How successful are HIV prevention programs? Which HIV prevention programs are most cost effective? Which programs are worth expanding and which should be abandoned altogether? This book addresses the quantitative evaluation of HIV prevention programs, assessing for the first time several different quantitative methods of evaluation.
The authors of the book include behavioral scientists, biologists, economists, epidemiologists, health service researchers, operations researchers, policy makers, and statisticians. They present a wide variety of perspectives on the subject, including an overview of HIV prevention programs in developing countries, economic analyses that address questions of cost effectiveness and resource allocation, case studies such as Israel’s ban on Ethiopian blood donors, and descriptions of new methodologies and problems.
About the Author
Edward H. Kaplan is William N. and Marie A. Beach Professor of Management Sciences, Yale School of Management, and professor of public health, Yale School of Medicine. Ron Brookmeyer is professor of biostatistics, Johns Hopkins University Bloomberg School of Public Health
Read an Excerpt
Quantitative Evaluation of HIV Prevention Programs
Yale University PressCopyright © 2002 Yale University
All right reserved.
Chapter OneOverview of HIV Prevention Programs in Developing Countries
Michael H. Merson and Julia M. Dayton
In this chapter we review the global epidemiology of HIV infection, particularly in developing countries. We summarize what has been learned about the efficacy of behavioral and medical interventions to prevent HIV infection in these countries, and offer a few ideas about how modeling could assist in planning public health prevention programs.
EPIDEMIOLOGY OF HIV IN DEVELOPING COUNTRIES
The United Nations Combined AIDS Program (UNAIDS) estimates that as of the end of 2000 a total of 57.9 million adults and children worldwide have been infected by HIV. About 36.1 million of these individuals have already died from AIDS, and 21.8 million are still living with HIV. Over 90 percent of those living with HIV are in developing countries (Table 1.1). 4.7 million adults and 600,000 children are believed to have acquired HIV during 2000, and almost all of these were living in developing countries.
HIV is a sexually transmitted disease (STD) that affects mostly adults in the prime of life (15 to 49 years). It is transmitted mainly by sex or other direct contact with bodily fluids of an infected person. Most adult transmission in the developing world is by heterosexual sex, though transmission by sex between men occurs often more commonly than realized and in all countries. The second most important mode of transmission is the sharing of unsterilized needles and syringes among injection drug users (IDUS), a practice which is common in many countries in Asia, Eastern Europe, and South America, particularly in Brazil and Argentina. Parenteral transmission also occurs by blood transfusion or medical injection, but these account for a relatively small percentage of infections. Mother-to-child transmission is a third mode of transmission, and is most common in sub-Saharan Africa, the region with the greatest number of HIV-infected women.
How the Epidemic Spreads
The extent to which an infectious disease spreads in a population depends on its reproductive rate, that is, the average number of susceptible people infected by an infected person over his or her lifetime. In order for an epidemic to be maintained, its reproductive rate must be greater than one. The greater the reproductive rate, the more rapidly it will spread. The amount of time a person remains infectious, his or her risk of transmission per sexual contact, and the rate of acquisition of new partners all affect the reproductive rate of any STD, including HIV. Biological, behavioral, and economic factors influence these three variables.
Recent evidence suggests that the average probability of transmission from an infected person to a noninfected person (called infectivity) is greatest during primary infection, the time between exposure to HIV and the appearance of HIV antibodies, and when an infected person is in the advanced stages of the disease and has developed AIDS (as indicated by a low CD4 T-lymphocyte count). Untreated STDs also increase the risk of HIV transmission, particularly genital ulcer disease. Simulations of the initial ten years of the HIV epidemic in Uganda indicate that over 90 percent of HIV infections can be attributed to STDs. Male circumcision has been shown to have a protective effect against HIV transmission.
The rate of partner change, the number of concurrent partners, and mixing patterns influence the reproductive rate. Both the average rate of partner change in a population and the variation across individuals are important. The greater the number of partners, the faster the spread of HIV. In almost all societies, most individuals have few sexual partners during their lifetime, but a small number have many partners. However, this latter group may be sufficient to sustain an epidemic and gradually extend its spread to the rest of the population. The number of concurrent partners appears to be more important than the total number of sexual partners. The more mixing between high-risk individuals (those with many partners) and low-risk individuals (those with few sexual partners), the faster the epidemic will spread in the general population. This is likely to occur, for example, in countries with an active commercial sex industry.
