A Risk Assessment and Seir Model of Rift Valley Fever Virus in the State of Florida

Abstract

Rift Valley fever (RVF) is a mosquito-borne disease of medical and agricultural importance, affecting humans and ungulates in a disease spectrum stretching from mild, self limited illness to serious disease, abortion and death. At present, RVF is geographically restricted to Sub-Saharan Africa, Egypt and the Arabian Peninsula, however, the risk of accidental or deliberate introduction of RVF to the United States of America (USA) is real. Florida has both the mosquito vectors (Aedes and Culex mosquitoes) and hosts (cattle) to support and potentially propagate an RVF outbreak. In addition, Florida has two of the USA's busiest airports and seaports, making it a possible ingress for a successful introduction of RVF. Therefore, we undertook an investigation to establish what factors exist in the USA and Florida that would affect an introduction. We accomplished this by creating a Haddon matrix, segmenting an RVF outbreak into discrete chronological and entitative units, and analyzing each cell for risk factors. We adapted a closed-system, deterministic, metapopulation SEIR model to force it by the dynamic effects of two of the physical factors identified in the Haddon matrix; temperature and precipitation. To model an RVF outbreak in Florida, we created a one-by-four spatial model where each patch was the non-spatial model with a migratory element between adjacent patches and the four patches resembled four arbitrary quadrants of Florida in temperature and precipitation. We identified the contact rates between Aedes and cattle and life expectancies as important risk factors in determining the epidemic intensity. To investigate countermeasure effects on epidemic intensity, we ran simulations modeling adulticiding of adult mosquitoes, and culling, vaccination and restricting of movement of cattle. We discovered that only vaccination, even when population coverage is partial with a moderately efficacious vaccine, was effective at reducing transmission and mitigating disease effects on the total cattle population. However, the length of time between reaching the disease prevalence threshold for vaccination and introducing vaccination positively correlated with the epidemic intensity. In order to maximize the mitigation effects of vaccination, it must be introduced as soon as possible after events trigger its requirement.