Decreasing trends of bacteraemia among HIV-infected Ugandan adults: incidence, aetiology, clinical outcomes and effect of antiretroviral therapy in a semi-urban setting (2000-2008).

OBJECTIVE: To investigate the effect of antiretroviral therapy on trends of incidence, aetiology and clinical outcomes of bacteraemia among HIV-infected Ugandans in a semi-urban setting. METHODS: A cohort of HIV-1-infected Ugandans aged 15 or older was followed from 2000 to 2008. Clinical, haematological and immunological measurements were taken at 6-monthly visits. Additionally, patients reported to outpatient clinics whenever they were ill. Patients with elevated axillary temperature above 37.4 °C consistently triggered clinical assessment (with mandatory blood cultures) and empirical management protocol. Daily cotrimoxazole prophylaxis and highly active antiretroviral therapy (HAART) were introduced stepwise to eligible patients in August 2000 and February 2003, respectively. We compared the rates of bacteraemia across five calendar periods using random-effects Poisson regression for the effect of HAART at the population level. RESULTS: A total of 246 bacteraemia episodes (including multiple episodes) were documented among 188 individuals (crude incidence: 42.4 events per 1000 person-years; 95% CI: 35.0, 51.4). The most common species isolated was Streptococcus pneumoniae. After adjustment for current age, clinical characteristics at enrollment (CD4+ T-cell counts and WHO stage) and time since enrollment, the incidence of bacteraemia dropped significantly when HAART was widely available compared with the period when treatment was not available (adjusted hazard ratio: 0.17; 95% CI: 0.09, 0.35). No poor health outcomes (death or lack of clinical response to antibiotics) after bacteraemia occurred after complete access to HAART. CONCLUSIONS: HAART availability in a resource-poor setting substantially reduced the trends of bacteraemia among HIV-infected adults. This may further impact on future morbidity and healthcare costs of HIV-infected people.