MPH, Public Health Data Science
Spatial correlation, tuberculosis, infections diseases, Amazon, Brazil, Moran’s I, Monte Carlo Method, Geographical Information System
Introduction The most recent report on Tuberculosis by Brazilian healthcare authorities (Ministerio da Saude, 2022) shows that the northern region of Brazil, where all the Amazon Riverine Municipalities are located, has the highest incidence and mortality rate compared to other regions of the country. Locating spatial clusters in this region is thus essential to reduce the disease burden.
Methodology The municipality-level information regarding tuberculosis case rate in the Riverine Municipalities of the Brazilian Amazon in the years 2019-2022 was collected from the DATASUS website. After data cleaning in RStudio, a CSV file was uploaded to ArcMap and joined with geographical data of Brazilian municipalities.
Spatial Cluster analysis was performed using the Global Moran’s I method. The cluster’s significance was estimated by Monte Carlo hypothesis testing.
Results The resulting Global Moran’s I value for the rate of Tuberculosis cases in the Riverine Municipalities was 0.13. This was found by utilizing the nearest neighbor’s method for spatial correlation between polygons. After running 10,000 simulations through the Monte Carlo Method, the calculated p-value for the Global Moran’s I was 0.02, below the predetermined significance value of 0.05.
Conclusion The resulting Global Moran’s I of 0.13 indicates that the riverine municipalities of the Amazonian Region of Brazil tend to present spatial clustering of Tuberculosis cases in the studied period. Through the Monte Carlo Method, the resulting p-value was determined to be below the predetermined significance value, indicating that the spatial correlation is likely to be found in the studied population.
- Ministerio da Saude. (2022). Boletim Epidemiológico Tuberculose 2022. (1st ed).