Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
Summary The human immunocleficiency virus (HIV), now known to be the cause of acquired immune deficiency syndrome, or AIDS, is only one element of the complex problem that is commonly called the AIDS epidemic. The spread] of HIV infection and, consequently, AIDS is the product of human behaviors enacted in social contexts. Both the behaviors and the circumstances in which they occur are conditioned and shaped by culture and larger social structures. The epidemic is thus as much a social and behavioral phenomenon as it is a biological one. Understanding how HIV infection is spread, encouraging be- havioral change so as to retard this spread, and coping with the social consequences of the epidemic raise questions that lie within the domain of the social, behavioral, and statistical sciences. Fol- Towing publication of the 1986 report on AIDS of the Institute of Medicine/National Academy of Sciences, the present committee was established in the fall of 1987 to provide a focus for AIDS activities within these disciplines at the National Research Council. At the re- quest of the Public Health Service (PHS) and with additional support from the Rockefeller and Russell Sage Foundations, the committee 1 this committee's review of the behavioral, social, and statistical issues related to HIV/AIDS builds on the work of the Institute of Medicine/National Academy of Sci- ences, which has produced two reports that focused on public health, biological research, and medical care issues: Confronting AIDS: Directions for Public Health, Health Care, and Research (1986) and Confronting AIDS: Update 1988 (both published by the Na- tional Academy Press, Washington, D.C.~. The committee also wishes to acknowledge the related activities being carried out or planned by the Academy complex. Some of the efforts under consideration including the future activities of our own committee and of the Institute of Medicine address topics that will be of concern to readers of this report, including drug and vaccine development, AIDS research policy, and the social impact of the AIDS epidemic. 1

OCR for page 1
2 | AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE has begun its work by reviewing the contributions that can be made by the paradigms, data, and methods of the social, behavioral, ant! statistical sciences2 in mounting an effective national response to the HIV/AIDS epidemic.3 The committee's report is diviclec3 into three parts. The first part presents evidence on the current extent of HIV infection in the U.S. population (Chapter 1) and on the patterns of sexual behavior and drug use (Chapters 2 and 3) that spread HIV infection. The second part describes intervention strategies and principles that hold promise for producing behavioral change to slow the spread of HTV infection (Chapter 4) and methods for evaluating the effectiveness of such interventions (Chapter 5~. The third part -(Chapters 6 and 7) discusses some of the barriers that impede effective research and intervention programs. The organization of this summary follows that of the report, and it includes some of the report's key recom- mendations. (All of the committee's recommendations are listed in Appendix A.) At the outset of its report, the committee believes it is important to comment on the term epidemic, which is sometimes misunderstood in connection with HIV/AIDS. During an epidemic, the occurrences of new cases of a disease in a community follows a well-known pattern: it may increase dramatically in a short period of time, peak, and then clecTine. During the course of an epidemic, there may be cycles of rise anti decline in the number of new cases. In 1989 the Unite(1 States stands at the base of a rapidly rising curve of AIDS cases and (leaths. Barring a dramatic breakthrough in treatment, it is projected that more than 50,000 Americans will die of AIDS during 1991. The number of deaths (luring this 12-month 2Including anthropology, economics, political science, psychology, sociology, and statis- tics, and their subdisciplines (e.g., demography, social psychology, biostatistics, etc.). 3Specifically, the committee was charged to (1) describe what is known about the spread of HIV infection and AIDS in the United States; (2) identify critical groups at risk of infection and how to reach them; (3) describe research findings from the social and behavioral sciences that should be helpful in planning and choosing among ways to in- tervene successfully to control the spread of HIV infection; (4) describe ways to evaluate the effectiveness of such interventions; and (5) recommend new research that can expand our understanding of the spread of HIV infection and improve the nation's ability to control this spread in the future. 4Two technical terms are frequently used in discussions of epidemic diseases: incidence and prevalence. Incidence denotes the rate of occurrence of new infections per unit of time (e.g., per year). Thus, an incidence of .03 per year in some group means that new infections occurred in 3 percent of the group during the year in question. Prevalence denotes that proportion of a group that is currently infected. A prevalence of .10 means that 10 percent of the group is currently infected.