Poverty may also increase transmission. Impoverished women are more vulnerable and less effective in negotiating safe sex with partners and more likely to engage in sex work. Poverty may motivate men to migrate for work, putting them at risk for having more sexual partners while away from the family. Empirical analysis by the World Bank of national-level aggregate data supports these hypotheses.
Factors that accompany economic growth, particularly improved infrastructure and increased travel, can facilitate spread of the epidemic. One reason is that an open economy facilitates the movement of individuals. Economic growth may also contribute to a shift from more conservative to more liberal social attitudes, which may lead to greater individual freedom and more risky sexual activity.
WOMEN ARE AT GREATER RISK THAN MEN
Although globally more men than women have been infected with HIV (with the exception of sub-Saharan Africa), new infections are increasing faster among women. Women are also becoming infected at a younger age than men. Biological, behavioral, and social factors explain these gender differences. Like other STDs, HIV is more easily transmitted from men to women than from women to men. Young women and girls are particularly vulnerable, as they have more cervicovaginal fragility. Recent studies in several African populations found that women aged 15 to 19 are five or six times more likely to be HIV-positive than men of the same age. Females are also more likely than males to have nonsymptomatic STDs, and, since these STDs are likely to go untreated, females have an elevated risk for HIV infection. Social factors, such as a lack of control over the conditions under which they have sex, can also contribute to increased risk for HIV infection. This applies to a spectrum of sexually active females-from sex workers to monogamous, married women.
Impact on Mortality Rates, Life Expectancy, and Population Growth
AIDS causes a large share of mortality in developing countries: at the end of 2000, a total of 17.5 million adults and 4.3 million children under the age of 15 had died since the beginning of the epidemic. By 1990, AIDS was the third leading cause of adult death in the developing world (following tuberculosis and other infectious diseases), but its share of mortality is growing much faster than any other disease. In African countries with the most severe epidemics, over half of adult mortality is attributed to AIDS. It has been estimated that HIV will be the second largest contributor to prime-age adult death in developing countries by the year 2020, accounting for over one-third of adult deaths from infectious disease. Infant mortality rates are higher owing to AIDS, with countries that had significantly reduced infant mortality and with high HIV prevalence rates being relatively more affected. For example, by the year 2010, infant mortality rates in Zimbabwe are estimated to be twice as high as they would have been without AIDS.
The HIV epidemic is also decreasing life expectancy rates in developing countries, reversing important gains made in recent decades. Life expectancies have been reduced by 4 to 26 years in those countries most severely affected by the epidemic. AIDS is also expected to reduce population growth rates in many countries, but rates will remain positive in the year 2010 (although some will be near zero). AIDS has also caused absolute declines in population size, and it is estimated that there are 16 million fewer people today in the 21 most affected countries in sub-Saharan Africa.
Stages of the Epidemic
There is wide variation in patterns of the epidemic across countries and regions of the developing world. A useful way to compare the severity of the epidemic in different countries is to use data on the prevalence of HIV infection among populations who do or do not engage in high-risk behavior. Based on these criteria, the World Bank delineated a typology of "stages" of the epidemic (Figure 1.1):
Nascent: HIV prevalence is less than 5 percent in all known subpopulations presumed to practice high-risk behavior.
Concentrated: HIV prevalence has surpassed 5 percent in one or more subpopulations presumed to practice high-risk behavior, but prevalence among women attending urban antenatal clinics is less than 5 percent.
Generalized: HIV prevalence has surpassed 5 percent in one or more subpopulations at high risk and among women attending urban antenatal clinics.
Sub-Saharan Africa has been most severely affected, with an estimated 25.3 million people living with HIV at the end of 2000. There were 3.8 million new infections in 2000, the most for any region. Heterosexual transmission accounts for over 90 percent of all HIV infections in the region. As discussed earlier, the average age of those contracting HIV has been decreasing, and rates of newly acquired HIV are highest among young women (and to a lesser extent, young men) aged 15 to 24.
At least 20 countries in this region have generalized epidemics. This includes most countries in southern Africa, where HIV is currently spreading most rapidly, probably owing to the recent increase in ease of transportation and economic growth. In South Africa, the epidemic exploded in the 1990s, as indicated by an increase in the prevalence from 4.2 percent among urban antenatal clinic attendees in 1991 to 16 percent among antenatal clinic attendees nationwide in 1997. In Botswana, Namibia, Swaziland, and Zimbabwe, 20 to 26 percent of people aged 15 to 49 are living with HIV today. The West African countries of Cote d'Ivoire and Burkina Faso also have generalized epidemics, but HIV rates appear to be stabilizing in this region.
