"Scope of the AIDS Epidemic in the United States"
CDC National AIDS Hotline Training Bulletin #171
February 13, 1996
These are answers from Centers for Disease Control and Prevention (CDC) to questions concerning an article on the "Scope of the AIDS Epidemic in the United States" in Science, November 24, 1995, by Philip S. Rosenberg, Ph.D.
1. How do these data compare with previous estimates of HIV prevalence?
This study estimates that between 630,000 and 900,000 Americans are living with HIV (this is called "HIV prevalence"). This estimate is derived from statistical techniques that have been refined and expanded since the 1990 Public Health Service (PHS) estimate of 800,000-1,200,000 people infected. Researchers now have additional years of data and more precise information on the completeness of AIDS reporting, which suggest that the upward range of the previous estimate was probably too high. The new estimate more accurately reflects the current state of knowledge.
2. What is this new information? And why does that change the estimate?
Because "back calculation" is based on actual AIDS cases reported to CDC, researchers must estimate how many HIV infections are never reported to CDC as cases of AIDS, and incorporate that number into the overall estimate. The calculation of the percentage of cases never reported must include both the number of people diagnosed with AIDS who are simply never reported and the number of people estimated to die with HIV before developing AIDS.
Recent examinations of the CDC reporting system, coupled with more complete information on the number of people who die with HIV before being diagnosed with AIDS, have shown that the system is more complete than previously thought. Researchers now believe that approximately 85% of people with HIV will eventually be reported to CDC with AIDS.
The previous estimate was based on the assumption that 75% of people with HIV were eventually reported to CDC with AIDS. Because the percentage of people missed is now thought to be significantly lower (15% rather than 25%), the percentage that must be added to the "back calculated" number of HIV infections to account for these individuals is lower. Therefore the estimated number of people infected in lower, likely providing a more accurate picture.
3. How accurate is this estimate?
This estimate is based on the best available information to date and is consistent with estimates from a number of different sources. However, there is no way to determine precisely how many individuals are infected with HIV. We can not directly observe changes in the numbers of people being infected with HIV (the way we can analyze AIDS cases). And determining where and how many new cases of HIV are occurring is an imprecise science that must rely on "snapshots" of the number of infections in specific subgroups of the population and complex statistical modeling.
Because the AIDS epidemic is continuously evolving and is currently behaving differently in different populations, the estimate could indeed change slightly as additional years of data become available. From a prevention standpoint, the absolute number of infections is less important than changes over time and changes in specific groups.
4. Is HIV prevalence increasing or decreasing?
The Rosenberg analysis, as well as serosurveys in subgroups of the entire U.S. population, indicate that prevalence has remained relatively stable in recent years. This means that roughly the same number of people are becoming infected with HIV as the number of people who are dying with severe HIV disease.
5. There seem to be a lot of conflicting news stories about the state of the epidemic. Some indicate the epidemic is increasing and others suggest that it is leveling. Why is this?
AIDS is not one but several epidemics. The epidemic continues to evolve and is behaving differently in different groups of people--by age, race, risk group, and geographic region. To evaluate the status of these diverse epidemics, CDC and others rely on several different data sets and studies.
Confusion most frequently stems from the generalization of findings from individual studies, rather than a combination of data. Additionally, some confusion has resulted from the intermingling of terms--such as HIV incidence, HIV prevalence, AIDS incidence, and AIDS deaths--all of which provide important but distinct information.
6. Is this just another way of looking at the data?
This is the first study to analyze national AIDS data and estimate HIV prevalence by age. And consequently, it provides the clearest picture to date of the continuing toll of AIDS in young Americans.
7. If each generation faces the same epidemic, why continue to invest in prevention?
These findings suggest that prevention has helped stabilize the rate of infection in the population overall, and may have contributed to a decline in infection in older white males.
This study underscores the need for sustained, targeted prevention efforts to reduce this repeated toll. While 40,000 new infections is much less than the dramatic rate of increase that characterized the mid 80's, it is not acceptable to allow another American to become infected for every one who dies from this preventable disease.
8. Why is HIV prevalence so low in white women?
A number of underlying factors may contribute to the disproportionate impact of AIDS on women of color. While it is impossible to determine what those factors are from this data, local prevention planners should consider social, economic, and cultural factors that may influence risk behaviors.
9. Why does this analysis only go through January 1993?
The Rosenberg study uses a method of "back calculation" that can currently only be applied to cases reported to CDC under the pre-1993 "case definition."
The "case definition," the criteria used to report people with severe HIV-related illness or AIDS, was changed in 1993. CDC added three medical conditions and a measure of a severely suppressed immune system to more accurately capture all of those individuals severely ill and in need of care.
Rosenberg and other researchers are now devising methods that can applied to the new reporting system. These estimates can then be updated to reflect the most recent data.
10. What does this study tell us about risk groups? How many new infections are related to heterosexual sex, drug use, and gay and bisexual sex?
While the Rosenberg study does not analyze the data by risk group, the characteristics of recently reported AIDS cases do provide indications of where recent infections may be occurring. The most recent reports of AIDS cases suggest that about 1/2 of cases are among men who have sex with men, about 1/4 of cases are among IDU, and about 1/8 among heterosexuals.
11. Will the repeated waves of HIV infection in young adults continue?
With sustained, targeted prevention efforts for each group entering young adulthood, we can help keep these waves from developing.
The Rosenberg study indicate that the rate of infection among young adults remained relatively stable through 1992. As each generation (through young adults born in 1969) has come of age, they have faced the same pattern of HIV infection--with a substantial increase in the rate of infection as individuals enter their late teens and early twenties, with rates peaking in the mid- to late-twenties. If we can intervene and change behaviors as young adults reach this age, perhaps, in the next analysis the pattern will change.
12. Do these findings surprise you?
The fact that new infections are occurring primarily in young adults is certainly not surprising. Many other studies have suggested a falling average age of infection. As the epidemic has matured in this country, one would expect the new entrants to the at-risk population to be young.
One would hope, however, that young adults today would be becoming infected at much lower rates than young adults several years ago--as the population has become more aware of how to prevent HIV infection. This study reinforces the need for sustained prevention, and perhaps also reinforces both the difficulty and the importance of changing behavior among our nation's young adults.
13. What should be done now?
This study points to the need for increased prevention efforts for young adults. Because of the diversity of the epidemic among different groups, communities must look more closely at what is going on locally and design programs to address the prevention needs of young adults--particularly minorities. And when designing prevention programs for young adults, young adults should be included in the design, development, and delivery of messages and services.
Disclaimer: CDC Hotline Training Bulletins
The information in the "CDC Hotline Training Bulletins" is provided by CDC and NIH for use by the CDC National AIDS Hotline in responding to general questions from the public about HIV and AIDS. The bulletins are not intended to be comprehensive discussions of the subject areas. Treatment and drug therapy options change as new research and clinical experiences broaden scientific knowledge. Therefore, persons seeking information on drug therapy should refer to the product information sheet included in all drug packages for the most current and accurate information about a particular drug, especially if the drug is new or infrequently used. HIV-infected individuals should consult their personal physician for specific concerns about their health. For persons desiring more information on a specific topic, public, medical, and university libraries can provide excellent references.
The AIDS Clinical Trials Information Service (800-874-2572) can provide information about ongoing HIV/AIDS clinical trials; the HIV/AIDS Treatment Information Service (800-448-0440) can assist with information about the latest treatments for persons with HIV infection or AIDS.
This article was provided by U.S. Centers for Disease Control and Prevention. Visit the CDC's website to find out more about their activities, publications and services.