Lameness is a behavioural indicator of pain that negatively affects dairy ruminants' health and welfare. Lameness is generally assessed by subjective methods, based on the observation of the animal's behaviour, using numerical rating scales (NRSs) - the most common scoring system - and visual analogue scales (VASs). A NRS consists of a set of different categories with descriptive definitions associated to each category. A VAS is a continuous scale, generally illustrated by a horizontal line, with descriptors as anchors on both ends of the line. Distinct drawbacks have been associated with both types of scales. NRSs have been associated with a reduced sensitivity to capture variations in lower levels of lameness that may adversely impact animals' welfare assessments. VASs are considered too subjective and associated with low user-acceptance. Recent literature on health scales has been focusing on the development of modified VASs that define equal ranges along the scales' continuum, with thresholds representing a NRS descriptor. Although good results have been reported in using these modified VASs for lameness scoring, the literature recognizes that it is paramount to test whether existing NRS descriptors are equal spaced in the VAS continuum, as well as research the extent to which lameness intensity varies for different lameness and posture signs used to define NRS descriptors. The answers to these questions are vital for the development of a new modified VAS to assess lameness in goats. Aiming to address these questions we collected and analyzed lameness scorings using individual VASs to score three lameness and posture signs (gait, head nodding and arched-back). Lameness scorings were performed through a video-based web-survey. We collected a total of 570 valid participations from respondents with different occupations and experience. Because of expected differences in the respondents' ability to assess lameness, we analyzed answers by levels of cardinal consistency. The cardinal consistency levels were designed as increasing filters of consistency between the respondents' assessment and the NRS model used by the experts to score the videos. Our results showed: (1) respondents' difficulties in recognizing and discriminating across some NRS descriptors; (2) these difficulties varied with the lameness severity and with the lameness sign; (3) gait, the basis for NRS lameness descriptors in goats, was not scored evenly spaced along the VAS continuum; (4) similar results were found for the head nodding and arched-back signs. In conclusion we suggest that the exact location of the thresholds along the VAS continuum should be reassessed, and the inclusion of different lameness and posture signs should receive further attention before new modified VASs are developed. Moreover, the use of NRSs in lameness scoring should only consider their ordinal measurement properties, therefore giving space for developing, validating and using alternative lameness scoring methods in farm animals that allow for higher measurement levels.
|Publication Title||Applied Animal Behaviour Science|
|Author Address||Animal Behaviour and Welfare Lab., Centre for Interdisciplinary Research in Animal Health, Faculdade de Medicina Veterinaria, Universidade de Lisboa, Avenida da Universidade Tecnica, Lisboa 1300-477, Portugal.email@example.com|
|Cite this work||
Researchers should cite this work as follows: