Prevalence of HIV infection among hospital patients in north west Tanzania.

In order to estimate hospital HIV prevalence, the economic impact of AIDS on health care and to assess the implications of HIV testing on clinical suspicion of AIDS this hospital based study was done at the government regional hospital of Kagera, Tanzania. Consecutive admissions were recruited into the study, and those consenting had a blood specimen taken, one portion of which was used to aid clinical diagnosis, while the other was tested anonymously for HIV antibodies using two ELISA systems. A short questionnaire was used to specify demographic characteristics, hospital ward of admission and diagnosis of each study subject. The overall age adjusted HIV-1 prevalence was 32.8% (N = 1422) and there was no significant difference in the age adjusted sex specific prevalence. The highest prevalence (53.3%) was found in the 25-34 years age group as well as in the gynaecological and medical wards (41.2% and 40.4%, respectively). The diagnostic category of clinical AIDS had a sensitivity of 11.3% and a specificity of 99.3%, indicating that only 11.3% of the HIV seropositives would have been HIV tested on clinical suspicion of AIDS. Similarly, the HIV-1 antibody sensitivity and specificity for tuberculosis were 5.9% and 97.9%, respectively. Patients who were HIV-1 infected were more likely to have a history of previous hospital admissions, RR = 1.34 (95% CI = 1.16-1.56), and were at an increased risk of developing tuberculosis, RR = 2.02 (95% CI = 1.50-2.70). The diagnostic categories with the highest HIV-1 infection prevalence were clinical AIDS (88.5%), herpes zoster and other HIV-1 skin manifestations combined (85.7%) and pulmonary tuberculosis (58.3%). In conclusion, the prevalence of HIV-1 infection was high among hospitalized patients in Bukoba hospital indicating that the major cause of illness leading to admission to the hospital may have been underlying HIV-1 infection. The findings also indicate that in a high HIV-1 prevalence area, testing for HIV infection on the basis of clinical suspicion of AIDS alone is not sufficient to provide rational care to the majority of HIV infected patients.