8% of HIV-infected children) [4]. Rotavirus infection appears perennially in South Africa with a peak during the cooler season in autumn–winter [7]. This aim of this study was to determine the incidence of hospitalisation for acute gastroenteritis in HIV-infected and HIV-uninfected children

from a cohort of children under five years of age in Soweto, South Africa, to assist in determining the burden of hospitalisation that would be preventable with rotavirus vaccine. The study population involved a cohort of 39,879 infants, enrolled at six weeks of age, from 2 March 1998 to 30 October 2000 into a phase III trial which evaluated the efficacy of a pneumococcal conjugate vaccine (PCV) as described [9]. Follow-up for severe illnesses find more in the cohort was undertaken through hospital-based surveillance of all-cause hospitalisation at Chris Hani Baragwanath Hospital (CHBH) until Quisinostat October 2005. CHBH is a secondary–tertiary levels care hospital and the only public hospital in the area. It is estimated that 90% of all admissions in children from the study area occur to this single hospital, where free health care is provided to all children.

All hospitalisations of study participants at CHBH for any cause were identified, clinical information obtained and an examination performed by a study doctor. The study doctors were not involved in the decision to hospitalise a child, or in the child’s management. Standard of care

of all children admitted with acute gastroenteritis included rehydration, either oral or intravenous, correction of many electrolyte abnormalities and early feeding. Antiretroviral therapy (ART) for HIV-infected children was not standard of care in South Africa during the study period. In addition, antiretroviral treatment for prevention of mother-to-child transmission of HIV was not routinely provided to mothers and their newborn infants during the study enrolment period. Based on the measured prevalence of HIV infection among women attending antenatal clinics during the duration of the study period, it was estimated that 24.87% of the children enrolled onto the study were born to HIV-infected mothers. The vertical transmission rate, in the absence of antiretroviral intervention, from mother to child was estimated to be 26%, and thus, 6.47% of the study-cohort was imputed to have been HIV-infected [10]. Children hospitalized for any illness at CHBH were evaluated for HIV infection as previously reported [9]. This included confirming HIV-infection status by HIV-PCR testing in children under 18 months of age and by HIV-ELISA testing in older children. This study involved a secondary analysis of the study database which has previously reported on the impact of PCV on pneumococcal disease, including respiratory illnesses [9] and [10].