OCR for page 1
SUMMARY ~ 3 period alone will exceed the total number of deaths in this country from the beginning of the epidemic through 1988. Such rapid growth in the occurrence of a disease is the defining characteristic of an epidemic, but it is important to recognize two further points about the HIV/AIDS epidemic. First, the occurrence of AIDS cases lags behind the spread of HIV infection. Several years typically elapse between the time an adult is infected! with HIV and the appearance of clinical signs sufficient to warrant the diagnosis of AIDS. The contemporary spread of HTV cannot therefore be cliscerned from the current counts of new ATDS cases. So, for example, in the absence of therapies that retard the progression from HIV to ATDS, the epidemic of AIDS cases will continue to rise for several years after the spread of HIV infection begins to decline in a population. Similarly, a sharp decline in the occurrence of new AIDS cases in a given year would not preclude the possibility that the occurrence of new HIV infections had increased during that same year. Unfortunately, the barriers that impede tracking of the spread of HTV infection exceed those that impede tracking of the spread of AIDS cases. Hence, currently available information about the spread of HIV infection is considerably less reliable than information about the occurrence of ATDS cases. Second, the committee would emphasize that a decline in either the spread of HIV infection or the occurrence of new AIDS cases (or both) wouIcl not signal that the danger has passed. HIV is already substantially seeded in the U.S. populationthe number of people who are now infected may surpass 1 million and the virus is likely to continue to spread, if not in epidemic form, then in a persistent, more stable "endemic" form (literally, Dwelling with the peopled. The threat of epidemic and endemic disease will be most serious for those groups that are most heavily seeded with HIV infection, including IV drug users and men who have sex with men, as well as for their sexual partners and offspring. Currently available data also inclicate that the black and Hispanic populations of the United States are experiencing a disproportionate burden of ATDS cases (in particular, cases associated with IV cirug-use, heterosexual, ant] mother-infant transmission). The ATDS case data suggest that these populations may be more heavily seeded with HIV infection than are other ethnic groups and may be disproportionately threatened with further spread of the virus. Our committee is concerned with understanding and reducing the spread of HIV infection, whether this spread be epidemic or endemic in character.

OCR for page 1
4 | AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE MONITORING THE SPREAD OF HIV AND AIDS The overall dimensions of the current HIV/AIDS epidemic in the United States (or anywhere) are hard to determine because the most observable component people who have AIDS is only a small part of the total epidemic. The largest component by far is composed of all those who have been infected by HIV, but the magnitude of this component is difficult to estimate because most infected persons are asymptomatic for several years after their infection. A key first step in controlling the spread of HTV infection and AIDS is the collection of reliable data on the prevalence and incidence of HIV infection and AIDS in the population. A further step requires an understanding of the sexual and IV drug-use behaviors that spread HIV from one person to another and thereby produce changes in HIV prevalence and incidence. The committee believes that more reliable systems must be developed for tracking the course of the epidemic. It also wishes to emphasize in the strongest possible terms that the development of such systems is a prerequisite for mounting a fully effective ant! efficient national response to AIDS. Statistics on AIDS Weekly data from the Centers for Disease Control (CDC) report past and current cases of AIDS. As of November 14, 198S, 7S,312 cases of AIDS had been reported to CDC, and 44,071 people had died as a result. Such statistics are important, and the panel recommends that the system for collecting them be maintained and strengthened. Yet the committee concludes that a fully adequate system for moni- toring the course of the epidemic must go beyond the current system for reporting AIDS cases and deaths: it must also provide reliable monitoring of the prevalence and incidence of HIV infection in the U.S. population. Developing accurate statistical systems to monitor HIV infection is critical for several reasons. . Counts of AIDS cases are out-of-date indicators of the present state of the epidemic because there is a Tong, asymptomatic latency period between HIV infection and the development of AIDS. For example, most adults who will be counted as new AIDS cases in 1989 were probably infected with HIV prior to 1986. . The lives of a substantial proportion of persons infected with HIV will be substantially shortened as a result

OCR for page 1
SUMMARY ~ 5 of that infection. However, these people do not al- ways manifest sufficient symptoms to be captured by the AIDS reporting system. All HIV-infected inclividuals have the potential to trans- mit the infection and thereby spread the epidemic. Statistics on HIV Infection At present, there are no reliable data on the current prevalence of HIV in the United States, although rough estimates have been con- structecl using two quite different methocls. One method aggregates estimates of the size of each major risk group (e.g., the number of persons who regularly inject IV drugs) multiplied by estimates of the HIV prevalence rate for that group. The second method exploits the necessary mathematical links among three time series: (1) the cumulative cases of AIDS to a given time; (2) the cumulative number of cases of HIV infection to that time; and (3) the distribution of the lengths of time that may elapse between infection with HTV anti the appearance of AIDS (the latency, or incubation, period). These two methods agree that the most plausible estimates of prevalence lie in the vicinity of 1 million infected persons (with a range of 0.5-2 million). Admittecily, both of these estimation meth- ods are vulnerable to many sources of uncertainty. These uncer- tainties are of very different kinds, however. The first method is subject to uncertainties about, for example, the number of persons who regularly use IV drugs. The second method is subject to un- certainties about the probabilities that an HIV-infected person will develop AIDS (i.e., 1, 2, 3, n years after infection). Confidence in the rough estimate produced by the two methods is strengthened by the fact that the uncertainties affecting each method are quite clifferent. The committee concludes, nonetheless, that more reliable data on HIV prevalence are needed. In recommending that reliable systems be developed for tracking the course of the HIV/AIDS epidemic, the committee wishes to reit- erate its firm belief that such systems are prerequisites for mounting an effective and efficient national response to AIDS. Without better information on the incidence of new HIV infections in the population, the United States will lack adequate means for determining whether current strategies for controlling the spread of the virus are working. Without better information on the prevalence of HTV infection, the nation will be unable to prepare adequately for future demands for hospital beds and other health care services. Without better data,

