Effect of pneumococcal conjugate vaccine introduction on childhood pneumonia mortality in Brazil: a retrospective observational study

Background: Understanding the real-world effect of pneumococcal conjugate vaccines (PCVs) on pneumonia mortality is crucial because of the expectation that increased PCV use will substantially reduce the burden of pneumonia deaths in children younger than 5 years. However, few post-vaccine introduction studies have estimated the benefits of PCV use on childhood mortality and results have been inconsistent. Therefore, we set out to assess the effect of introduction of ten-valent pneumococcal conjugate vaccine (PCV10) on pneumonia mortality in children in Brazil.

Methods: In this retrospective observational study, we used publicly available mortality data of children aged 3-59 months in Brazil. We separated data by age group (3-11 months, 3-23 months, and 3-59 months) and stratified data by three different socioeconomic factors of Brazilian municipalities (in 2010): Human Development Index, proportion of children living in extreme poverty, and proportion of mothers with no primary education. We first examined long-term trends in childhood pneumonia mortality in Brazil (from 1980 to 2014). We then assessed the effect of PCV10-introduced in Brazil in 2010-both nationally and in municipalities stratified by socioeconomic status, with a synthetic control approach as our primary analytical method.

Findings: Between 1980 and 2010, a period during which Brazil's Human Development Index rose substantially, national pneumonia mortality in children younger than 5 years decreased from about 150 to 15 deaths per 100 000 children younger than 5 years. Despite rapid uptake of PCV10 after its introduction in 2010, we observed a further vaccine-associated decline of about 10% in national childhood pneumonia mortality with our primary analytical method, with a high degree of uncertainty in the estimates. We observed larger reductions in municipal childhood pneumonia mortality in all three age groups (3-11 months, 3-23 months, and 3-59 months) in municipalities with a high percentage of extreme childhood poverty and mothers with no primary education, with the largest decrease observed in children aged 3-23 months in municipalities with low maternal education (24%, 95% credible interval 7-35).

Interpretation: The large reduction observed from 1980 to 2010 in national pneumonia mortality in children younger than 5 years underscores that improvements in nutrition, hygiene, education, and health care have an important role in reducing pneumonia mortality. Although the PCV-associated reduction in childhood pneumonia mortality at the national level was modest, we found that PCV led to larger reductions in low-income municipalities. Similarly, large benefits might occur when PCVs are introduced in other low-income settings.

Funding: Bill & Melinda Gates Foundation and National Institute of Allergy and Infectious Diseases.