Inter-observer agreement, diagnostic sensitivity and specificity of animal-based indicators of young lamb welfare

Abstract

A scientific literature review and consensus of expert opinion used the welfare definitions provided by the Farm Animal Welfare Council (FAWC) Five Freedoms as the framework for selecting a set of animal-based indicators that were sensitive to the current on-farm welfare issues of young lambs (aged <6 weeks). Ten animal-based indicators assessed by observation -demeanour, response to stimulation, shivering, standing ability, posture, abdominal fill, body condition, lameness, eye condition and salivation were tested as part of the objective of developing valid, reliable and feasible animal-based measures of lamb welfare The indicators were independently tested on 966 young lambs from 17 sheep flocks across Northwest England and Wales during December 2008 to April 2009 by four trained observers. Inter-observer reliability was assessed using Fleiss's kappa (k), and the pair-wise agreement with an experienced, observer designated as the 'test standard observer' (TSO) was examined using Cohen's k. Latent class analysis (LCA) estimated the sensitivity (Se) and specificity (Sp) of each observer without assuming a gold standard and predicted the Se and Sp of randomly selected observers who may apply the indicators in the future. Overall, good levels of inter-observer reliability, and high levels of Sp were identified for demeanour (k 5 0.54, Se > 0.70, Sp > 0. LCA predicted that randomly selected observers had Se . 0.77 (acceptable), and Sp > 0.98 (high) for assessments of demeanour, lameness, abdominal fill posture, body condition and eye condition. The diagnostic performance of some indicators was influenced by the composition of the study population, and it would be useful to test the indicators on lambs with a greater level of outcomes associated with poor welfare. The findings presented in this paper could be applied in the selection of valid, reliable and feasible indicators used for the purposes of on-farm assessments of lamb welfare.

Animal-Based Measures to Assess the Welfare of Extensively Managed Ewes

“…On-farm welfare assessments can be used for immediate or ongoing on-farm monitoring and benchmarking by farmers and veterinarians, and to demonstrate compliance with national and international legal welfare standards and farm assurance schemes [1,2]. For welfare assessments to be effective and acceptable to all key stakeholders (i.e., industry, animals, scientists, consumers and society), they must incorporate welfare measures that are meaningful with respect to animal welfare, provide repeatable outcomes when applied by different observers and practical under farm conditions; that is they must be valid, reliable and feasible [2,3].…”

“…For welfare assessments to be effective and acceptable to all key stakeholders (i.e., industry, animals, scientists, consumers and society), they must incorporate welfare measures that are meaningful with respect to animal welfare, provide repeatable outcomes when applied by different observers and practical under farm conditions; that is they must be valid, reliable and feasible [2,3]. Welfare measures can be classified into categories that assess housing and facilities (resource-based measures), farmer policies and management strategies (management-based measures), and direct animal observations (animal-based measures) [4,5].…”

“…Previously, sheep have received considerably less attention in the development of welfare measures when compared with animals farmed intensively such as pigs, poultry and dairy cattle. Furthermore, most of the research has been conducted in European countries [10,13,15,17,18], where sheep are managed in small flocks and usually in more intensive, indoor-lambing systems [2]. The aim of the present study was to test the reliability and feasibility of some animal based measures for sheep welfare assessment.…”

Feasibility and validity of animal-based indicators for on-farm welfare assessment of thermal stress in dairy goats

“…However, our results suggest that shivering in goats may occur even at higher environmental temperatures. Although shivering has been used to detect cold stress in farm animals (Redbo et al 1996;Mejdell and Bøe 2005;Welfare Quality® 2009;Phythian et al 2013), this behaviour has been reported to occur with low prevalence in other species (e.g. a horse was found shivering in a rainy day at 5°C; Mejdell and Bøe 2005).…”

“…Mount (1979) reported that goats may shiver if they are exposed to low temperatures (i.e. −5°C), but this indicator has never been used for welfare evaluation in goats, whereas it is widely used in other species, such as pigs (Welfare Quality® 2009), horses (Mejdell and Bøe 2005), lambs (Phythian et al 2013) and cattle (Redbo et al 1996).…”