Inferring epicenters of vector-borne epidemics from vector biology, with an example of Chagas disease in Peru

Background/Question/Methods

Understanding the static and dynamic properties of infectious diseases is critical to curbing epidemics. The insect vector and parasitic agent of Chagas disease have become urban problems in the city of Arequipa, Peru, yet the debilitating symptoms of the chronic form of the disease are absent from hospitals in the city. We performed a cross-sectional entomologic survey for the triatomine bug vector, Triatoma infestans, in a periurban community of Arequipa, followed by a serologic survey of residents of the community for infection by the parasite agent, Trypanosoma cruzi. We use a new method, ‘Epicenter Regression', to test the hypothesis that transmission of T. cruzi is in an epidemic phase in the city and to make inference on the sites of introduction of the pathogen. We use Bayesian methods and include prior information on the spread of the disease derived from entomologic studies on T. infestans. Results/Conclusions We show that transmission of T. cruzi in a community of Arequipa occurred in a series of focal microepidemics, rather than a single epidemic, and that the lion's share of infections occurred over a brief period prior to testing. The clinical implications of our finding are crucial: drug therapy, effective early in the course of T. cruzi infection, would be beneficial to infected individuals of all ages in our study site. Our results also suggest that the absence of symptoms of Chagas disease in Arequipa is due to the long asymptomatic stage of the disease and short history of transmission.