HIV PREVALENCE ESTIMATES, FLORIDA, 2002
FLORIDA DEPARTMENT OF HEALTH
BUREAU OF HIV/AIDS

Assessment of the extent of the HIV epidemic is an important step in community planning for HIV prevention and HIV/AIDS patient care. HIV prevalence—the estimated number of persons living with HIV infection—includes those living with a diagnosis of HIV or AIDS and those who may be infected but who are unaware of their serostatus.

New HIV prevalence estimates have been developed for Florida and are higher than the estimates distributed by the department five years ago. The new estimates are predicated on those recently developed by CDC, i.e., approximately 850,000–950,000 persons are living with HIV infection in the United States (Fleming, et al., HIV Prevalence in the United States, 2000. [Oral abstract session 5] 9th Conference on Retroviruses and Opportunistic Infections, Seattle, Feb. 24-28, 2002 [http://63.126.3.84/2002/Abstract/13996.htm]).

Florida has consistently reported 10–11% of the national AIDS morbidity and currently accounts for 11% of all persons living with AIDS in the U.S. The Department of Health now estimates that approximately 95,000 persons, or roughly 10–11% of the national total, are currently living with HIV infection in Florida. A plausible range around this point-estimate would be approximately 88,000–102,000. Previously, in 1997, the department estimated that Florida's HIV prevalence was 82,500 (plausible range of 65,000–100,000).

 The Bureau of HIV/AIDS has developed statewide HIV prevalence point-estimates characterized by race/ethnicity, sex, age at diagnosis, current age, and mode of exposure (Table 1). The estimated numbers of HIV-infected persons in each demographic subgroup are based on the proportion of reported cases of those living with HIV/AIDS in each subgroup, according to the HIV/AIDS Reporting System (HARS) as of 12/31/01. Those with no identified risk (NIR) have been redistributed into exposure categories according to how a large sample of persons living with HIV/AIDS initially reported as NIR has been historically reclassified. All the HIV prevalence estimates by demographic characteristics are considered to be approximations, though exact numbers are shown.

Partnership- or area-specific HIV-prevalence point-estimates (Table 2) may be regarded as being more provisional in nature than the statewide estimates. Estimates based on smaller numbers of reported living HIV/AIDS cases would have broader plausible ranges associated with them. In addition, completeness and timeliness of HIV/AIDS reporting may vary by area; to the extent that reportable cases are missed, the share of the state's total HIV prevalence in a given area may be underestimated. Out-of-state residents who receive in-state, HIV-related services may not be captured by the HIV/AIDS Reporting System (HARS), though they could contribute to the community's burden of HIV prevalence. Inter-area movement and migration out of state by HIV-infected persons could also be factors whose precise impact on local HIV prevalence and services may be difficult to assess.

The provisional area-specific HIV prevalence estimates have been developed with an understanding that local community planners may feel there is a sound basis for developing their own estimates, which may not agree with the ones presented here. By May 2002, comprehensive epidemiological profiles for each area will be disseminated to the Florida Community Planning Group and others. The profiles will include provisional, area-specific HIV prevalence estimates, formatted as in Table 1, by race/ethnicity, sex, age at diagnosis, current age, and exposure category with NIRs redistributed.

All available HIV/AIDS surveillance indicators based on reportable data point to certain racial/ethnic disparities. Non-Hispanic blacks, who represent only 14% of Florida's population in the 2000 census (Table 3), account for the majority of recently reported HIV cases, AIDS cases, and HIV/AIDS deaths. Approximately 50% of the statewide estimated HIV prevalence are similarly thought to be among blacks, as indicated in Table 1. The department and the Bureau of HIV/AIDS have taken numerous steps to aggressively address such disparities. As detailed in an addendum to this report, strategies include development of a minority media campaign, establishment of the Florida HIV/AIDS Minority Network, deployment of eight regional minority AIDS coordinators, funding of a minority AIDS initiative and a Closing the Gap initiative, and contracts with community-based organizations to conduct culturally specific HIV/AIDS outreach, early intervention, and linkage into care.

Little is known about those who are HIV infected but unaware of their serostatus or those who have been diagnosed but not reported. However, CDC has estimated that in the past few years the undiagnosed group has diminished from one-third of all HIV-infected persons to one-quarter of all those infected (Fleming, et al., 2002). In Florida, since HIV infection reporting was implemented in mid-1997, we are rapidly learning more about the epidemic as an annual average of about 5,000 newly diagnosed HIV cases and 5,000 newly identified AIDS cases have been reported in HARS. An estimate of statewide HIV incidence would be about 4,000 new infections per year (assuming Florida claims 10% of CDC's estimate of a currently stable national incidence of 40,000 per year), while an annual average of 2,000 deaths have occurred since 1997 among Florida patients reported with AIDS. The current annual growth in statewide reported HIV/AIDS cases (about 10,000 per year) thus appears to be outpacing the estimated annual growth in HIV prevalence (equal to approximately 4,000 [annual HIV incidence] less 2,000 [annual AIDS deaths]). This enables increasingly valid generalizations about the demographics of Florida's epidemic from the known data in HARS to the total population of HIV-infected persons.

A number of other assumptions that underlie the development of HIV prevalence estimates are detailed in Table 4 and Table 5. A general limitation is that the smaller the demographic or geographic subgroup under consideration, the greater the uncertainty about point-estimates. Statistical methods are not currently available to establish plausible ranges around each estimate. For HIV/AIDS community planning purposes, however, there are tangible benefits to referring to statewide and local HIV prevalence in terms of point-estimates, rather than ranges, as long as the caveats about the uncertainties and underlying assumptions are recognized and acknowledged.