OCR for page 1
6 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE scientists and the American public can anticipate endless debates about whether the disease is spreading "rapidly" or "slowly." To the extent that opposing sides in these debates produce ''evidence" from convenience samples,5 inconsistency in conclusions is to be expected, and there is no basis for an informative scientific debate. Reliable as- sessment of the prevalence of HIV infection in a population requires drawing a sample from that population, obtaining a blood specimen from each person in the sample, and accurately testing the specimens for the presence of HIV. O ~ ~ ~ ~ . ~ ~ The validity and hence the usefulness of such HTV prevalence data depend critically on how the sample is chosen from the popu- lation. Fifty years of theory and practice have provider! a valuable statistical toot for this purpose: probability sampling. Drawing prob- ability samples of U.S. households is a well-developec! art; drawing probability samples of populations of special interest (for example, clients of sexually transmitted disease clinics and drug treatment centers) is also within the reach of current statistical technology.6 The use of such methods will allow the monitoring of prevalence over time and the estimation of the incidence of infection, not only for the national population but for specific geographic areas and for groups define by demographic characteristics and behavior. CDC's Family of HIV Seroprevalence Surveys CDC has launched a program to survey HIV prevalence among sev- eral population groups, including clients of drug treatment centers, clinics for sexually transmitted diseases (STDs), tuberculosis clinics, and clinics serving women of reproductive age; patients at general hospitals; and newborn infants. With the exception of the newborn survey, the clinics, centers, and hospitals that will furnish data in this survey program have been purposively selected to facilitate pub- lic health management of the epidemic, and many survey sites have been chosen because they serve populations that are presumed to be especially vulnerable to infection. Such purposive selection, however, compromises the usefulness of the data for estimating prevalence and 5In a "convenience" sample, respondents are selected in a manner that precludes gen- eralization of the statistical findings (e.g., prevalence of infection) with known margins of sampling error to any population beyond the particular individuals included in that sample. 6As discussed in the following section, "National Seroprevalence Survey," the execution of such surveys requires that survey designers grapple with the potential problem of sample bias owing to selective nonresponse in the survey.

OCR for page 1
SUMMARY ~ 7 incidence in any well-defined population of interest. With the excel tion of the survey of newborns, the committee finds that none of these surveys (as currently designed) will provide estimates of HIV prevalence that can be generalized with known margins of error to the population groups of interest (e.g., all clients of STD clinics or patients at general hospitals).7 It is likely, however, that some or all of these surveys can be augmented so as to become probability sam- ples. In such augmented surveys, the present sample elements (the clinics or hospitals included in the present surveys) would constitute one stratum in a stratified probability sample of the populations of general hospitals, STD clinics, and so on. The committee recommencis that efforts be made to re- formulate the CDC family of seroprevalence surveys as prob- ability samples. The committee recognizes that these surveys may serve other purposes, and it acknowle(lges the difficulties and effort involved in such a reformulation and the operational constraints that undeniably weigh heavily on CDC. Nevertheless, the committee be- lieves that wider, if not total, use of probability samples is feasible. Greater involvement of the National Center for Health Statistics (which has recently been made a part of CDC) in the design ant! execution of these surveys may be helpful in achieving this objective. One component of the family of seroprevaTence surveys tests blood specimens from newborn babies, and the committee considers this effort to be a very promising enterprise. Because data are ob- tainecT from all newborns, the survey is free of many kinds of bias. In addition, this survey provides a basis for monitoring the seropreva- lence of childbearing women, a substantial and important component of the sexually active adult population.8 The committee recom- mencis that the newborn infant seroprevalence survey be extencled to include all chilciren born in the Uniter! States. Supplementing the newborn survey with surveys of probability samples of women who have abortions would provide a more complete picture of HIV prevalence among sexually active women of reproduc- tive age. The committee recommencIs instituting a continu- ing anonymous probability survey of the HIV serostatus of 7Chapter 1 of the main report discusses problems that affect the use of data from special populations (e.g., military recruits, blood donors, etc.) to infer HIV prevalence or changes in prevalence in the population at large. 8Newborns of HIV-seropositive women carry the maternal antibody to HIV, even though the infants themselves may not be infected. After some time, the maternal antibod- ies disappear from the infant's blood if the baby is not infected. The Institute of Medicine/National Academy of Sciences 1988 report estimated that there is a 30-50 percent risk of perinatal HIV transmission from an infected mother to her child (p. 35~.

OCR for page 1
8 | AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE women who are clients of clinics that provide abortion ser- vices. This survey will be most valuable if the universe surveyed includes all women who have abortions. National Seroprevalence Survey A seroprevaTence survey based on a national probability sample of households is currently undergoing feasibility testing at CDC's Na- tional Center for Health Statistics. The iclea of such a survey has much appeal, but there are formidable barriers that will have to be overcome before implementation could proceed. A pilot test in Washington, D.C., was recently cancele(1 after protests from the com- munity and the city health department. This experience suggests the extreme sensitivity of all such ciata collection programs; it also un- clerscores the need to fully inform and involve local communities, public health departments, and all groups that might be at risk if the confidentiality of the data collections were to be compromised. The greatest technical barrier to obtaining an accurate estimate of HIV prevalence is the possibility of bias from selective nonresponse. This kind of bias can plausibly occur if, for example, individuals who belong to groups with elevatecl HIV prevalence rates (e.g., gay men, {V drug users) are more likely than people in other segments of the population to refuse to supply a blood specimen. Although the obstacles to conducting a national seroprevalence survey are substantial, they are not necessarily impossible to over- come, and success, if obtained, would be rewarding indeed. Thus, the committee commends the exploratory spirit in which CDC has begun the development of this survey, and it applauds the strategy of using pilot experiments to test the survey's capacity to provide useful direct estimates of prevalence anal, ultimately, of trends in prevalence. The outcome of these experiments should play a decisive role in the final decision of whether to go forward with the national survey. Assuring Confidentiality Much of the information needed to understand and cope with the spread! of HIV infection is obtainable only with the consent of a person who may be harmed if confidentiality is breached. Thus, guaranteeing confidentiality helps protect individual respondents, and it also serves society's interest in obtaining statistical information to help combat the disease.

