Estimating Valley Fever Endemicity with Environmental Parameters

Valley Fever is a disease caused and transmitted by the fungi Coccidioides immitis and Coccidioides posadasii. The disease affects the lungs and is endemic in arid regions within the southwest United States. Although studies have been conducted to study disease severity and quantify risk factors, few have focused on creating predictive models aimed at forecasting case occurrences and evolving endemicity of the disease. Valley Fever is not transmitted from person to person, as virtually all transmission occurs from the environment to the host. Thus, we can hypothesize that specific environmental factors and conditions will have measurable impacts on the spread and prevalence of the disease. Using publicly available data from the last two decades, we have developed multiple linear regression, decision trees, and random forest models using parameters such as air quality, precipitation, wildfires, and earthquakes to provide monthly-level predictions of the disease rate across six different counties in Central California. Our goal is to inform future predictions of case occurrences within California's Central Valley and to identify possible parameters based on nonhuman and human that can serve as the basis for future studies. Our findings show that certain parameters, such as those related to fire and air quality, seem to show promise in helping to predict disease cases.