Generalizability and comparability of prevalence estimates in the wild bird literature: methodological and epidemiological considerations. (18th June 2020)

• Record Type: Journal Article
• Authors: Vogt, Nadine A.; Stevens, Christian P.G.; Pearl, David L.; Taboada, Eduardo N.; Jardine, Claire M.
• Is Part Of: Animal health research reviews. Volume 21:Number 1(2020)
• Page Start: 89
• Page End: 95
• Publication Date: 2020-06-18
• Subjects: Epidemiology, -- study design, -- wildlife, -- zoonoses; Animal health -- Periodicals; Veterinary medicine -- Periodicals; 636.089
• DOI: 10.1017/S1466252320000043
• Languages: English
• ISSNs: 1466-2523

Abstract

Abstract: Wild birds have been the focus of a great deal of research investigating the epidemiology of zoonotic bacteria and antimicrobial resistance in the environment. While enteric pathogens (e.g. Campylobacter, Salmonella, and E. coli O157:H7) and antimicrobial resistant bacteria of public health importance have been isolated from a wide variety of wild bird species, there is a considerable variation in the measured prevalence of a given microorganism from different studies. This variation may often reflect differences in certain ecological and biological factors such as feeding habits and immune status. Variation in prevalence estimates may also reflect differences in sample collection and processing methods, along with a host of epidemiological inputs related to overall study design. Because the generalizability and comparability of prevalence estimates in the wild bird literature are constrained by their methodological and epidemiological underpinnings, understanding them is crucial to the accurate interpretation of prevalence estimates. The main purpose of this review is to examine methodological and epidemiological inputs to prevalence estimates in the wild bird literature that have a major bearing on their generalizability and comparability. The inputs examined here include sample type, microbiological methods, study design, bias, sample size, definitions of prevalence outcomes and parameters, and control of clustering. The issues raised in this review suggest, among other things, that future prevalence studies of wild birds should avoid opportunistic sampling when possible, as this places significant limitations on the generalizability of prevalence data.