OCR for page 1
SUMMARY | 9 To maintain confidentiality, safeguards to prevent both deliber- ate disclosure and inadvertent "deductive" disclosure must be put into place. Deductive disclosure can be precluded by coarsely group- ing, modifying, or withholding part of the information before releas- ing or publishing the data. The committee believes that policies for sharing statistical data on HIV and AIDS must provide absolute pro- tection of confidentiality and should seek to provide this protection at the least practical cost in information. Three additional strategies can help in this regard. First, con- fidentiality can be buttressed with legal penalties in the event of a breach. Second, legal protection against discrimination based on HIV status can be established. Third, anonymous testing can be conducted so that the identity of the donor is neither known nor traceable to the blood specimen.9 The committee believes that each of these strategies should be vigorously pursued. Sexual Behavior and AIDS The need to control the spread of HIV infection has forced a recogni- tion of the underdeveloped state of sex research in the United States. Information about sexual conduct is necessary to understand both the epidemiology of the spread of the disease and the social processes that are involved in behavioral change. Yet current understanding is fragmentary, and the underlying research data are often unreliable. Alfred Kinsey and his colleagues pioneered the use of social sci- ence techniques to document the sexual behavior of Americans in the 1940s. The defects of this work are widely known: for example, respondents were disproportionately drawn from the Midwest and from college campuses, and the research did not use probability sam- pling. Still, there can be no denying the crucial historical importance of that work in ushering in a new era in which social science has played a larger role in understanding human sexuality. Since the original Kinsey studies were published in 1948 and 1953, there has been an uneven effort in sex research, in terms of volume and quality, and especially in research relevant to the behaviors that are known to spread HIV. The paucity of solid research contributes to the dilemma now faced by scientists and policy makers 9 Blind testing has been widely employed in studies using blood specimens that have been collected for other purposes. It is, however, feasible to use analogous methods in studies that collect blood specimens for the specific purpose of testing for HIV. In this case, all identifying information would have to be destroyed prior to the HIV test to ensure anonymity.

OCR for page 1
10 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE who are trying to develop intervention strategies to retard] the spread of the virus. Studies are especially lacking on at least five topics: 1. sexuality outside of marriage; 2. sexuality with persons of the same gender; 3. sexuality with persons of both genders; 4. sexual contacts for pay; and 5. variations in sexual techniques among the various types of sexual partnerings. In the past, federal agencies have supported some behavioral research on sexual practices, but much of it has been focused on the sexual behavior of female adolescents, with the goal of preventing teenage pregnancies. The committee believes such research is a valu- able ant} necessary part of the fecleral research portfolio, but basic knowledge of human sexual behavior is needed in many other areas as well. The committee recommends that the Public Health Service support vigorous programs of basic social and be- havioral research on human sexual behavior, particularly through such agencies as the National Institutes of Health; the Alcohol, Drug Abuse, and Mental Health Aciministra- tion; and the Centers for Disease Control. Same-Gender Sex Among Men Because the initial spread of the HIV epidemic was first identified among men who have sex with other men, there has been an upsurge of interest in the number of such men, their sexual practices, and the organization of their social life. The relationship of persons with same-gen(ler sexual orientation to the larger U.S. society has been undergoing substantial change during this century, and that change in itself has affected the sexual and social lives of these men. Estimates of the number of men who engage in same-gender sex- ual behavior figure prominently in the attempts (mentioned earlier) to calculate HIV prevalence. The estimates used in those attempts were derived from Kinsey's studies on male sexual behavior in the period 1938-1948. In addition to the defects in that work that were noted above, the committee finds that the Kinsey studies are not an adequate base on which to formulate estimates of the number of persons in the contemporary population who have sexual relations with persons of the same gender. New fiats, however, are available from two national surveys con- ducted in 1970 and 1988. These studies have their own methodolog- ical difficulties, but data from both of them suggest that a minimum

