Refinement and partial validation of the UNESP-Botucatu multidimensional composite pain scale for assessing postoperative pain in horses

Abstract

Background: Quantification of pain plays a vital role in the diagnosis and management of pain in animals. In order to refine and validate an acute pain scale for horses a prospective, randomized, blinded study was conducted. Twenty-four client owned adult horses were recruited and allocated to one of four following groups: anaesthesia only (GA); pre-emptive analgesia and anaesthesia (GAA,); anaesthesia, castration and postoperative analgesia (GC); or pre-emptive analgesia, anaesthesia and castration (GCA). One investigator, unaware of the treatment group, assessed all horses at time-points before and after intervention and completed the pain scale. Videos were also obtained at these time-points and were evaluated by a further four blinded evaluators who also completed the scale. The data were used to investigate the relevance, specificity, criterion validity and inter-and intra-observer reliability of each item on the pain scale, and to evaluate construct validity and responsiveness of the scale.Results: Construct validity was demonstrated by the observed differences in scores between the groups, four hours after anaesthetic recovery and before administration of systemic analgesia in the GC group. Inter-and intra-observer reliability for the items was only satisfactory. Subsequently the pain scale was refined, based on results for relevance, specificity and total item correlation.Conclusions: Scale refinement and exclusion of items that did not meet predefined requirements generated a selection of relevant pain behaviours in horses. After further validation for reliability, these may be used to evaluate pain under clinical and experimental conditions.