February 2, 2007
In 2004, Florida accounted for 11% of the total number of acquired immunodeficiency syndrome (AIDS) cases in the United States, ranking third behind New York and California. Florida also had the second-highest reported AIDS diagnosis rate (behind New York).1 During 2004, non-Hispanic blacks* accounted for 14% of the Florida population but 52% of the 77,421 persons in Florida living with human immunodeficiency virus (HIV)/AIDS. This report describes trends in diagnoses of HIV/AIDS cases that occurred among blacks during 1999-2004 and were reported to the Florida Department of Health. These data indicate that, during 1999-2004, the annual rate of HIV/AIDS diagnosis among blacks decreased more than the rates among other racial/ethnic groups. To examine possible explanations for this decline, HIV/AIDS diagnosis rate trends were compared with trends in gonorrhea diagnosis and publicly funded HIV testing in Florida. The results indicated that gonorrhea diagnosis rates also decreased among blacks in Florida during 1999-2004, whereas the number of HIV tests increased. These findings suggest that HIV/AIDS diagnoses in blacks might be associated with a reduction in high-risk sexual behavior and that the decline was not the result of less testing.
The diagnosis date for HIV/AIDS was defined as the diagnosis date for HIV infection, regardless of whether AIDS was diagnosed subsequently. HIV/AIDS case data were adjusted for reporting delays and insufficient information to enable classification into risk-factor categories.2§ Estimated annual percentage changes (EAPCs) in diagnoses were calculated by Poisson regression. A z test for differences in two parameters was used to evaluate the statistical significance of racial/ethnic differences in EAPCs.
During 1999-2004, the HIV/AIDS diagnosis rate among blacks in Florida decreased from 224.4 cases per 100,000 population in 1999 to 134.0 in 2004 (Table 1). This decrease resulted in a decrease in the disparity in diagnosis rates between blacks and Hispanics (with 47.9 cases per 100,000 population in 1999 and 46.1 in 2004) and between blacks and whites (with 18.8 cases per 100,000 population in 1999 and 18.4 in 2004); the diagnosis rate among blacks was 11.9 times higher than that among whites in 1999 but 7.3 times higher in 2004.¶ During 1999-2004, the rate decreased among black, Hispanic, and white females and among black males. The annual percentage decrease in the rate was greater among black women (EAPC = -10.2) than white women (EAPC = -3.3) and Hispanic women (EAPC = -2.9) (p<0.05) (Table 1). Among blacks, the total number of diagnoses decreased from 1999 to 2004 among men and women with a history of injection-drug use (IDU), men with a history of both male-to-male sexual contact and IDU, and men and women with a history of high-risk heterosexual contact (i.e., sexual contact with a person known to be HIV infected or at high risk for HIV infection [e.g., history of IDU or male-to-male sexual contact]) (Table 2). The EAPC decreased more among blacks than among whites and Hispanics in all risk-factor categories (p<0.05) except men with a history of both male-to-male sexual contact and IDU, among whom the difference between blacks and Hispanics was not significant.
Data reported to the Florida Department of Health regarding the number of diagnosed gonorrhea cases and publicly funded HIV tests indicated that, during 1999-2004, when HIV/AIDS diagnosis rates significantly decreased among blacks, gonorrhea rates also significantly decreased among black males (EAPC = -8.7, 95% confidence interval [CI] = -9.2 to -8.2) and black females (EAPC = -7.4, CI = -7.9 to -6.8) (Table 3). Conversely, during 1999-2004, the annual number of publicly funded HIV tests in Florida increased significantly among blacks, from 81,101 tests in 1999 to 105,072 in 2004 (EAPC = 5.7, CI = 5.5-5.9), among whites, from 108,680 tests in 1999 to 114,103 in 2004 (EAPC = 1.7, CI = 1.6-1.9), and among Hispanics, from 32,050 tests in 1999 to 64,472 in 2004 (EAPC = 15.3, CI = 15.0-15.5).
Reported by: S Lieb, MPH, M LaLota, MPH, T Liberti, L Maddox, MPH, D Thompson, MPH, Florida Dept of Health. R Selik, MD, Div of HIV/AIDS Prevention, National Center for HIV, Viral Hepatitis, STDs, and Tuberculosis Prevention (proposed), CDC.
