Interobserver Reliability of the Animal Welfare Indicators Welfare Assessment Protocol for Horses.

The research article presents a study on the reliability of the Animal Welfare Indicators (AWIN) welfare assessment protocol for horses. The authors found that there is a good level of agreement between different observers using this protocol to assess horse welfare.

Introduction and Methodology

• The motivation for this work comes from the need for standardised, objective tools to assess animal welfare, in this case, specifically for horses.
• Two trained observers conducted a total of 18 assessments on farm at the same time, observing the same animals. The aim was to determine how reliable the assessments were between different observers.
• The assessment protocol used in these observations was the Animal Welfare Indicators (AWIN) welfare assessment protocol for horses.
• The results were compared at two different scales: the individual level and farm level.

Statistical Analysis

• The researchers employed several statistical measures to verify the reliability of the observations. These measures include Cohen’s kappa (κ), weighted kappa (κ), and prevalence-adjusted, bias-adjusted kappa (PABAK) for the individual level.
• Spearman rank correlation coefficient (RS), intraclass correlation coefficient (ICC), limits of agreement (LoA), and smallest detectable change (SDC) were used for farm level analyses.
• To further evaluate the Qualitative Behaviour Assessment, the researchers used principal component analysis.

Findings of the Study

• Upon analysing the results, the study demonstrated that for most indicators, there was an acceptable to good level of agreement between different observers. This was the case for both at the individual animal level and the overall farm level.
• The notable exceptions where the interobserver reliability was not as strong were the indicators “moderate presence of tension above eye area (score 1)” and “orbital tightening (score 1)” on the Horse Grimace Scale, as well as the indicator for the presence of swollen joints.
• The authors comment on the need for more detailed descriptions for different scores for some indicators like the Body Condition Score.

Conclusions

• In general, the AWIN welfare assessment protocol for horses demonstrated good interobserver reliability, indicating that it could be a reliable tool for the objective assessment of horse welfare.
• However, certain specific indicators within the protocol do need refinement and improvements to ensure consistent scoring between different observers.