Assessment of acute and chronic pain in working horses and donkeys in Egypt and the influence of observer training

BACKGROUND: Pain assessment can support monitoring welfare in working equids. AIMS/OBJECTIVE: To assess agreement of structured pain scores before and after training observers and to assess differences in pain scores between working equids with or without acute or chronic pain and to compare them to non-working equids. METHODS: In part 1, eleven veterinarians and nine technicians performed pain scores before and after training. Agreement with reference scores was analyzed with Intra Class Correlation (ICC) analysis. In part 2, pain assessments were performed in 96 donkeys (n = 40 fit-to-work, n = 21 acute pain, n = 35 chronic pain) and 104 horses (n = 42 fit-to-work, n = 25 acute pain, n = 37 chronic pain). RESULTS: Median agreement in pain scores significantly increased from ICC=0.66 (range:0.06-0.89) before to ICC=0.89 (range:0.44-0.93) after training for technicians (P = 0.024) while a non-significant increase from ICC=0.87 (range:0.23-0.94) to ICC=0.92 (range:0.48-0.93) was found for veterinarians (P = 0.096). For chronic pain scores, fit-to-work animals showed scores similar to those in non-working equids. For acute pain scores, fit-to-work animals showed scores higher than those of non-working equids (P < 0.001). Animals suffering from acute or chronic pain showed significantly higher scores compared to baseline (P < 0.001). Sensitivity and specificity were 91% and 73% for donkeys suffering from acute pain, 83% and 85% for donkeys suffering from chronic pain, 92% and 75% for horses suffering from acute pain and 87% and 81% for horses suffering from chronic pain. CONCLUSIONS: Training in pain assessment significantly improved accuracy of pain scoring for technicians. Pain scales can help differentiate animals suffering from pain from fit-to-work equids.