OCR for page 1
SUMMARY ~ 1 1 of 2-3 percent of American men have sex with other men with some frequency during adulthood. Data from the 1970 survey also suggest that a minimum of 20 percent of adult American mates had at least one sexual experience to orgasm with another male during their lives, and 7 percent of men have such an experience in adulthood (age 20 or oIcler). The Tong history of social intolerance toward same-gender sexu- ality introduces considerable uncertainty about the accuracy of these estimates, which are derived from self-reports obtained in two na- tional surveys. The committee believes that the foregoing estimates are best treated as setting "lower bounds" on the actual number of men who have such experiences. This conclusion follows from the assumption that the number of men in a survey who will conceal the homosexual experiences they have had is greater than the number of men who will report homosexual experiences that never actually occurred. Although our understanding of AIDS in the male homosexual population is far from complete, longitudinal studies initiated during the early years of the epidemic have provided a rich and expanding data base on patterns of HIV transmission and sexual behavior. Early studies among gay men, including the Multicenter AIDS Cohort Studies (MACS), established AIDS as a sexually transmitted disease and identified important risk factors for its spread, including multiple sexual partners and unprotected anal intercourse. In addition to the longitudinal studies that have delineated the risk factors and natural history of AIDS, other studies have been following cohorts of gay men to compile detailed behavioral data over time. These studies offer some indication that behavioral changes to reduce the risk of HIV infection have been occurring in many groups of gay men. Significant decreases in the prevalence of unprotected anal intercourse have been reported in studies undertaken in such urban areas as San Francisco, New York City, Chicago, and Boston. In acldition, significant declines in numbers of sexual partners have been reported in numerous studies of gay mate sexual behavior as it relates to AIDS. Unfortunately, changes in risk-associated behavior have not been universal: high rates of unprotected anal intercourse have been reported in areas that are not foci of the epidemic (e.g., upstate New York and New Mexico). Identifying the factors responsible for behavioral change among gay men is methodologically and conceptually complex. The factors responsible for initial reductions in risk-associated behavior may not be the same factors that are involved in maintaining those behaviors.

OCR for page 1
18 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE nor the seroprevaTence rate is known with any certainty. Current estimates rely on data that were collected for other purposes and that were acquired through efforts intended to measure only crude trends. In sum, the committee recommends that the appropriate government authorities take immediate action to 1. provide drug treatment upon request for IV drug users throughout the country; 2. sustain and expand current programs that pro- vide for "safer injection" to reach al-l current IV drug users in the nation on a continuing basis and with appropriate research evaluation; and 3. establish data collection systems for monitoring present AIDS prevention efforts for IV drug users. The {V drug-using population is also at risk of acquiring and spreading HIV infection through unprotected sexual behaviors. Little is known about the sexual, contraceptive, and childbearing practices of IV drug users, although early studies indicate that more risk- reclucing change has occurred in injection practices than in sexual behaviors. The committee recommends that high priority be given to studies of the sexual ant} procreative behavior of IV drug users, including methods to recluce sexual ant! perinatal (mother-infant) transmission of HIV. Although the committee urges that more basic behavioral re- search be undertaken to improve unclerstanding of risk-associated behaviors ant! how to change them, it also finds that the implemen- tation of intervention programs cannot wait upon the findings of such research. The severity of the AIDS epidemic demands innovative ap- proaches to prevent the spread of infection among {V drug users, with special attention to collecting good evaluation data. Reaching and serving {V drug users will require innovative methods and acIdi- tional resources. Slowing the spreac! of HIV infection in this country depends on the ability to find new ways to reach and influence this population. Plannecl variations of intervention strategies, accompa- nied by sound! evaluation measures, will enable a determination of which kinds of programs are most successful in facilitating change in risky behaviors in this population. 1lFor example, some AIDS intervention programs use mobile vans and cadres of "out- reach workers" who can go into "shooting galleries" and other places in which drug use occurs which have proven helpful in serving people who have not been reached by other services or agencies.

OCR for page 1
SUMMARY ~ 19 LIMITING THE SPREAD OF HIV INFECTION Facilitating Change in Health Behaviors Preventing the spread of HIV infection will require changing those behaviors that are known to transmit the virus and then maintain- ing those changes over time. Whether these goals can, in fact, be accomplished will depend in large part on the effectiveness of the intervention strategies that are brought to bear on the problem. Providing accurate, appropriate information is a logical starting point for any intervention program, although information by itself is unlikely to be sufficient to alter risk-associated behavior. Yet even the accomplishment of this prologue to more complex efforts requires an understanding of the target audience in order to formulate and deliver persuasive messages. To produce action (in this case, behavioral change), a message must reach the appropriate audience ant! be understood. The committee recommends making information available in clear, explicit language in the idiom of the target audience. Furthermore, the committee recommends that sex education be available to both male and female students and that such education include explicit information relevant to the prevention of HIV infection. Health education campaigns of the past, notably those to prevent STDs and drug use, have tencled to rely on fear to motivate those at risk. Research has shown, however, that fear alone is unlikely to succeed. Fear-arousing health promotion messages must also provide specific information on the steps that can be taken to protect an individual from the threat to his or her well-being. The excessive fear generated by heightened perceptions of seriousness and suscep- tibility can be offset by providing assurances that there is, indeed, something that can be done to prevent infection. Therefore, AIDS prevention messages should strike a balance in the level of threat that is conveyed: the level should be sufficiently high to motivate indivicluals to take action but not so high that it paralyzes them with fear or causes them to deny their susceptibility. The mass media can play an important role in providing infor- mation about risk, as well as in molding both the skills and behavior of individuals and the norms of the community to support that be- havior. The committee recommends that television networks present more public service messages on those behaviors as- sociated with HIV transmission and practical measures for interrupting the spread of infection. The committee further

