RESUMO
A Q fever epidemic occurred in 2013 in a small military residential area in Cayenne, French Guiana. A retrospective cohort study was conducted to identify Q fever risk factors. Confirmed acute Q fever case was defined as positive serology (IgMâ¯≥â¯50 and phase II IgGâ¯≥â¯200) and/or positive qPCR on serum or blood. In addition, wild mammals were captured at the study site and tested by serology and real-time PCR performed on blood, vaginal swabs and ticks. The attack rate was 20 percent (11/54). All the cases were symptomatic with fever >38.5⯰C and community-acquired pneumonia for four cases. Log binomial multivariate models identified two independent risk factors associated with Q fever: to clean the house (RRaâ¯=â¯7.5 CI95% [1.03-55.3]) and to carry a three-toed sloth in arms (RRaâ¯=â¯2.6 CI95% [1.1-5.8]). Eighteen marsupial individuals were captured, all PCRs were negative but 17% (3/18) had a positive serology. Another study conducted after the epidemic found only one (1/4) three-tooth sloth (Bradypus tridactylus) with feces highly infectious for C. burnetii MST17. The same strain C. burnetii genotype 17 has been laboratory- confirmed in this mammal and in human cases. These results support the implication of three-toed-sloth in this epidemic. Human contamination mainly occurs through inhalation of infectious aerosols as suggested by high relative risk associated with house cleaning activities and pulmonary forms of the disease, and through direct contact with three- toed-sloth. Positive serological results among marsupials confirm wildlife exposure and suggest a more complex sylvatic transmission cycle among wild mammals.