Validation of two behaviour-based pain scales for horses with acute colic

Two behaviour-based scales for evaluating abdominal pain in horses (Equine Acute Abdominal Pain Scales, EAAPS-1 and -2) and a numerical rating scale (NRS) were compared for reliability and validity. Forty-one equine veterinarians randomly assigned into three groups were each presented a different set of 28 moving picture films randomly chosen among 36 films of horses with colic and four controls. One randomly chosen film was embedded twice within each set. The first (n = 15) and the second (n = 16) groups scored pain with the EAAPS-1 and EAPPS-2, respectively, while the third (n = 10) used the NRS. The intra-class correlations (ICC) for EAAPS-1 (0.80) and EAAPS-2 (0.76) vs. NRS (0.67) indicated superior inter-rater reliability for both EAPPS scales. The intra-rater reliability of the EAAPS-1 was superior to both the other scales (weighted κ = 0.9) vs. 0.5 and 0.7 for EAAPS-2 and NRS, respectively). The convergent validity between both EAAPS scores and the NRS scores was substantial (weighted κ = 0.64). Both EAAPS scales discriminated well between extreme groups (areas under the receiver operating characteristic curve, ROC, area under the curve, AUC, >0.9) to differentiate severe from mild pain, as judged by the NRS. Both EAAPS scales showed predictive validity comparable to NRS (AUCs for dichotomous treatment modality (none and medically treated vs. surgically treated or euthanased) and with mortality (dead vs. alive) between 0.6 and 0.7 for all three scales. EAAPS-1 was the most reliable of the three scales and both EAAPS scales demonstrated validity comparable to the NRS scale.