The AIDS epidemic is also generalized in most countries in East Africa, which was one of the first areas to suffer from AIDS. There is evidence that prevalence is declining in at least one East African country, Uganda, particularly among young people living in urban areas (see below). A few sub-Saharan African countries-Cape Verde, Madagascar, Mauritania, Mauritius, and Somalia-are still at the nascent stage of the epidemic.
In South and Southeast Asia, 5.8 million persons were estimated to be living with HIV at the end of 2000 and infection rates are increasing rapidly in this region. Most countries in the region, including China, India, and Malaysia, have concentrated epidemics. However, levels of infection and modes of transmission vary greatly across the region. Thailand has been most severely affected. In Bangkok the prevalence among IDUS rose from 1 percent in late 1987 to over 30 percent eight months later. In parallel to the epidemic among IDUS, HIV was transmitted among sex workers and their clients, and then to the general population. By mid-1993, the prevalence was 35 percent among IDUS, 29 percent among brothel-based sex workers, and 1 percent among pregnant women. Rates were much higher in the northern part of the country. Since 1993, the overall epidemic has declined, as reflected by a decrease in prevalence among young army conscripts and childbearing women, as discussed later in this chapter.
There are two major epidemics in China, where about half a million people were infected at the end of 1999, up from 200,000 at the end of 1996. One epidemic is occurring among IDUS, who are mostly in the southwest region of the country. As of 1999, prevalence rates in IDU populations reached as high as 70 to 80 percent in some areas, and were on average 30 to 50 percent in this region. The other, newer epidemic is among heterosexuals, mainly along the economically prosperous eastern seaboard, where STDS are also increasing.
In India, prevalence rates in some populations are over ten times those in neighboring China. Although surveillance is limited, it is estimated that about 4 million people were living with HIV at the end of 1999, making India the country with the second highest number of HIV-infected persons (South Africa is higher). In at least 5 states, more than 1 percent of pregnant women are now infected. In Mumbai (formerly Bombay) the percentage of patients at STD clinics who were HIV-infected rose from 23 percent in 1994 to 56 percent in 1999. In Tamil Nadu, prevalence of HIV was as high as 6.5 percent in antenatal clinics in one city. In the northeast, transmission is mainly by needle sharing among IDUS. Manipur State reported 6 percent of the AIDS cases in 1997, all among IDUS, 73 percent of whom were infected by 1996.
In Cambodia, HIV infection has spread to the general population, with 3.76 percent of married women of reproductive age infected with HIV in 1998. In Myanmar, infection among sex workers increased from 4 percent in 1992 to over 20 percent in 1996, and two-thirds of IDUS and about 2 percent of pregnant women were found to be infected in 1996. In Vietnam, over 18 percent of high-risk groups and 0.1 percent of women attending urban antenatal clinics were HIV-infected in 1997.
The epidemic is nascent in other countries of the region. It is not clear why the epidemic has not spread more widely in the Philippines, Indonesia, and Singapore, where the virus has been for several years and sex work is common. One possible explanation regards differences in the commercial sex industry. Evidence suggests that the intensity of epidemics associated with sex workers is determined primarily by the number of sex partners (clients) per sex worker, the frequency of use of commercial sex by men, and the rate of regular condom use in commercial sex. In Indonesia and the Philippines, it is thought that there are fewer customers per sex worker and a smaller percentage of men engaging in commercial sex. Countries like the Philippines, Bangladesh, and Indonesia also have much less intravenous drug use. Another potential explanation for the low rates in Indonesia is the widespread availability of STD treatment.
In Latin America, about 1.4 million people were estimated to be living with HIV as of December 2000. More than half of the countries in Latin America have concentrated epidemics, including Brazil and Mexico. In several countries of the region, including Brazil, Mexico, and the countries of the Southern Cone (Argentina, Chile, Paraguay, and Uruguay), HIV infections were initially concentrated among men who have sex with men and among IDUS.
Excerpted from Quantitative Evaluation of HIV Prevention Programs Copyright © 2002 by Yale University. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.