Tuberculosis is a disease transmitted by aerosols through person-to-person contact, caused by M. tuberculosis. Although BCG vaccine, which protects against the most severe forms of the disease, is available nationwide, vaccination coverage has not reached the ideal rate of 90% recommended by the World Health Organization. This contributes to disease persistence and increasing case numbers, as reported by WHO, which estimated 10 million new cases and 1.6 million deaths in 2021 alone. To spatially estimate BCG vaccination coverage rates in the state of São Paulo between 2018 and 2024 and correlate them with relevant economic variables. Ecological and exploratory study with information obtained from DATASUS on BCG vaccination coverage in São Paulo municipalities between 2018 and 2024, correlated with economic variables such as the Paulista Municipal Development Index (IPDM), obtained from Fundação SEADE, and per capita income from IBGE. Data were entered into TerraView to identify spatial autocorrelation, estimated by Moran’s I (IM), and to construct thematic maps. Data were also analyzed in GeoDa to estimate Bivariate Moran’s I (IMb) and LISA, with α<5%. Mean coverage across the years studied was 85.79%. In all seven years, coverage remained below the 90% target, with the lowest rates recorded in 2020 (71.45%) and 2021 (68.76%). IM for coverage over the period was 0.09 (p<0.01), and the thematic map indicated below-ideal coverage in 309 municipalities, particularly in the regions of Sorocaba, Greater São Paulo, Vale do Paraíba, and the central region of the state. In total, 110 municipalities had very low coverage (<75%), and another 199 municipalities had below-ideal coverage (75–90%). IMb revealed a paradoxical negative autocorrelation with socioeconomic variables: IPDM (IMb = -0.16) and per capita income (IMb = -0.09). The study evidenced clusters of municipalities with low tuberculosis vaccination coverage, data that can support municipal and regional health managers. Additionally, paradoxical IMb results highlight the need to include maternal variables, such as maternal age, number of children, and marital status, in future analyses.
Souza et al. (2026) studied this question.