| Abstract
|
:
|
Cholera, an acute diarrheal infectious disease caused by the bacterium Vibrio cholerae, displays a seasonal pattern that still remains poorly understood. Two epidemic peaks per year is the typical pattern in endemic areas in Bangladesh and former Bengal, with a decline in the summer during the monsoons, but only one peak coincident with the rainy season is present in other epidemic regions. In this work, the seasonal patterns are studied in the context of the different routes of transmission of the disease. New infections via primary, or environmental, transmission occur from contaminated water sources regardless of previous levels of infection whereas, secondary or human-to-human transmission, links current and previous levels of infections in the population occurring via the fecal-oral route. The analyses on monthly cholera mortality in Chapter 2, allowed the identification of endemic and epidemic areas in former British India, and indicated a dual role for rainfall, influencing positively the occurrence of disease outbreaks in epidemic areas, whereas negatively in the endemic areas. The analysis in Chapter 3 examined two decades of high resolution spatio-temporal cholera data from Matlab, Bangladesh, from a static statistical perspective and found evidence for spatial aggregation of cases, supporting secondary transmission. In chapter 4 we extended this analysis using a metapopulation approach for modeling cholera, including its temporary immunity and, seasonal and interannual variation in the routes of transmission, at the spatial resolution of the data. A Markov Chain Monte Carlo approach was used to estimate all the parameters of the family of discrete state models (i.e., Susceptible-Infected-Recovered-Susceptible) developed here. The results on the force of infection suggests a central role for secondary transmission during the whole year. Whereas, primary transmission seems to play a weak role, except in a short window of a few weeks preceding the spring peak of cholera cases.
|