Comparative study of the epidemiology and aetiology of bloodstream infections in hospitalized adult patients in Tanzania, Malawi, and Thailand : the role of human immunodeficiency virus type 1 (HIV-1) infection
Before 1995, the frequencies of mycobacterial and fungal bloodstream infections (BSI) in human immunodeficiency virus (HIV)-infected populations in sub-Saharan Africa and Southeast Asia were unknown. Therefore, a prospective survey of febrile (oral temperature >38 C or axillary temperature >37.5 C) adult patients who presented to sentinel teaching hospitals in Tanzania (1995), Thailand (1997), and Malawi (1997 dry season and 1998 wet season) was conducted. The objectives were to (i) determine the aetiology, prevalence, and clinical correlates of BSL and (ii) characterise the role played by HIV infection. After informed consent, a detailed history was recorded for each patient followed by physical examination. Next, blood was cultured for bacteria, mycobacteria, and fungi, and tested for HIV and malaria. Data were collected for 517 patients in Tanzania, 246 in Thailand, and 471 in Malawi. Respective BSI, HIV, and malaria parasitaemia rates were: Tanzania: 28%, 55%, 9.5% Thailand: 48%, 74%, 0 Malawi dry season: 30%, 74%, 4% Malawi wet season: 28%, 73%, 31%. The most frequently isolated bloodstream pathogens were Mycobacterium tuberculosis (MTB) and non-typhi Salmonella species (NTS) in Tanzania MTB and Cryptococcus neoformans in Thailand MTB and Streptococcus pneumoniae during Malawi dry season and MTB and NTS during Malawi wet season. In each country, HIV-infected patients were significantly more likely to acquire BSI all patients with mycobacteraemia were HIV-infected. The Malawi findings are the first description of seasonal variation in the occurrence of S. pneumoniae and NTS bacteraemias. Logistic regression models yielded predictors of BSI in Thailand (HIV infection, chronic diarrhoea, lymphadenopathy, or splenomegaly) and Malawi (HIV infection, chronic fever, oral candidiasis, or acute diarrhoea). In populations with high prevalence rates of HIV infection, MTB has become the foremost cause of documented BSI. Similar season- and country-specific surveys, performed periodically in HIV-endemic regions will provide data on the aetiology and predictors of BSI, and facilitate empirical therapy of febrile illnesses.