TOWARDS A LEPTOSPIROSIS EARLY WARNING SYSTEM IN NORTH- EASTERN ARGENTINA

Abstract

Intro: Leptospirosis is a zoonotic disease that occurs in many regions around the world. Due to the link between extreme climatic events (such as heavy rainfall and flooding) and the risk of leptospirosis, it is considered a climate sensitive infectious disease. In Argentina, leptospirosis is of major public health concern in the Central and North-Eastern provinces. This study aims to quantify the effects of climate on leptospirosis risk in this region and to move towards the development of an Early Warning System. Methods: We developed a Bayesian hierarchical mixed-modelling framework to quantify the effects of lagged Niño 3.4 index, precipitation and river height, on leptospirosis risk in Santa Fe and Entre Ríos provinces between 2009 and 2020. We then tested a set of highest fitting candidate models on their ability to produce subseasonal forecasts of leptospirosis outbreaks in each province. Findings: We found that all climate variables were positively associated with leptospirosis cases in both provinces. Due to the direct influence of ENSO on the local climate pattern, we developed separate models for ENSO and for the local climate indicators, namely precipitation and river height. We found that ENSO models correctly detected 89% of outbreaks in both Santa Fe and Entre Ríos with a three month lead time. Additionally, the inclusion of local climate indicators (precipitation and river height), with a one month lag, reduced the number of false positives in both provinces. Conclusion: Model results suggest that climatic events are strong drivers of leptospirosis incidence, although the association was weaker in Santa Fe compared with Entre Ríos. Furthermore, this study highlights how risk models could be used as prediction tools in an early warning system, by taking advantage of the predictive power of long-lead ENSO predictors and short-lead local climate predictors in determining the risk of leptospirosis outbreaks