Trends in reported AIDS cases do not provide a complete picture of the prevalence of the public health problem that HIV infection poses for a population group, community, or nation, because HIV infection per se precedes the clinical diagnosis of AIDS by many years. In some groups, reported AIDS continues to increase after HIV infection has declined. For example, despite very dramatic declines in HIV incidence associated with blood and plasma transfusions after 1985, when antibody screening of blood donations was instituted in the United States, reported cases of AIDS associated with transfusions and in persons with hemophilia continued to increase through the end of the decade. Conversely, reported AIDS case rates may underestimate greatly the future impact of HIV infection, especially in communities or populations more recently affected by the epidemic.

Most notable are the emerging HIV infection epidemics in Thailand, India, Viet Nam, and other areas of Asia, where few cases of AIDS have reached the clinical horizon. For this reason, AIDS surveillance must be accompanied by carefully conducted HIV serosurveys to accurately monitor the public health problem.

Estimates of the prevalence of HIV infection in the United States by the U.S. Public Health Service (USPHS) in 1997 are between 750,000 and 900,000. In addition, it is estimated that approximately 0.6% of men and 0.1% of women are infected; however, the prevalence is much higher for blacks, where 2% of men and 0.6% of women are estimated to be infected. These estimates were based on both available HIV seroprevalence data and on statistical models using AIDS surveillance data and information on the natural history of infection.

National data on HIV prevalence have been directly measured from HIV testing of first-time blood donors, military recruit applicants, job corps applicants, and surveys of antibody status of newborn infants. The prevalence of HIV infection in both military recruit applicants and blood donors grossly underestimates true HIV prevalence rates because homosexual men, IDUs, and persons with hemophilia are discouraged from applying for military service and actively deferred from donating blood.

The highest HIV prevalence rates detected have been among homosexual or bisexual men, IDUs, and persons with hemophilia who received untreated factor concentrates. Prevalence rates in these groups ranged widely in studies-homosexual/bisexual men, 10 to 70%; IDUs, 1 to 50%; and persons with hemophilia, 15 to 90%. Because most surveys in homosexual men and IDUs were conducted among persons seeking medical care for sexually transmitted diseases (STDs) or treatment for drug abuse, the data may not be completely representative of these populations. HIV prevalence rates in persons with hemophilia A and B were directly related to the amount of clotting factor received prior to 1985. HIV seroprevalence rates among female prostitutes varied widely, from 0 to >50%, with the differences largely attributed to the extent of injecting drug use in the population surveyed and the HIV prevalence among IDUs in the community at that time. HIV prevalence rates among male prostitutes parallel rates in homosexual and bisexual men seen in STD clinics in the same communities.

Standardized HIV serosurveys conducted in STD clinics among heterosexual men and women who do not inject drugs showed a median seroprevalence of 0.9% for men and 0.6% for women in 1992. Among adolescents, as with other STDs, HIV infection rates were higher in women than in men. The close association of clinical tuberculosis with HIV infection is apparent from surveys in tuberculosis clinics, in which HIV seroprevalence had a median rate of 10% by 1990.

Check Table 409-3

HIV seroprevalence rates in childbearing women have been measured by blinded testing of residual blood samples collected on filter paper from newborns for routine metabolic screening such as for phenylketonuria. Based on data from 35 states, approximately 7000 births occurred in HIV-infected women annually in 1991-1992, for an annual rate of 1.7 per 1000 childbearing women nationwide. At a perinatal transmission rate of 20 to 30%, 1400 to 2100 infants were infected with HIV perinatally in the United States in 1992. Seroprevalence rates varied widely among states, from <1 per 1000 to >1 to 3% in northeastern urban areas. The survey results in New York State (HIV prevalence of 0.67% statewide and >1.4% in New York City in childbearing women in 1989) resulted in a state policy that encourages HIV counseling of all women of childbearing age and offers counseling and HIV testing to women contemplating pregnancy or already pregnant. Seroprevalence approached or exceeded 0.5% in New Jersey, Maryland, Florida, Puerto Rico, the District of Columbia, and New York by 1992. Nationwide, these blinded surveys were discontinued in 1995.

Because incident HIV infections seldom cause persons to seek medical care, direct measurement of HIV incidence is very difficult in most populations. Using a combination of approaches, the USPHS estimated that 55,000 new HIV infections occurred in adults and adolescents in 1996. HIV incidence estimates must be refined to measure the growth of the epidemic, as well as the effectiveness of prevention efforts.