Attributing Ethiopian animal health losses to high-level causes using expert elicitation

Andrew Larkins; Wudu Temesgen; Gemma Chaters; Carlotta Di Bari; Stephen Kwok; Theo Knight-Jones; Jonathan Rushton; Mieghan Bruce

doi:10.1016/j.prevetmed.2023.106077

Attributing Ethiopian animal health losses to high-level causes using expert elicitation

Prev Vet Med. 2023 Dec:221:106077. doi: 10.1016/j.prevetmed.2023.106077. Epub 2023 Nov 11.

Authors

Andrew Larkins¹, Wudu Temesgen², Gemma Chaters³, Carlotta Di Bari⁴, Stephen Kwok¹, Theo Knight-Jones⁵, Jonathan Rushton³, Mieghan Bruce⁶

Affiliations

¹ School of Veterinary Medicine, Harry Butler Institute, Murdoch University, 90 South Street, Murdoch, Western Australia 6150, Australia.
² International Livestock Research Institute, PO Box 5689, Addis Ababa, Ethiopia; University of Gondar, Department of Veterinary Epidemiology and Public Health, PO Box 196, Gondar, Ethiopia.
³ Department of Livestock and One Health, Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, 146 Brownlow Hill, Liverpool L3 5RF, United Kingdom.
⁴ Department of Epidemiology and Public Health, Sciensano, Rue Juliette Wytsman 14, Brussels 1050, Belgium.
⁵ International Livestock Research Institute, PO Box 5689, Addis Ababa, Ethiopia.
⁶ School of Veterinary Medicine, Harry Butler Institute, Murdoch University, 90 South Street, Murdoch, Western Australia 6150, Australia. Electronic address: Mieghan.Bruce@murdoch.edu.au.

PMID: 37976968
DOI: 10.1016/j.prevetmed.2023.106077

Abstract

The Global Burden of Animal Diseases programme is currently working to estimate the burden of animal health loss in Ethiopia. As part of this work, structured expert elicitation has been trialled to attribute the proportion of animal health losses due to three independent and exhaustive high-level causes (infectious, non-infectious, and external). Separate in-person workshops were conducted with eight cattle, nine small ruminant, and eight chicken experts. Following the Investigate-Discuss-Estimate-Aggregate protocol for structured expert elicitation, estimates were obtained for the proportion of animal health loss due to high-level causes in different combinations of health loss, species, age-sex class, and production system. Three-point questions were used to inform beta-pert distributions and capture uncertainty in estimates. Individual expert estimates were aggregated by quantile mean to produce average distributions. Random samples from these average distributions estimated that infectious causes inflict the highest proportion of health loss in Ethiopia, with at least 40 % of health losses estimated to be due to infectious causes in all categories. This study provides a rapid, simple, and engaging method to attribute the burden of animal health loss at a high-level. Results are informative, however will become increasingly useful once they can be compared with results from more sophisticated, data-driven models.

Keywords: Animal health economics; Burden of animal disease; Ethiopia; Livestock.

MeSH terms

Animal Diseases*
Animals
Cattle
Cattle Diseases* / epidemiology
Cattle Diseases* / etiology
Ethiopia / epidemiology
Humans
Uncertainty