Floods and leptospirosis in Brazilian municipalities from 2003 to 2013: use of data mining techniques

Leptospirosis is related to problems with environmental sanitation, and the incidence tends to increase during flood periods. Considering issues related to climate change, floods can be expected to increase. Floods do not affect populations homogeneously, and communities with worse socioeconomic conditions tend to be impacted more heavily. In order to determine whether the number of floods increases the incidence of leptospirosis and its relationship to contextual variables, the study used socioeconomic, environmental, and disease occurrence data at the municipal (county) level. Municipalities suffering problems with sewage disposal showed a higher risk of leptospirosis incidence. Total flooding since the municipality’s declaration of flood emergency was an important risk marker for leptospirosis incidence. Regression tree modeling proved useful for estimating leptospirosis incidence in Brazil.

Keywords:
Leptospirosis; Floods; Dry-Pit; Decision Trees; Climate Change