Background. Rotavirus vaccines are moderately protective against illness in settings with high compared with low mortality rates. Vaccine effectiveness (VE) evaluations may clarify our understanding of these disparities, but estimates among key subpopulations and against rare outcomes are not available in many analyses due to sample size. We combined 25 data sets from test-negative design case- control evaluations in 24 countries that enrolled children with medically attended diarrhea, laboratory-confirmed rotavirus stool testing, and documented vaccination status. We calculated rotavirus VE stratified by country-level characteristics.
Methods. Children 3–59 months old with birthdates and surveillance hospital arrival dates were included; other variables were standardized as available. Children were considered vaccinated if they received ≥1 dose of rotavirus vaccine >14 days before arrival. We summarized child- and country-level characteristics, including national <5-year-old mortality rate (U5M). Following the manufacturer recommended dose schedule, complete- and partial-series adjusted VE were estimated using logistic regression models.
Results. We included 6626 rotavirus-positive children (case patients) and 19 459 rotavirus negative children (controls). Adjusted complete-series VE was significantly higher among children from countries in the low and medium U5M stratum (74% [95% confidence interval, 64%–81%]) compared with all groups within the high U5M stratum (range, 52% [42%–60%]) to 46% [31%–57%]). Partial- series estimates were lower than complete-series estimates.
Conclusions. These findings are consistent with the published literature, though they suggest heterogeneity in vaccine performance within broad child mortality rate levels. Our findings also highlight the importance of complete-series vaccination.
Keywords. rotavirus; diarrhea; vaccine; immunization; rotavirus vaccine.

rotavirus; diarrhea; vaccine; immunization; rotavirus vaccine