Use of a generalized additive model for a spatial analysis of bovine brucellosis risk in the state of Mato Grosso in 2002 and 2014

Diseases that affect cattle represent obstacles to the development of livestock activity. Brucellosis is a significant such disease because it is transmissible, has a chronic nature, and causes health and economic damages to the herd and rural producer. Data from surveys performed in 2002 and 2014 were compared to identify the spatial distribution of bovine brucellosis and to evaluate clusters of outbreaks and areas of greater risk to have infected cattle in the state of Mato Grosso, Brazil. The present study analyzed the data obtained in the aforementioned investigations with a statistical model based on a spatial point process called a generalized additive model (GAM). The analysis made it possible to identify the regions of highest and lowest risk in the state of Mato Grosso. Of the 1001 properties analyzed in 2002, 198 were in areas with high-odds ratio, and 121 were in a low-odds ratio area. Of the 1248 properties sampled in 2014, 119 were in a high-odds ratio area, and 162 were in a low-odds ratio area. Areas with high-odds ratio are more likely to have infected cattle and can be considered to be at higher risk for the disease. The results of the present study highlight the reduction in foci, prevalence, and its relationship with the spatial distribution of bovine brucellosis. The study results should help the official defense service of Mato Grosso direct its activities according to the profile of each region.