Q fever is a widely distributed, yet, neglected zoonotic disease, for which domestic ruminants are considered the main reservoirs in some countries. There are still many gaps in our knowledge of its epidemiology, and the source of sporadic cases is often not determined. In this study, we show how Q fever surveillance data in combination with information routinely collected by government agencies in Minnesota during 1997 to 2015 can be used to characterize patterns of occurrence of Q fever illnesses and detect variables potentially associated with increased human illness. Cluster analysis and Bayesian spatial regression modeling revealed the presence of areas in Southern Minnesota at higher risk of Q fever. The number of sheep flocks at the county level helped to explain the observed number of human cases, while no association with the cattle or goat population was observed. Our results provide information about the heterogeneous spatial distribution of risk of Q fever in Minnesota.
Keywords: Q fever; United States; livestock; sheep; spatiotemporal analysis; zoonoses.