The grimace scale as a method of acute pain assessment for neonatal livestock

University of Minnesota M.S. thesis. August 2020. Major: Animal Sciences. Advisor: Yuzhi Li. 1 computer file (PDF); xiii, 125 pages.Neonatal livestock animals commonly undergo management procedures that induce acute pain and have negative implications on their well-being. This research consists of two studies, one focused on piglets and one on goat kids. The first study was designed to assess behavior of male piglets as an indicator of pain before, during, and after surgical castration. The four objectives of this study were: 1) To assess acute pain during castration through behavioral indicators, 2) To evaluate the utility of the Piglet Grimace Scale (PGS) to detect acute pain, and 3) To refine the current method of data collection for the PGS using photogrammetry and 3D landmark-based geometric morphometrics. The third objective had two sub-objectives: a) To determine if photogrammetry and 3D landmark-based geometric morphometrics can obtain clear, analyzable images of piglet faces and b) To evaluate 3D landmark-based geometric morphometrics to quantify changes in piglet facial shape. Eighty-eight male piglets were randomly allocated to one of two treatments: surgical castration (C; n=43) and sham-castration (S; n=45). Within 24 hours after birth, identical procedures were followed for both treatment groups, except sham-castrated piglets were not castrated. For objective 1, struggle behavior (curl ups, leg kicks, and body flailing) and vocalization parameters (duration and peak frequency) during the castration (or sham-castration) period were recorded and analyzed. For objectives 2-4, photographs of piglets were taken at four time points using 11 cameras mounted on a photogrammetry rig: baseline (T1 - immediately before castration), post-castration (T2 - immediately after castration), 4 h post-castration (T3), and 22 h post-castration (T4). Four trained raters scored the piglet images using five facial action units (FAUs): orbital tightening, ear position, temporal tension, lip contraction, and nose bulge/cheek tension. 3D facial models were generated for each piglet at each time points using the Agisoft Metashape software. Landmarks corresponding to 3 facial action units (FAUs) (orbital tightening, lip contraction, and nose bulge/cheek tension) were placed onto the models using the Markers feature for geometric morphometric analysis. Data were analyzed using the Glimmix, Mixed, and Genmod procedures of SAS software, and Generalized Procrustes Analysis (GPA) procedure of Morpheus software. During castration, castrated piglets kicked more frequently than sham-castrated piglets (28.8 vs. 21.3 kicks/min, SE=0.09; P=0.02). Additionally, 51.2% of castrated piglets displayed body flailing, whereas only 4.4% of sham piglets displayed the same behavior (P=0.03). Castrated piglets also responded with more high frequency (≥1,000 Hz) calls than sham-castrated piglets (23.6 vs. 18.6 calls/min, SE=0.26; P=0.04) and high frequency calls tended to be of longer duration for castrated piglets (0.45 vs. 0.27 sec/call, SE=0.04; P=0.08). These results indicate that surgical castration increased the frequency of leg kicks, body flailing, and high frequency calls compared to sham-castration, suggesting these may be useful behavioral indicators of acute pain in piglets. The reliability of the PGS was tested by evaluating the agreement among the four raters. The intra-class correlation coefficient (ICC) for orbital tightening, ear position, nose bulge/cheek tension, temporal tension, and lip contraction were 0.68, 0.67, 0.54, 0.40, and 0.28, respectively. For all time points (T1-T4), the odds of castrated to sham-castrated piglets for all FAUs did not differ from 1, suggesting that castration did not change any FAU. Moreover, geometric morphometric analysis did not discriminate differences between the treatment groups or among time points (all P > 0.10). These results suggest orbital tightening and ear position are more reliable than all other FAUs for a PGS, as indicated by the ICCs. However, neither the PGS nor 3D landmark-based geometric morphometrics were sensitive enough to detect pain in piglets post-castration in the current study. The objective of the second study was to develop a Goat Kid Grimace Scale (GKGS) as a novel method of pain assessment following disbudding. Goat kids (n=42) of mixed sex and breed between the age of 2-15 days were randomly assigned to one of seven treatment groups via block randomization (6 blocks, 7 kids per block): 1) simulated disbudding (SHAM), or thermal disbudding after administration of 2) 0.05 mg/kg IM xylazine (X), 3) 1 mg/kg oral meloxicam (M), 4) 4 mg/kg SQ buffered lidocaine (L) 5). xylazine + buffered lidocaine (XL), 6) xylazine + oral meloxicam (XM), and 7) xylazine + oral meloxicam + lidocaine (XML). All pain agents were administered 20 min prior to disbudding. Photographs of kids’ faces were taken at six time points: baseline (T1 – before administration of pain agents), pre-blood collection (T2), pre-disbudding (T3), post-disbudding (T4), and directly before (T5) and after (T6) mechanical nociceptive testing (4 h post-disbudding). Data were analyzed using the Mixed and Genmod procedure of SAS software. Four trained raters scored the kid photographs over two sessions using four FAUs: orbital tightening, ear position, lip tightening, and nostril dilation. The ICCs were as follows: orbital tightening (0.79 & 0.84), ear position (0.67 & 0.61), nostril dilation (0.45 & 0.56), and lip tightening (0.45 & 0.56) for the first and second session, respectively. Prior to disbudding (T1-T3), the odds of X, XL, XM, and XML scoring in a higher grimace category for orbital tightening were 79.1, 164.80, 128.0, and 86.0 to SHAM, respectively (all P < 0.001). For all times points (T1-T6), the odds of X, XL, XM, and XML scoring in a higher grimace category for orbital tightening were 43.5, 89.0, 82.8, and 71.2 to SHAM (all P < 0.001), respectively. The odds of X, M, L, XL, XM, XML scoring in a higher grimace category for ear position were 23.1, 5.1, 12.2, 23.6, 32.1, and 45.9 to SHAM (all P < 0.05), respectively. The odds of X, L, XL, and XML scoring a higher grimace category for lip tightening were 8.3, 12.3, 15.9, and 16.0 to SHAM (all P < 0.05), respectively. The odds of XM scoring in a higher grimace category for nostril dilation was 5.89 to SHAM (P = 0.0005). Results suggest that orbital tightening and ear position may be promising FAUs for a GKGS. Application of some pain mitigation treatments (X, XL, XM, and XML) alone induced a change in orbital tightening, while all FAUs changed in response to pain mitigation treatments (X, L, XL, XM, and XML) together with disbudding