In malaria-endemic areas, fever is often assumed to stem from malaria. The objective of this study was to determine the microbiological causes of fever among HIV+ and HIV-Ugandans attending outpatient clinics.

Adult subjects attending rural outpatient clinics Nov 1 2006-April 11 2007 with axillary/oral temp ?37.5°C underwent clinical evaluation and screening for malaria with Binax NOW malaria rapid diagnostic test (RDT) and blood smear (BS) . Most subjects were HIV+, on Septrin prophylaxis and first-line ART. Sterile blood samples collected included 2 Septi-Chek and 1 Bactec Lytic blood culture bottle, serum and whole blood. Sensitivity testing was by BioMerieux ATB system and Kirby-Bauer disk diffusion for chloramphenicol. Sputum Ziehl-Nielsen staining was performed for subjects with respiratory findings. Viral etiologies were not assessed. Treatment was provided using a syndromic management algorithm designed for the study. Definitive therapy was provided for microbiologic diagnoses.

91 subjects had 101 blood samples collected at enrollment or follow-up visits. 82 (90%) subjects were HIV+; 9 (10%) were HIV-. There were 13/101 (13%) positive bacterial blood cultures; 6 (6%) were considered contaminants. Pathogens isolated included non-typhoidal Salmonella, Strep. pneum., and Staph.aureus. 1/61 (2%) mycobacterial blood cultures grew M. tuberculosis. 20/101 (20%) samples were malaria RDT+; 17/101 (17%) samples were BS+. 6/16 (38%) sputum samples were AFB+.

33/91 (36%) subjects had at least one cause of fever demonstrated. Most fevers were not due to malaria, and were unexplained. Implementation of malaria RDT at point-of-care resulted in decreased provision of antimalarials and increased provision of empiric therapy for alternate diagnoses.

For more information on the Rakai Health Sciences Program, please visit our website: www.jhsph.edu/rakai/