Abstract

Outbreaks of shiga-toxigenic Escherichia coli (STEC) O157 have been associated with human–animal contact in public settings, including open farms, State and County Fair livestock exhibits, and petting zoos. These outbreaks have sickened hundreds of people, sometimes severely, and caused at least two deaths in the past 5 years. Human–animal contact settings vary in hygiene and sanitation practices, the degree of supervision, the extent of animal contact permitted, and in facility design. Recent surveys have shown that livestock on farms and at fairs often shed both STEC O157 and Salmonella spp. in their feces. Our goal was to estimate American Zoo and Aquarium Association (AZA)-institutional and individual animal fecal prevalence rates for both pathogens. We hypothesized that the standardized conditions and generally higher levels of hygiene in AZA contact settings would result in lower pathogen prevalence compared to animals in production or fair settings.

American Zoo and Aquarium Association zoos with human-animal contact areas (such as children’s zoos) were recruited to participate voluntarily and confidentially. Feces collected in the summer of 2003 and 2004 from animals in contact areas, were cultured for Salmonella spp. and STEC O157. Thirty-six zoos provided feces from 997 animals, including 526 goats, 192 sheep, 59 equids, 49 cattle, 45 pigs, 33 deer, 26 llamids, 17 birds, 16 rabbits, 10 rodents, 5 tortoises, and 3 carnivores. STEC O157 was isolated from one quarantined bovid. Salmonella spp. was isolated at four zoos from three goats, one horse, one bovine and one giraffe. STEC O157 or Salmonella spp. was isolated from an animal in five of 36 contact areas (13.9%). Fecal prevalence of STEC O157 (1/997=0.1%) and Salmonella spp. (6/997=0.6%) was very low in animals connected with AZA contact areas, both in absolute terms and relative to the high prevalence (often >25%) for these pathogens in farm or fair setting. Contact areas at AZA zoos appear to present a low zoonotic risk to human visitors compared to other settings where human–animal contact may occur. Understanding the basis for this low prevalence may have future application towards lowering enteric zoonotic bacteria prevalence in animals in farm settings.