United States: results of the first four years of screening among civilian applicants for military service. JAIDS. 1990;3:1168-1180. 12. Garland FC, Douglas ML, Hickey TM, et al. Incidence of human immunodeficiency
.-----.... i ....
.....
..
virus seroconversion in US Navy and Marine Corps personnel, 1986 through 1988. JAMA 1989;262:3161-3165. 13. McNeil JG, Brundage JF, Wann ZF, et al. Direct measurement of human immunodeficiency virus seroconversion in a serially tested population of young adults in the United States Army, October 1985 to October 1987. NEnglJMed 1989;320:15811585. 14. Kramer MS. Clinical Epidemiology and Biostatistics. New York, NY: SpringerVerlag; 1988:32-35. 15. Lilienfeld AM, Lilienfeld DE. Foundations of epidemiology. 2nd ed. New York, NY: Oxford University Press; 1980:335-337. 16. Langmuir A. AIDS projections are too high. Pan Am Health Qrg BulL 1989;23: 121-126. 17. Bregman DJ, LangmuirAD. Farr's law applied to AIDS projections. JAMA. 1990;263:1522-1525. 18. McNeil JG, Brundage JF, Gardner LI, et al. Trends of HIV seroconversion among young adults in the US Army, 1985 to 1989. JAMA. 1991;265:1709-1714. 19. McGrane WL, Toth FJ. The use of inter-
active media for HIV/AIDS prevention in the military community. Milit Med. 1990;6:235-240. 20. Hahn RA, Onorato IM, Jones TS, et al. Prevalence of HIV infection among intravenous drug users in the United States. JAMA 1989;261:2677-2684. 21. Schoenbaum EE, Hartel D, Selwyn PA, et al. Risk factors for human immunodeficiency virus infection in intravenous drug users. NEngIJMedL 1989;321:874-879. 22. Chaisson RE, Moss AR, Onishi R, et al. Human immunodeficiency virus infection in heterosexual drug users in San Francisco. Am J Public Health. 1987;77:169172. 23. Herbold JR. AIDS policy development within the Department of Defense. Milit Med. 1986;151:623-627. 24. Hoffman KJ. 1989 Sir Henry Welicome medal and prize winner: mandatory HIV testing and the duty to screen. Milt Med. 1990;155:195-201. 25. Tramont EC, Redfield R, Burke D, et al. HTLV-EIILAV infections in the military. Milit Med. 1987;152:105-106. 26. MayerW. Whatwe are doing about AIDS. Defense. 1986;1:30-32.
Fear of Dying and HIV Infection vs Hepatitis B Infection Lawrence J. Schneidennan, MD, and Robert M. Kaplan, PhD
Intrducion Health care providers are at risk of infection resulting from accidental exposure to blood, such as needlestick injury. For many years before a vaccine became available, providers risked infection and death from hepatitis B. These risks were faced without the outcry and debate and the disruption of services that now attend the acquired immunodeficiency syndrome (AIDS) crisis. One of us, in a lecture to health care providers, pointed out this apparent irrationality of health providers' behavior in response to human immunodeficiencyvirus (HIV). As can be demonstrated by simple calculations, the probability ofdeath from accidental exposure to HIV is no worse than was the probability of death from accidental exposure to hepatitis B in the recent past. The chance of hepatitis B infection following accidental exposure is in the range of 25%; we conservatively assume that about 5% of those infected die either from fulminant or chronic progressive disease.1'2 The chance of HIV infection following acci-
dental exposure is much lower, apparently no greater than 1%, and it is presently assumed that 100% of patients who are HIV infected will develop AIDS and its complications and die.3 Multiplying the probabilities ofrisk of infection following exposure by the risk of death following infection yields approximately equal outcomes. That is, in both circumstances there is an approximately 1% chance of dying following accidental exposure. Why then are health providers, who took relatively few precautions in the hepatitis B era, now taking exceptional precautions, some of them even refusing The authors are with the Division of Health Care Sciences, Department of Community & Family Medicine, University of California, San Diego. Requests for reprints should be sent to Lawrence J. Schneiderman, MD, Division of Health Care Sciences, 0622, Department of Community & Family Medicine, University of California, San Diego, La Jolla, CA 920930622. This paper was submitted to the Journal March 25, 1991, and accepted with revisions July 9, 1991.
