RESUMO
Background: Studies from Guinea-Bissau and Bangladesh have shown that campaigns with oral polio vaccine (C-OPV) may be associated with 25-31% lower child mortality. Between 1996 and 2015, Ghana had 50 national C-OPVs and numerous campaigns with vitamin A supplementation (VAS), and measles vaccine (MV). We investigated whether C-OPVs had beneficial non-specific effects (NSEs) on child survival in northern Ghana. Methods: We used data from a health and demographic surveillance system in the Navrongo Health Research Centre in rural northern Ghana to examine mortality from day 1-5 years of age. We used Cox models with age as underlying time scale to calculate hazard ratios (HR) for the time-varying covariate "after-campaign" mortality versus "before-campaign" mortality, adjusted for temporal change in mortality, other campaign interventions and stratified for season at risk. Findings: From 1996 to 2015, 75,610 children were followed for 280,156 person-years between day 1 and 5 years of age. In initial analysis, assuming a common effect across all ages, we did not find that OPV-only campaigns significantly reduced all-cause mortality, the HR being 0.96 (95% CI: 0.88-1.05). However, we subsequently found the HR differed strongly by age group, being 0.92 (0.75-1.13), 1.29 (1.10-1.51), 0.79 (0.66-0.94), 0.67 (0.53-0.86) and 1.03 (0.78-1.36) respectively for children aged 0-2, 3-5, 6-8, 9-11 and above 12 months of age (p < 0.001). Triangulation of the evidence from this and previous studies suggested that increased frequency of C-OPVs and a different historical period could explain these results. Interpretation: In Ghana, C-OPVs had limited effects on overall child survival. However, triangulating the evidence suggested that NSEs of C-OPVs depend on age of first exposure and routine vaccination programs. C-OPVs had beneficial effects for children that were not exposed before 6 months of age. These non-specific effects of OPV should be exploited to further reduce child mortality. Funding: DANIDA; Else og Mogens Wedell Wedellsborgs Fond.