Inter- and Intra-individual Agreement of a Scoring Model for Pododermatitis in Greater Flamingos (Phoenicopterus roseus)
American Association of Zoo Veterinarians Conference 2019
Joanna K. Webb1,2, DVM; Krista A. Keller1,2, DVM, DACZM; Kenneth R. Welle1, DVM, DABVP (Avian Practice); Peter Burvenich3; Grace Mitchell3; Samantha Bradley1,2; Eliana Foltin1,2; Matthew C. Allender1,2, MS, DVM, PhD, DACZM
1Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL, USA; 2Wildlife Epidemiology Laboratory, College of Veterinary Medicine, University of Illinois, Urbana, IL, USA; 3Miller Park Zoo, Bloomington, IL, USA
Pododermatitis is a leading cause of morbidity in managed Phoenicopteridae. The individual feet of 50 greater flamingos (Phoenicopterus roseus) were photographed in fall 2018 while being managed primarily in an outdoor enclosure and in December 2018 after being managed primarily within indoor holding for 1.5 mo. Each randomized image (n=96) was scored based upon a previously described pododermatitis classification rubric that assigns a 0–2 score for hyperkeratosis, fissures, nodules and papillomatous lesions.1 In addition, a subjective overall pododermatitis score (none, mild, moderate, severe) was assigned. Scoring was performed by 7 blinded individuals: 3 specialists, 2 animal care staff (ACS) and 2 veterinary students. Inter-individual reliability and agreement were evaluated using Krippendorff’s alpha and weighted kappa, respectively, to evaluate the scoring model. Randomization of images was repeated, and each image scored again by specialists at least 1 mo later to evaluate intra-individual agreement. Overall, the scoring of fissures exhibited good intra- and inter-individual reliability and agreement; there was reliable agreement between specialists (α=0.782) and between all individuals (α=0.807), good agreement between specialists and ACS (κ=0.853) and good or very good intra-individual agreement for all of the specialists. Scoring of the other categories and the subjective score resulted in variable reliability and agreement. Based upon results presented here, it is recommended that modification of the current scoring system be performed to provide a classification scheme that allows for repeated measures of lesions, allowing for assessment of alterations in lesions.
1. Nielsen AM, Nielsen SS, Kin CE, Bertelsen MF. Classification and prevalence of foot lesions in captive flamingos (Phoenicopteridae). J Zoo Wildl Med. 2010;41(1):44–48.