Because few gastroenteritis Trichostatin A episodes met the ≥17 score criterion used to define severe in the traditional Clark score applied in health facilities (i.e. 1.6% of episodes), we considered a score of ≥16 as severe using the modified Clark score for this analysis. Secondary objectives in the home visit analysis included evaluation of all gastroenteritis episodes regardless of severity, the incidence of febrile illness and acute

lower respiratory illness (ALRI), medication use, and healthcare-seeking. In Kenya, stools were transported in cool packs from the rural clinics to KEMRI/CDC laboratories within 6 h of collection. Stools were cultured and assessed for pathogenic enteric bacteria (excluding E. coli) using standard microbiologic methodologies [16]. For rotavirus testing, stool specimens were stored at −20 °C until

shipment to Merck Research Laboratories. The rotavirus testing methods, including genotyping, used in this study have been previously described [7], [10], [17] and [18]. Voluntary HIV counseling and testing was offered to all children. The Determine® HIV-1/2 rapid test (Abbott Laboratories, Tokyo, Japan) was performed to detect HIV antibodies. The Roche Amplicor HIV-1 DNA test version 1.5 (Roche Diagnostic System, Branchburg, NJ, USA) was also performed on all infants 6 weeks of age or greater, to confirm HIV infection by polymerase-chain-reaction (PCR). The PCR result was taken as the definitive result for infant HIV infection for the purposes of analysis, Selleckchem Screening Library and all positive PCR tests were repeated for verification. Children with presence of HIV antibodies with negative PCR results were considered HIV-exposed. Children were also tested for HIV (both antibody and PCR) at 9, 12, and 18 months from enrollment to detect MTMR9 acquisition of new HIV infection. For the clinic-based catchment surveillance, overall efficacy was defined as 1 − Rvaccine/Rplacebo × 100%, where R represented the incidence

for the respective groups, as has been described before [7] and [10]. The primary analysis of efficacy was based on the per-protocol subject population. No specific sample size calculations were done for the Kenya site separately from the main study. In the home visit analysis, the denominator for incidence calculations was the person-time determined from the 14 days of observation at each home visit. Time to incidence episode was calculated as symptom free days preceding the episode. Only one episode of gastroenteritis could be reported for each 2-week period. Unlike in the facility-based analysis, episodes occurring after the first episode, in subsequent home visits, were included in the numerator, as it was not possible to determine which episodes were caused by rotavirus. Both severe and all episodes of gastroenteritis were compared between groups.