Combining species distribution models and big datasets may provide finer assessments of snakebite impacts.

Background: Snakebite is a major poverty-related neglected tropical disease. An integrated scientific approach is needed to understand the dynamics of this important health issue. Our objective was to estimate snakebite occurrence in a tropical area by using a blend of ecological modelling and robust statistical analysis. Methods: The present study used climatic, environmental, and human population density data to determine the area with snakebite occurrence-probability for the first time in Bangladesh. We also analysed a large, 16-year dataset of hospitalized snakebite cases to reveal the epidemiology of snakebite in the south-eastern zone of the country. Findings: Our results show that cobra bite is the most commonly occurring venomous snakebite in humans (around ~12% of the total yearly snakebite records), and men are more frequently bitten than women (2/3 of human victims are men). Most bites occur during the rainy season for cobra and green pit viper, while krait bites are not restricted to any particular season. As snakebite incidents are closely related to climate conditions, we can model snakebite risk using temperature and precipitation variables. Whereas there is a lack of snakebite reports from several parts of the study area in official records, our models predict that the entire study area is favourable for snakebite incidents. Based on the combined evidence we estimate that about 200,000 snakebite events occur every year in the south-eastern part of Bangladesh alone. Considering future global climate change, our model projections show that snakebite incidence in Bangladesh might not significantly decrease in the future (- 2070-); however, the distribution of probabilities might change, with a predicted increase of snakebite incidence in the hilly areas of the country. Conclusions: Using climatic data to predict snakebite incidence in Bangladesh allowed us to provide estimations of the total annual number of snakebite cases in the study area. As in most countries, the scarcity of accurate epidemiological data in official records might have masked the real magnitude of this problem. Our analysis suggests that the problem of snakebite envenoming in Bangladesh might be worse than currently perceived. A long-term sustainable snakebite program plan should be designed and institutionalized, considering climatic, geographical and human demographic variables, to obtain better data and facilitate the implementation of accurate snakebite management programs for this country. Author summary: Snakebite envenoming is an important health problem in Bangladesh. A community-based survey conducted in 2009 indicated that around 700,000 snakebites/year may occur in this country. However, until now, there is no study estimating the incidence of snakebite in Bangladesh from the aspect of ecology. Considering the climatic, environmental, and human population variables, a large, 16-year clinico-epidemiological dataset of snakebite patients was analysed to portray snakebite occurrence probability in Chattogram Division, which comprises the south-eastern part of Bangladesh. Five species distribution models were used to map snakebite-prone areas. We found that the study area has large differences in the predicted incidence of snakebite. Our predictions indicate that around 200,000 snakebite cases per year might occur in the study area, where the latest estimate on 2009 estimated less than half that number for this division. Further, predictions for the year 2070 provide us with similar numbers, independent of the global increase in temperature. As snakebite will remain a major health problem for this country in the next 50 years, we propose the development of a sustainable management plan aiming to limit human-snake conflicts and their consequences while safeguarding also snakes and the ecosystem services they provide especially to agriculture. [ABSTRACT FROM AUTHOR]