OCR for page 1
20 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE recommends that television networks accept condom aciver- tisements. Efforts should be made to link media representatives with local public health agencies to ensure that the messages are am propriate for the local audiences and that well-clesignec3 evaluations of media efforts are conducted. Although public information campaigns are a sound first step for prevention programs, they are generally insufficient on their own to induce widespread behavioral change. For people to initiate changes in their behavior, they must be motivated, they must believe that the changes being proposed will do some good,- and they must believe they have a reasonable chance of successfully accomplishing those changes. Past research on behavioral change indicates the following: Changes that are consistent with an individual's existing beliefs and values are more likely to be adopted. For some people, modifying behavior through incremen- tal changes is easier to achieve than a global change in life-style. Offering alternative courses of behavior from which an individual can choose is preferable to dictating one "ap- propriate" behavior. . For those who believe in the efficacy of a particular behavior but do not believe they can successfully execute it, skills training can be helpful. Most people need assistance and support to change un- healthy behaviors, ant! most will not be completely suc- cessful in adhering to new patterns. AIDS prevention programs should also incorporate what is al- ready known about the adoption and diffusion of new ideas. For example, opinion leaders of target populations should be identified and used to maximize a program's credibility and persuasiveness with the target audience and to shape more effective messages and programs. Also to be considered! is the community context in which a prevention program is implementecl. If they are to be successful, new programs should be carefully reviewed before being implemented for characteristics that might impede their acceptance in the community. Targeting programs to the community rather than to inclividuals may bring adclitional benefits. Community-level programs have two important points of impact: (1) they can reach a critical mass of individuals to provicle information, motivation, and skills training; and (2) by working through a variety of community agencies, they

OCR for page 1
SUMMARY ~ 21 can foster changes in the norms that stipulate appropriate behavior for community members. One further approach to facilitating behavioral change is the use of HIV antibody testing, an approach that has been found to be helpful in changing behaviors in some groups. The committee recommends that anonymous HIV antibody testing with ap- propriate pre- ant} posttest counseling be made available on a voluntary basis for anyone desiring it. However, to maximize the usefulness of antibody testing in facilitating behavioral change, more knowledge is needed about why individuals seek testing, how testing affects behavioral change in different populations, and how it affects psychiatric morbidity. The committee also emphasizes that HIV testing is not a substitute for broader efforts in education and intervention. Evaluating Interventions The role of evaluation is to allow a determination of which strate- gies actually change people's behavior and which c30 not. Making these determinations requires a systematic process that produces a reliable account of a program's effectiveness. Indeed, preparing for an evaluation can often increase program specificity and quality at the outset. Program innovations that are informed by feedback from careful (and prompt) evaluations can lead to the more rapicl cliscard- ing of poor ideas and the adoption of good ones. The eventual result is a more effective intervention program. The committee recommencis that the Office of the Assis- tant Secretary for Health take responsibility for an evalua- tion strategy that will provide timely information on the rel- ative effectiveness of different AIDS intervention programs. Such a strategy should consider both short- and Tong-term benefits and should be applied to a variety of programs. When possible, for at least each major type of intervention and each major target population, a minimum of two intervention pro- grams should be subjected! to rigorous evaluations that are designed to produce research evidence of the highest possible quality. Vari- ants of intervention programs should be developed for and tested in different populations and in different geographic areas using random assignment strategy accompanied by careful evaluation. When ethi- cally possible, one of the variants should be a nontreatment control.

OCR for page 1
22 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE Randomized Field Experiments One of the most crucial aspects of evaluation involves inferring whether an intervention has had an effect on the target popula- tion. To determine such effects, one must compare what did happen with what wouIc3 have happened if the intervention had not taken place. Because it is not possible to make this comparison directly, inference strategies turn to various proxiesfor instance, extrapolat- ing a trend from past history (before treatment) or using comparison groups. Frequently, however, no similar comparison group can be defined or recruited. Moreover, although adjustments can be made for known differences between two groups, such adjustments may be difficult to make, and there is no way to account for unrecognized differences in two clifferently constituted groups. The remedy for this problem is to establish a singly constituted group in which to assess treatment effects. To be inclucled in the group, a person must satisfy the criteria for inclusion in the program. Then, a subset of that one group is randomly chosen to receive the intervention, thus producing two comparable subgroups that are not identical but are as alike as two random samples drawn from the same population. To maintain the comparability of the two groups, outcome mea- surements must be performed symmetrically for all program par- ticipants (treated and untreated). This idea underlies ranclomized clinical trials and randomized field experiments. The design is a powerful tool, although there will be some cases in which it cannot be applied. The committee recommencis the expanded use of ranclomizec! field experiments for evaluating new interven- tion programs on both individual and community levels. Resources for Evaluation Carrying out evaluations that produce reliable data about the effec- tiveness of interventions to stop the spread of HIV infection requires creative leadership on the part of the management of an organi- zation ant! an attitude among its staff that evaluation is positive and constructive. In addition, priorities must be set, and adequate resources must be made available for evaluation activities. Unfor- tunately, there are seldom enough dollars, expert people, or time to evaluate everything in detail. The selection of a particular in- tervention for in-depth evaluation should depend on several criteria: the importance of the intervention, the extent of the knowledge al- ready in hand, the perceived value of additional information, and

