A framework for predicting endemic cholera using satellite derived environmental determinants

Cholera remains one of the most prevalent water-related infections in many tropical regions of the world. Macro-environmental processes provide a natural ecological niche for Vibrio cholerae and because powerful evidence of new biotypes is emerging, it is unlikely that the bacteria will be fully eradicated. Consequently, to develop effective intervention and mitigation strategies, it is necessary to develop cholera prediction models with several months' lead time. Almost all cholera outbreaks originate near the coastal areas and cholera bacteria exhibit a strong relationship with coastal plankton. Using chlorophyll as a surrogate for plankton bloom in coastal areas, recent studies have postulated a relationship between chlorophyll and cholera incidence. Here, we show that seasonal cholera outbreaks in the Bengal Delta can be predicted two to three months in advance with an overall prediction accuracy of over 75% by using satellite-derived chlorophyll and air temperature data. Such high prediction accuracy is achievable because the two seasonal peaks of cholera are predicted using two separate models representing distinctive macro-scale environmental processes. We have shown that interannual variability of pre-monsoon cholera outbreaks can be satisfactorily explained with coastal plankton blooms and a cascade of hydro-coastal processes. Post-monsoon cholera outbreaks, on the other hand, are related to macro-scale monsoon processes and subsequent breakdown of sanitary conditions. Our results demonstrate that satellite data over a range of space and time scales are effective in developing a cholera prediction model for the Bengal Delta with several months' lead time. We anticipate our modeling framework and findings will provide the impetus to explore the utility of satellite derived macro-scale variables for cholera prediction in other cholera endemic regions.