April 1992, Vol. 82, No. 4
Public Health Briefs
to care for patients who are IRV infected? Are they sharing in society's "AIDS hysteria," much of which is known to be fanned by prejudice and ignorance? A member of the audience provided a thought-provoldng answer: "The problem is that once you have HIV infection you have no chance. You know you're going to die." The answer suggested that even though the probability of infection is very low, the certainty of death might be so undesirable that it might account for the greater concern about HIV than about hepatitis B-a hypothesis we chose to test in an empirical study. Expected utility theory"5 has traditionally served to explain decision-making processes that involve risk. According to this theory, individuals should favor the alternative that maxinizes gains and mininizes losses. In the case of HIV and hepatitis B exposure, the expected utility of either alternative is nearly equal. Thus, it would be expected that the motivation to avoid each condition is approximately equal. An alternative to expected utility theory was introduced more recently. This theory, known as prospect theory,6 places greater emphasis on context effects. The theory suggests that a choice involving risk is strongly influenced by other contextual information. Unlike expected utility theory, prospect theory suggests that individuals do not take a comprehensive view of all possible outcomes that might be associated with a particular choice. Instead, they view the consequences of a particular choice as a change, which is judged as a gain or loss rather than the summation or final state of all possible outcomes. According to this theory, high probabilities of loss, such as certain death, might overcome low probabilities of the situation developing. Thus, prospect theory might suggest that fear of HIV should exceed fear of hepatitis B.
MetdW4o Subjects The subjects were 264 undergraduate students at the University of California, San Diego, and San Diego State University. The students were enrolled in lower division courses in psychology and health sciences. These courses had not, at the time of the study, considered the AIDS epidemic. Sixty-one percent of the subjects were female and 39%o were male. The mean age of the participants was 20.30 years (SD = 3.87 years).
April 1992, Vol. 82, No. 4
Questionnaire Each subject responded to one of three versions of a questionnaire. The first version descnibed a situation explicitly specifying hypothetical hepatitis B and HV exposures. Group 1 received a questionnaire that stated: Two disesewhich can be acquired by accidental needlestick exposure to an infected patient are hepatitis B and AIDS. The risks of contracting hepatitis B from a needlestick are approximately 25%, and if one gets the disease there is approximately a 5% chance of dying. The risks of contracting AIDS from a needlestick exposure to an infected patient are approximately 1%, but 100%o of those infected will die. Assume that a preventive vaccine is available for both diseases. You are a health care worker in a foreign country in which the prevalence of both diseases appears to be equal, but which can afford only one or the other vaccine for you. Which vaccine would you choose?
The subject could choose a hepatitis B or an AIDS vaccine. The second and third groups were presented scenarios with similar contrasts in risks as in the first group, but in descnbing the hypothetical situations the scenarios eliminated the wordsAIDS and hepatisRB. The first of these conditions stated: You have been shipwrecked on the shore of a tropical country. The only passage to safety is by one of two routes, both of which expose you to insects of varying population intensity and lethality. Route A sendsyou to avalleywhere your likelihood of being stung is 25%. The chance of dying if you are stung is 5%. Route B sends you through a valley in which the insects are sparser but more lethal.
Through this route you have a 1% chance of being stung, but if you are stung you are 1O00o certain to die. Which route would you choose?
The third set of scenarios (given to Group 3) was identical to the set given to the second group, except the probabilities of being stung and dying were reversed. Route A involved a 100%o chance of being stung but a 1% chance of dying, while Route B involved a 5% chance of being stung and a 25% chance of dying.
Statistical Methods Comparisons between conditions were achieved using x2 tests. T tests and X2 tests were used to evaluate the relationships between choice and demographic variables.
RemuS Table 1 summarzes the choices for the three groups. Each choice was conceptualized as either avoiding HIV or avoiding hepatitis B risk situations. Because of the wording of the scenarios, this risk component was systematically varied across groups. In the insect scenario, 64% chose the route that avoided the 1% chance exposure and 100% death; in the explicit HIV-hepatitis B scenario, 67% preferred an AIDS over a hepatitis B vaccine. The difference between these two groups was nonsignificant (X2 = 0.26). When the probabilities in the insect scenarios were reversed (100% chance of being infected, 1% chance of dying; 5% chance of being infected, 25% chance of dying), the results changed. Under these circumstances, only 31% preferred the route with the 100%o chance of being infected and 1% chance of dying. The difference between these two insect scenar-
American Journal of Public Health 585
Public Health Briefs
ios (i.e., reversing the chances of being stung and dying) was statistically significant(X2 = 19.31,P < .001). The subjects' response to the insect scenario with the reversed probabilities was also significandy different from the response to the explicit HIV-hepatitis B scenario (X2 = 22.85, P < .001). Age and sex did not significantly influence choice, as evaluated respectively by t tests (age) and x2 tests (sex). For example, in Group 2, 69% of the women and 66% of the men selected Choice 1. The mean age of those selecting Choice 1 was 21.61 years, while the mean age of those selecting Choice 2 was 20.34 years
(t8, = 1.33, NS). Discussion Expected utility theory suggests that decisions should be based on the product of outcome and expectancy.4'5 In this experiment we created scenarios in which the expected utilities of two outcomes were nearly equivalent. In all scenarios the subjects avoided outcomes that had a low probability ofcertain death. In the one condition that did not include certain death, subjects preferred the choice that offered the lowest probability of death. The effect of context observed in this study lends support to prospect theory.6 In comparison to expected utility theory, prospect theory is better able to account for the effects of certain outcomes upon judgment. Since responses to the insect scenarios and the identified disease scenario were not significantly different, data from this study may explain the differential attention to hepatitis B and HIV by health care providers. If fear of AIDS caused alarm independent of the probability information, we should have observed differential responses to the insect and identified disease scenarios with the same probabilities. Although the expected value of death due to exposure to HIV is about equivalent to the expected value of death from hepatitis B, the fear of HIV appears to exceed the fear of hepatitis B.
