Heterogeneities in Cholera Transmission and their Implications for Vaccination

Despite being one of the first infectious diseases to be isolated and studied in modern times, cholera remains a public health threat throughout the world. Like many diseases, cholera transmission potential can vary widely from setting to setting with heterogeneity resulting from numerous factors including the human immune response and transmission conduits provided by the built environment and local hygiene practices. Recent scientific and policy developments have made the reactive use of oral cholera vaccines a potential tool in the fight against cholera, but limited vaccine supply and transmission heterogeneity leave many open questions about to to best use these vaccines. In this dissertation, we first present an in-depth case study on a 2009 cholera epidemic in Bissau City, Guinea Bissau where we fit a stochastic disease transmission model to spatially explicit surveillance data. With this model we explore the impact of different reactive vaccination strategies under limited vaccine supply and logistic constraints. We find large heterogeneity in the transmission potential between neighborhoods in Bissau City and find that a single neighborhood was a necessary driver of the epidemic. If vaccine supply were limited but could have been distributed within the first 80 days of the epidemic, targeting vaccination at this `hotspot' neighborhood would have averted the most cases both within the hotspot and throughout the city. While we find that vaccine should have been targeted at the transmission hotspot in Bissau City, this may not be the case elsewhere or when considering hotspots at different spatial scales. We use deterministic meta-population models to explore the conditions under which reactive vaccination targeted at transmission hotspots leads to the most cases averted when compared to targeting non-hotspots, and random vaccination. We show that transmission potential imbalance between locations, epidemic connectivity, vaccine availability, and timing all play a large role in shaping impact of different targeting strategies, and that targeting vaccination at hotspots is only preferable in a limited number of scenarios. Finally, we address one aspect of heterogeneity in cholera transmission that is driven in part by the host response - the incubation period. We reviewed the literature for data pertaining to the incubation period of cholera and conducted a pooled analysis to estimate its distribution. The incubation period did not differ by a clinically significant margin between strains (except O1 El Tor Ogawa). We estimate the median incubation period of toxigenic cholera to be 1.4 days (95% CI, 1.3-1.6). Five percent of cholera cases will develop symptoms by 0.5 days (95% CI 0.4-0.5), and 95% by 4.4 days (95% CI 3.9-5.0) after infection. These estimates can serve as useful tools in outbreak investigations, disease control policy, and computational models of disease transmission. Cholera vaccine is likely to be used more in the coming years both reactively in epidemic settings, and proactively in highly endemic settings. More in-depth case studies and theoretical analyses are needed in the future to understand differences in transmission dynamics around the globe, and how locale specific attributes may shape the outcome of vaccination campaigns. In this dissertation we demonstrate the importance of understanding local disease transmission dynamics in designing a vaccination campaigns and provide new evidence decision makers to consider when allocating limited oral cholera vaccine