Spatial analysis of pneumococcal meningitis in São Paulo in the pre- and post-immunization era
Keywords:Meningitis, Pneumococcal, prevention & control, Pneumococcal Vaccines, supply & distribution, Vaccination Coverage, Spatial Analysis, Geographic Information Systems, utilization
OBJECTIVE: To analyze the pneumococcal meningitis incidence rates in the State of São Paulo, Brazil, by age group, municipalities and micro-regions, as well as the spatial distribution of pneumococcal meningitis incidence rates among children under 5 years old in the pre- (2005–2009) and post-vaccination (2011–2013) periods and its associations with socioeconomic variables and vaccination coverage. METHODS: The data source was the Brazilian Notifiable Diseases Information System. For the pre- and post-vaccination periods, thematic maps were built for pneumococcal meningitis incidence in under-5 children, by São Paulo state micro-regions, vaccination coverage and socioeconomic variables, using QGIS 2.6.1 software. Scan statistics performed by the SatScan 9.2 software were used to analyze spatial and spatiotemporal clusters in São Paulo municipalities and micro-regions. A Bayesian inference for latent Gaussian model with zero-inflated Poisson model through the integrated nested Laplace approximation was used in the spatial analysis to evaluate associations between pneumococcal meningitis incidence rates and socioeconomic variables of interest in São Paulo micro-regions. RESULTS: From 2005 to 2013, 3,963 pneumococcal meningitis cases were reported in São Paulo. Under-5 children were the most affected in the whole period. In the post-vaccination period, pneumococcal meningitis incidence rates decreased among this population, particularly among infants (from 4.17/100,000 in 2005 to 2.54/100,000 in 2013). Two clusters were found in pre-vaccination – one of low risk for pneumococcal meningitis, in the northwest of the state (OR = 0.45, p = 0.0003); and another of high risk in the southeast (OR = 1.62, p = 0.0000). In the post-vaccination period, only a high-risk cluster remained, in the southeast (RR = 1.97, p = 0.0570). In Bayesian analysis, wealth was the only variable positively associated to pneumococcal meningitis (RR = 1.026, 95%CI 1.002–1.052). CONCLUSIONS: Pneumococcal meningitis is probably underdiagnosed and underreported in São Paulo. Differentiated rates of pneumococcal meningitis diagnosis and reporting in each microregion, according to the São Paulo Index of Social Responsibility, might explain our results.