Decline in diarrhea mortality and admissions after routine childhood rotavirus immunization in Brazil: a time-series analysis

Abstract

Background: In 2006, Brazil began routine immunization of infants <15 wk of age with a single-strain rotavirus vaccine. We evaluated whether the rotavirus vaccination program was associated with declines in childhood diarrhea deaths and hospital admissions by monitoring disease trends before and after vaccine introduction in all five regions of Brazil with varying disease burden and distinct socioeconomic and health indicators.

Methods and findings: National data were analyzed with an interrupted time-series analysis that used diarrhea-related mortality or hospitalization rates as the main outcomes. Monthly mortality and admission rates estimated for the years after rotavirus vaccination (2007-2009) were compared with expected rates calculated from pre-vaccine years (2002-2005), adjusting for secular and seasonal trends. During the three years following rotavirus vaccination in Brazil, rates for diarrhea-related mortality and admissions among children <5 y of age were 22% (95% confidence interval 6%-44%) and 17% (95% confidence interval 5%-27%) lower than expected, respectively. A cumulative total of ~1,500 fewer diarrhea deaths and 130,000 fewer admissions were observed among children <5 y during the three years after rotavirus vaccination. The largest reductions in deaths (22%-28%) and admissions (21%-25%) were among children younger than 2 y, who had the highest rates of vaccination. In contrast, lower reductions in deaths (4%) and admissions (7%) were noted among children two years of age and older, who were not age-eligible for vaccination during the study period.

Conclusions: After the introduction of rotavirus vaccination for infants, significant declines for three full years were observed in under-5-y diarrhea-related mortality and hospital admissions for diarrhea in Brazil. The largest reductions in diarrhea-related mortality and hospital admissions for diarrhea were among children younger than 2 y, who were eligible for vaccination as infants, which suggests that the reduced diarrhea burden in this age group was associated with introduction of the rotavirus vaccine. These real-world data are consistent with evidence obtained from clinical trials and strengthen the evidence base for the introduction of rotavirus vaccination as an effective measure for controlling severe and fatal childhood diarrhea.

Figure 1. Trends in childhood diarrhea deaths and admissions in Brazil.

Each analysis examines trends from 2002–2009, by age, including comparison of observed events (blue dots) after rotavirus vaccination (2007 to 2009) in Brazil with expected events (solid line) and 95% CIs (gray shaded area) in the absence of vaccination. Expected number of events and 95% CIs are based on predictions from regression models fitted to historic data from each region (2002 to 2005).

Figure 2. Impact of rotavirus vaccination on monthly events of childhood diarrhea deaths in Brazil by region.

Each analysis compares the monthly observed events among children under 5 y after rotavirus vaccination (2007 to 2009) with expected events in the absence of vaccination, by region. Expected number of events and 95% CIs are based on predictions from regression models fitted to historic data from each region (2002 to 2005).

Figure 3. Change in diarrhea mortality and admission rates after vaccination by year and region.

The maps depict the percent change in observed rates of diarrhea mortality and admission after rotavirus vaccination in Brazil compared to expected rates without a vaccination program, by year and region. Expected numbers of events are based on predictions from regression models fitted to historic data from each region (2002 to 2005). The percent declines are computed as one minus the RR of diarrhea events post-vaccine compared to the pre-vaccine era from Poisson regression models, adjusting for seasonality and secular trends.

Figure 4. Impact of rotavirus vaccination on monthly events of diarrhea admissions in Brazil by region.

Each analysis compares the monthly observed events among children under 5 y after rotavirus vaccination (2007 to 2009) with expected events in the absence of vaccination, by region. Expected number of events and 95% CIs are based on predictions from regression models fitted to historic data from each region (2002 to 2005).