CDC encourages health departments to use multiple data sources to develop epidemiologic profiles of populations at risk for HIV/AIDS, which can help improve prevention and treatment programs.3 The Florida Department of Health has analyzed data such as those described in this report by county and presented them to county health departments and the Florida public for HIV-prevention planning and community mobilization.
Trends in gonorrhea diagnosis were examined as a possible reflection of trends in high-risk sexual behavior and because gonorrhea typically is diagnosed soon after sexual transmission. Like HIV/AIDS diagnosis rates in Florida, gonorrhea diagnosis rates decreased both among black males and females. Although increases or decreases in diagnosis rates for both HIV/AIDS and gonorrhea might reflect changes in methods of diagnosis, treatment, or surveillance, rather than changes in sexual behavior, the finding that gonorrhea and HIV/AIDS diagnosis rates both decreased suggests that high-risk sexual behavior also decreased.
The possibility that decreases in HIV/AIDS diagnoses were a result of decreased HIV testing among blacks was not supported by data, which indicated a significant increase in testing among blacks at publicly funded HIV testing sites. Approximately 45% of all non-AIDS HIV diagnoses in Florida were reported from these sites.
The findings in this report are subject to at least four limitations. First, because retroactive reporting of HIV cases diagnosed before July 1997 (the implementation date of HIV reporting) was not allowed for persons without AIDS, some persons whose initial HIV diagnosis occurred before mid-1997 might have been misclassified with a later diagnosis date in this analysis if they were retested for HIV during the study period. Depending on whether the diagnosis dates were misclassified to the study period's early (1999-2001) or late (2002-2004) years, the decrease in diagnoses might have been overestimated or underestimated, respectively. Second, data on annual numbers of HIV tests were restricted to public clinics, which might have resulted in an overall underestimation of numbers of tests or a more pronounced underestimation for one racial/ethnic population compared with others. Third, persons who had multiple HIV tests in a certain year might have been counted multiple times in the annual HIV test data. Finally, the validity of the findings might be reduced by any inaccuracy of the adjustments for reporting delay and by missing risk-factor information.
Trends in the diagnosis of HIV/AIDS do not necessarily reflect trends in the transmission of HIV infection because diagnosis trends might be affected by other factors, including changes in testing behavior, clinical practice, or public health surveillance. CDC plans to address these factors by supplementing the HIV/AIDS case surveillance system with estimates of HIV incidence using a previously described serologic testing strategy.4 Meanwhile, examining data from other surveillance systems can assist public health professionals with interpreting HIV/AIDS diagnosis trends, as indicated by this report, in which multiple data sources support the finding that HIV/AIDS diagnosis rates among blacks decreased in Florida and that this decrease might have been associated with a decrease in high-risk sexual behavior.
The continuing high rates of HIV/AIDS and gonorrhea diagnoses among blacks and the significantly increasing numbers of HIV/AIDS diagnoses for white and Hispanic men with a history of male-to-male sexual contact underscore the need for additional and improved prevention measures. Higher rates of HIV testing can be expected to increase the number of HIV- infected persons who are aware of their infection, decrease HIV transmission to others, link infected persons to care and counseling services earlier, and ultimately reduce progression to AIDS and death.5
This report is based, in part, on contributions by P Arons, MD, and R Grigg, PhD, Florida Dept of Health.
Reporting of positive HIV tests by laboratories and of HIV cases (without AIDS) by health-care providers who diagnose or treat a case has been mandatory since mid-1997, and reporting of AIDS cases by health-care providers who diagnose or treat a case has been mandatory since 1983; cases from both the private and public sectors are reported to county health departments. A case is defined as a newly diagnosed condition; a positive test (e.g., from a publicly funded test site) does not become a reported case until it is investigated, at which time the case registry is searched for duplicate reports. Data on numbers of tests conducted at publicly funded counseling and testing clinics include both negative and positive tests and might include more than one test per person.
§Florida Department of Health staff obtained risk information for approximately 80% of HIV/AIDS cases from medical records in hospitals and doctors' offices and from data collected by publicly funded HIV testing sites; in addition, certain patients were interviewed. The remaining cases (i.e., without an identified HIV risk factor) were statistically redistributed into categories for specified risk factors, based on expected results of follow-up investigations.