OCR for page 1
SUMMARY ~ 23 the estimated feasibility of the assessment. To use available program evaluation resources most efficiently, the committee recommencis that only the best-designe~i and best-implementec} interven- tion programs be selected to receive those special resources that will be needed to conduct scientific evaluations. There is also a need to upgrade the capacity for evaluation at the local level. Producing quality ciata that will allow program planners to learn from ongoing intervention efforts will require additional re- sources, including appropriately trained personnel. Individuals with expertise in program evaluation need to be identified and brought into the AIDS prevention arena. Unfortunately, many state and local agencies have few connections with inclividuals in this field and will need assistance to identify appropriate people and recruit them. The committee believes it is critical that technical assistance for evalu- ation activities be made available. The committee recommends that CDC substantially increase efforts, with links to extra- mural scientific resources, to assist health departments and others in mounting evaluations. CDC (and any other agency that undertakes AIDS prevention programs) should assign to some administrative unit the responsibility for ensuring the use of planned variants of intervention programs and for overseeing a system of evaluation. OVERCOMING BARRIERS TO RESEARCH AND INTERVENTION Much needs to be done to improve available knowledge about the behaviors that transmit HIV and to control further spread of the infection. The committee finds, however, that some of the needed! actions have encountered and will continue to encounter resistance. Some obstacles arise from the structure of the scientific disciplines involved and the historical lack of support for the kinds of research that are now urgently needed. Other barriers come from within our culture, and they find practical expression in political decisions that restrict the types of AIDS education and intervention activities that governments are willing to funs] or permit. Barriers to Research Although there is growing appreciation of the need for behavioral and social research related to HIV transmission, the personnel to conduct such efforts are currently in short supply at agencies in- volved in AIDS activities. CDC, for example, is managing more than

OCR for page 1
24 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE $150 million of AIDS behavioral research and intervention programs with a small and severely overextended cadre of people trained in rel- evant disciplines. Of approximately 4,500 total employees at CDC's Atlanta facility, fewer than 40 are Ph.D.-leve] behavioral and so- cial scientists, and only a few of those individuals are working on AIDS-related projects. The committee recommends that the number of trained behavioral ant} social scientists employed in AIDS-relatect activities at fecleral agencies responsible for preventing the spread of HIV infection be substantially in- creased. - - Since the early years of the epidemic, CDC has had primary responsibility for AIDS data collection. Some surveys have involved only the collection of physical specimens (e.g., blood) along with a very restricted set of demographic characteristics (e.g., age, sex, residence). At present, CDC's Atlanta stab does not include a sam- pling statistician. The agency's recent acquisition of the National Center for Health Statistics (NCHS) can provide some of the needed statistical expertise, but the role of NCHS in CDC's data collection programs is still being defined. The committee recommends that the CDC AIDS program increase its staff of persons knowI- edgeable about survey sampling and survey clesign, and that it exploit the methodological expertise of the National Cen- ter for Health Statistics. Finally, in adclition to experienced sur- vey scientists, CDC needs technical assistance to evaluate currently funded intervention programs. The committee recognizes that it may be difficult to attract a sufficient number of senior scientists to Atlanta on a permanent basis. One- or, preferably, two-year visiting scientist appointments might provide quick access to needed personnel and allow CDC man- agement greater flexibility in meeting changing staff needs. The committee recommencis the use of PHS fellowship programs and Intergovernmental Personnel Appointments (IPAs) as an interim means for rapidly enlarging the ca(lre of senior behavioral and social scientists working on AIDS programs at CDC and other PHS agencies. Collaborative Research Much of the best behavioral and statistical research on AIDS has oc- curred through collaborations among scientists from universities, the staffs of government agencies (at all levels), and organizations rooted

OCR for page 1
SUMMARY | 25 in the communities that have borne the brunt of the AIDS epidemic. Collaborations of this kind are seldom without conflict; differences of social origin, ethnicity, economic status, or sexual orientation may sometimes lead to misunderstandings. Yet such misunderstancI- ings shouIc3 not deter collaboration; they should be seen as part of an indispensable process of accommodation of varying viewpoints. Efforts to design and implement effective AIDS education without taking into account the idioms and milieu of the target population are doomed to failure. The active and equitable collaboration of persons in the target populations with "outside" scientists and researchers can provide an important safeguard against such failures. Talented, well-trained, and dedicates! workers will be needed for research and intervention efforts at all levels of involvement, from the coordination and funding centers of the federal government to local outreach and education programs. To have credibility in the communities suffering the highest rates of HIV infection and to max- imize the likelihood of successfully preventing the further spread of infection, intervention programs at all levels must increase the in- volvement of minority researchers and minority health care workers in the black, Hispanic, and gay communities. In addition, the com- mittee recommends that special support be provided to foster what are often weak linkages among practitioners (those best positioned to deliver services) and researchers and to remove or reduce organi- zational impediments to the establishment of those relationships. One creative mechanism that has been developed to foster col- laboration is the multidisciplinary AIDS research center. Much of the needed behavioral and social research on AIDS prevention re- quires large, multidisciplinary teams of scientists with close working relationships with many of the different communities in which inter- ventions must be conducted. Recent initiatives, particularly those of the National Institute of Mental Health (NIMH), have been designed to stimulate the formation of multidisciplinary AIDS research centers in those cities that are the current foci of the AIDS epidemic. The committee believes that these initiatives are an appropriate compo- nent of rational strategies for the support of behavioral research on AIDS. The NIMH centers have shown evidence of involving scientists who have valuable links to the communities in which prevention re- search is needled. The committee recommends that support of multiclisciplinary centers for research on AIDS prevention be viewed as a long-term commitment to allow sustained collaborative efforts, inclucling valuable prospective studies.

