Validation of a Facial Expression based Pain Scale (FEPS) for objective pain assessment of donkeys with chronic painful diseases.

Abstract

Over the last decades animal welfare has become more important. In order to assess a donkeys welfare, we would like to develop an objective way to assess chronic pain in donkeys. Facial expression based pain scales have been developed for horses. Because of the differences between horses and donkeys, they need some adjustment to be useful for donkeys. The aim of this study is to validate the Facial Expression Pain Scale for donkeys (FEPSdonkey) for chronic pain, which was already previously investigated for validation for acute pain. A total of 77 donkeys were observed by 2 observers. The 38 patient donkeys were all previously diagnosed with a chronic painful disease by the veterinarians from the Donkey Sanctuary in England. The 39 healthy control donkeys were matched and selected by the observers. The donkeys were scored twice a day for 3 consecutive days. Furthermore a videotape of the donkeys faces during locomotion was made in order to assess facial expressions during locomotion. A survey was developed and filled out by the grooms to get a better impression of the donkeys’ overall health. A high interobserver agreement was found between observer 1 and 2 (Intraclass Correlation Coefficient (ICC) = 0,96 with p-value <0,001). There was a significant difference between patient donkey scores and control donkey scores with the FEPSdonkey (p<0,001). Overall sensitivity of 80,6% and specificity of 56,3% was found. There was no correlation between the FEPSdonkey score and the survey score and a weak correlation between the FEPSdonkey score and the Visual Analogue Scale (VAS) score (ICC = 0,56 with p<0,01). There was a good interand intra-observer reliability for scoring the videos with FEPSdonkey (ICC respectively 0,69 and 0,76 with p-value <0,001), however no significant difference was found between patients and controls (p-value = 0,19). FEPSdonkey proved to be reliable and clinically applicable to use for recognising chronic pain in donkeys. A larger amount of donkeys used in a future study is needed to further validate pain assessments. The survey and the FEPSdonkey scoring during locomotion did not prove to be clinically applicable in this set-up. Further research is needed to develop a survey that can aid in assessment of donkeys with possible chronic pain.