RESUMEN
BACKGROUND: Brazil initiated universal immunization of infants with the G1P[8] human rotavirus (RV) vaccine in March 2006. This study evaluated vaccine effectiveness (VE) against severe rotavirus gastroenteritis (RVGE) hospitalizations. METHODS: Matched case-control study conducted at 4 hospitals in Belém from May 2008 to May 2009. Cases were children hospitalized with RVGE age-eligible to have received 2 doses of the human RV vaccine (≥ 12 weeks of age and born after March 6, 2006). For each case, 1 neighborhood and 1 hospital control without gastroenteritis was selected, matching by birth date (± 8 and ± 6 weeks, respectively). Matched odds ratio of 2-dose RV vaccination in cases versus controls was used to estimate VE (1 - odds ratio × 100%). RESULTS: Of 538 RVGE cases, 507 hospital controls and 346 neighborhood controls included, 54%, 61%, and 74% had received both RV vaccine doses. VE against RVGE hospitalization was 75.8% (95% confidence interval [CI]: 58.1-86.0) using neighborhood controls and 40.0% (95% CI: 14.2-58.1) using hospital controls. VE in children 3 to 11 months and ≥ 12 months of age was 95.7% (95% CI: 67.8-99.4) and 65.1% (95% CI: 37.2-80.6) using neighborhood controls, and 55.6% (95% CI: 12.3-77.5) and 32.1% (95% CI: -3.7-55.5) using hospital controls. G2P[4] accounted for 82.0% of RVGE hospitalizations. G2P[4]-specific VE was 75.4% (95% CI: 56.7-86.0) using neighborhood controls and 38.9% (95% CI: 11.1-58.0) using hospital controls. CONCLUSIONS: Although fully heterotypic G2P[4] was the predominant RV strain, good VE was demonstrated. VE was highest in children aged 3 to 11 months. However, protection in children ≥ 12 months of age, important for optimal public health impact, was significantly sustained based on estimates obtained using neighborhood controls.