OCR for page 1
26 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE Social Barriers to Intervention Because AIDS and HIV infection are mainly transmitted by sexual activity and IV drug use, controlling the epidemic requires both a scientific understanding of these behaviors and a social commitment to behavioral interventions that are sufficient to reduce the transmis- sion of HIV infection to a level that cannot sustain epidemic growth. As noted at the beginning of this summary, epidemic disease refers not so much to the number of people who are affected but rather to the sudden appearance and rapid spread of a disease in a commu- nity, a phenomenon that usually evokes strong political and social responses. AIDS education efforts provide a number of examples in which potentially effective prevention activities have become entangled in social conflicts that caused delays in their implementation or a weak- ening of their message. Specific controversies have involved restric- tions on the use of explicit language in educational materials, con- flict between providing scientific information or promoting particular moral values in AIDS prevention messages, and prohibitions against condom advertising on network television. Conflicts of this sort are not unique to this epidemic. Indeed, there are a number of similarities between the AIDS epidemic and epidemics of the past. For example, history suggests that cultural, political, and economic institutions faced with the need to respond quickly to catastrophe often cling to familiar practices, even in the face of an unusual threat. Certainly, instances of such institutional "behavior" can be noted in efforts to mobilize society's forces against AIDS and HIV infection. A further similarity between this epidemic and most others is the disproportionate effect on the poor. Because education is critical to controlling the spread of HIV/AIDS, the ability to reach across class boundaries is vital. Yet those who must design and provide this education sometimes have little knowledge of the culture or conditions of poorer people. Differences of language, values, and life- style can make effective collaboration and communication difficult. Complicating these factors is the additional element of stigma- tization the phenomenon of marking indivicluals or groups as war- ranting exclusion from human society. In its sociological meaning, stigma is the set of ideas, beliefs, and judgments the dominant group in a society holds about another group that it has selectee} as de- serving of scorn or blame (in this case, those who are infected or who have the disease. These beliefs are not merely negative; often,

OCR for page 1
SUMMARY ~ 27 members of the stigmatized group are characterized as dangerous or as deserving of punishment for some vague offense or moral impro- bity, a pattern of thought that has led throughout human history to the blaming and persecuting of minorities as the cause of plagues or scourges. The HIV epidemic has lee! to stigmatization since its beginning. The fact that the disease has been largely confined to mate homosex- uals and IV drug users has made stigmatization almost inevitable, for these groups were already the objects to some degree of the dep- recating judgments that constitute this phenomenon. Fortunately, even though stigmatization has occurred, the public thus far has repudiatecl the worst forms of stigmatizing punishment. Calls for quarantining those infected with HIV have been rejected, and some protections for ensuring confidentiality have been erected in the areas of antibody testing and serostatus disclosure. Nevertheless, more remains to be done. Health professionals have a particular responsibility to counter stigmatization, especially in light of their past success in destigmatizing other conditions: leprosy, epilepsy, and, to some extent, mental retardation. The media also bear a particular responsibility in what they present and how they present it. Similarly, churches and educators have important roles to play. Finally, politicians and the American legal system can affect the extent to which any group in society is subject to stigmatization. A politician who becomes an advocate for the stigmatized obviously takes on an unpopular task. Laws, however, can retard the social pro- cess of stigmatization by prohibiting some of the behaviors that are inspired by it. Research has shown, for example, that the possibility of legal prosecution can alter discriminatory behaviors in various set- tings even in the presence of discriminatory attitudes. The law can also protect those who are infected with HTV from discrimination, and the educational message conveyed by such protection can help to reduce the underlying current of stigmatization that pushes those infected with HIV and AIDS to the outskirts of society. * * * In closing, we return to the theme with which we began: the HIV/AIDS epidemic is partly a social phenomenon, and the ma- jor weapons that are currently available to contain it seek to change the behaviors that spread the disease. Even if fully effective therapies or vaccines were to be found, it is likely that there will be a continuing

OCR for page 1
28 ~ AIDS, SEXUAL BEHAVIOR, AND IV DRUG USE role for behavioral intervention. It is more than 40 years since e~ec- tive drugs against syphilis and gonorrhea became widely available, but those diseases have not yet been eradicated in the United States. Similarly, the struggle to retard the spread of HIV is likely to persist well into the next century. Improved understanding and prevention of the behaviors that spread HIV/AIDS will be needed not only in the short run, however many years that may be, but in the decades that follow any medical breakthrough as well. NOTE: Reference documentation for the material in this Summary is presented in the respective chapters of the full report.