586 American Journal of Public Health
Perhaps the most important explanatory variable is the probability of certain death. The chances of being infected with HIV are small, yet the consequences are very serious. Fear of living for any length of time infected with a virus for which the outcomes are always fatal apparently overrides the concern that the initial chances of being infected are small. Conversely, accidental exposure to hepatitis B, although highly infectious, offers the hope that even if infected the victim has only about a 5% chance of ultimately dying of the illness. Thus, what at first appears to be irrational behavior in the face of alternative risks might actually be rational. Who would not prefer to live, no matter how long or short a time, in a state of hopeful uncertainty when it comes to death (as most of us do every day) than to be forced to alter our perception and live in a state of dismal certainty? The results might also explain fears associated with other low-probability events that are perceived to have certain mortal outcomes. For example, the fear of being involved in an airplane crash appears to greatly exceed the fear of becoming involved in an auto crash, even though the probability of injury and death from auto crashes exceeds that from airplane crashes, given that many more auto crashes occur. We hope these results also offer insights into public attitudes that might lead to practical approaches to education. For one thing, it seems apparent that the customary use of statistics in allaying anxieties and enforcing precautions must be improved by incorporating the "rationally irrational" fears experienced by most, if not all, human beings. This study has a number of limitations. The scenarios were hypothetical, and participation was restricted to college students. It might be argued that fears of accidental exposure to HIV and hepatitis B are related to factors other than those that we studied. For example, there is the issue of medical secrecy. Although information about hepatitis B is typically listed in the patient record, information about the HIV status of patients is typically ab-
sent. Thus, providers can take precautions against hepatitis B but do not have the information to alert them to take precautions against HIV. We cannot discount this explanation; however, we point out that in this study the greater fears of HIV exposure over hepatitis B exposure were demonstrated by subjects who were not health care providers and were not familiar with patient records. Furthermore, the island scenario produced results almost identical to those of the HIV-hepatitis B scenario and did not mention the names of the diseases. Another limitation ofthe study is that these scenarios are not entirely realistic. For example, the AIDS vs hepatitis B scenario asked which vaccine the respondent would prefer. Unfortunately, there is no real AIDS vaccine, making the generalization to the real-life situation difficult. However, we suggest that the apparently excessive fear of HIV exposure is not an irrational manifestation, but rather the rational reaction to the prospect of living with the certainty of proximate death. l
Ackmowledgments The authors thank Cathie Atkins, Bea Roppe, Liz Eakil, John Polich, and Amy Good for assistance in acquiring the data.
References 1. Centers for Disease Control. Guidelines for prevention of tansmission of human immunodeficiency virus and hepatitis B virus to health-care and public-safety workers.
MMWR 1989;38(suppl 6):5-6. 2. Centers for Disease Control. Recommendations for protection against viral hepatitis. MMWR. 1985;34:313-324, 329-335. 3. Marcus R. CDC cooperative needlestick group: Surveillance of health care workers exposed to blood from patients infected with the human immunodeficiency virus. NEngi JMed 1988;319:1118-1123. 4. Raiffa H, Schlaifer R. Applied Statistical Decision 7heory. Cambridge, Mass: Harvard Business School; 1961. 5. Von Neumann J, Morgenstern 0. Theor of Games and Economic Behavior. Princeton, NJ: Princeton University Press; 1944. 6. Kahneman D, Tversky A. Choices, values and frames. Am PsychoL 1983;39:341-350.
April 1992, Vol. 82, No. 4
