Objective assessment of acute pain in foals using a facial expression-based pain scale

Abstract

Background: Pain assessment based on facial expressions has been described in foals. Objectives: To extend previous pilot findings of the Equine Utrecht University Scale for Facial Assessment of Pain in Foals (EQUUS-FAP FOAL). Study design: Prospective blinded case–control study (known groups analysis). Methods: Video recordings (30–60 s) of n = 100 foals (38 patients and 62 controls) were collected. Videos were randomised and scored by two observers, blinded for the condition of the animals. Inter- and intra-observer agreement was tested by Intra Class Correlation (ICC) analysis. Differences between patients and controls were tested using Mann–Whitney U test, differences before and after analgesic treatment were tested with Wilcoxon signed rank test. Multilevel binomial logistic regression was performed for statistical weighting of each pain scale item. Receiver operating characteristic (ROC) analysis was performed to determine sensitivity, specificity and cut-off values. Results: EQUUS-FAP FOAL showed good inter- and intra-observer reliability (ICC coefficient of 0.85 and 0.90 respectively, p < 0.001). Patients had significantly higher pain scores (median score 7, IQR 5.3–7.5) compared with controls (median score 2.5, IQR 1–3, p < 0.001) and the patient pain scores decreased after analgesic treatment (median score 6.5, IQR 5–8 vs. median score 4.5, IQR 3.5–6.5, p < 0.05). Multilevel binomial logistic regression showed the relative importance of the individual scale items. Using the original scale, sensitivity and specificity were 88.2% (95% CI 76%–96%) and 87.9% (81%–98%) respectively, with an area under the ROC curve of 94.9% (92%–98%), based on a cut-off value of ≥4. These values remained similar with the weighted scale based on logistic regression output. Main limitations: Videos of included patients were short and acquired by a person in front of the stable. Conclusions: The current study shows that EQUUS-FAP FOAL can be effectively used to assess acute pain in young foals.