Validation of the Portuguese Version of the Short-Form Glasgow Composite Measure Pain Scale (CMPS-SF) According to COSMIN and GRADE Guidelines.

We aimed to validate the CMPS-SF according to COSMIN and GRADE guidelines. Four trained evaluators assessed 208 videos (pre-operative-M1, peak of pain-M2, 1 h after the peak of pain and analgesia (rescue)-M3, and 24 h post-extubation-M4) of 52 dogs, divided into negative controls (n = 10), soft tissue surgeries (n = 22), and orthopedic surgeries (n = 20). The videos were randomized and blinded as to when they were filmed, and were evaluated in two stages, 21 days apart. According to confirmatory analysis, the CMPS-SF is a unidimensional scale. Intra-observer reliability was between 0.80 and 0.99 and inter-observer reliability between 0.73 and 0.86. Criterion validity was confirmed by the correlation between the CMPS-SF and other unidimensional scales (≥0.7). The differences between the scores were M2 ≥ M3 > M4 > M1 (responsiveness), and the scale presented construct validity (higher postoperative pain scores in dogs undergoing surgery versus control). Internal consistency was 0.7 (Cronbach's α) and 0.77 (McDonald's ω), and the item-total correlation was between 0.3 and 0.7, except for "A(ii)-Attention to wound". Specificity and sensitivity were 78-87% and 74-83%, respectively. The cut-off point for rescue analgesia was ≥5 or ≥4 excluding item B(iii) mobility, and the GRADE classification was high, confirming the validity of the scale.

Deep learning for video-based automated pain recognition in rabbits.

Despite the wide range of uses of rabbits (Oryctolagus cuniculus) as experimental models for pain, as well as their increasing popularity as pets, pain assessment in rabbits is understudied. This study is the first to address automated detection of acute postoperative pain in rabbits. Using a dataset of video footage of n = 28 rabbits before (no pain) and after surgery (pain), we present an AI model for pain recognition using both the facial area and the body posture and reaching accuracy of above 87%. We apply a combination of 1 sec interval sampling with the Grayscale Short-Term stacking (GrayST) to incorporate temporal information for video classification at frame level and a frame selection technique to better exploit the availability of video data.

Explainable automated pain recognition in cats.

Manual tools for pain assessment from facial expressions have been suggested and validated for several animal species. However, facial expression analysis performed by humans is prone to subjectivity and bias, and in many cases also requires special expertise and training. This has led to an increasing body of work on automated pain recognition, which has been addressed for several species, including cats. Even for experts, cats are a notoriously challenging species for pain assessment. A previous study compared two approaches to automated 'pain'/'no pain' classification from cat facial images: a deep learning approach, and an approach based on manually annotated geometric landmarks, reaching comparable accuracy results. However, the study included a very homogeneous dataset of cats and thus further research to study generalizability of pain recognition to more realistic settings is required. This study addresses the question of whether AI models can classify 'pain'/'no pain' in cats in a more realistic (multi-breed, multi-sex) setting using a more heterogeneous and thus potentially 'noisy' dataset of 84 client-owned cats. Cats were a convenience sample presented to the Department of Small Animal Medicine and Surgery of the University of Veterinary Medicine Hannover and included individuals of different breeds, ages, sex, and with varying medical conditions/medical histories. Cats were scored by veterinary experts using the Glasgow composite measure pain scale in combination with the well-documented and comprehensive clinical history of those patients; the scoring was then used for training AI models using two different approaches. We show that in this context the landmark-based approach performs better, reaching accuracy above 77% in pain detection as opposed to only above 65% reached by the deep learning approach. Furthermore, we investigated the explainability of such machine recognition in terms of identifying facial features that are important for the machine, revealing that the region of nose and mouth seems more important for machine pain classification, while the region of ears is less important, with these findings being consistent across the models and techniques studied here.

Effects of Human Observer Presence on Pain Assessment Using Facial Expressions in Rabbits.

The goal of this study was to evaluate the effect of a human observer on Rabbit Grimace Scale (RbtGS) scores. The study scored video footage taken of 28 rabbits before and after orthopedic surgery, as follows: 24 h before surgery ( baseline), 1 h after surgery ( pain), 3 h after analgesia administration ( analgesia), and 24 h after surgery ( 24h) in the presence and absence of an observer. Videos were assessed twice in random order by 3 evaluators who were blind to the collection time and the presence or absence of an observer. Responses to pain and analgesia were evaluated by comparing the 4 time points using the Friedman test, followed by the Dunn test. The influence of the presence or absence of the observer at each time point was evaluated using the Wilcoxon test. Intra- and interrater reliabilities were estimated using the intraclass correlation coefficient. The scale was responsive to pain, as the scores increased after surgery and had decreased by 24 h after surgery. The presence of the observer reduced significantly the RbtGS scores (median and range) at pain (present, 0.75, 0 to 1.75; absent, 1, 0 to 2) and increased the scores at baseline (present, 0.2, 0 to 2; absent, 0, 0 to 2) and 24h after surgery (present, 0.33, 0 to 1.75; absent, 0.2, 0 to 1.5). The intrarater reliability was good (0.69) to very good (0.82) and interrater reliability was moderate (0.49) to good (0.67). Thus, the RbtGS appeared to detect pain when scored from video footage of rabbits before and after orthopedic surgery. In the presence of the observer, the pain scores were underestimated at the time considered to be associated with the greatest pain and overestimated at the times of little or no pain.

Reliability and Validity of UNESP-Botucatu Cattle Pain Scale and Cow Pain Scale in Bos taurus and Bos indicus Bulls to Assess Postoperative Pain of Surgical Orchiectomy.

Pain assessment guides decision-making in pain management and improves animal welfare. We aimed to investigate the reliability and validity of the UNESP-Botucatu cattle pain scale (UCAPS) and the cow pain scale (CPS) for postoperative pain assessment in Bos taurus (Angus) and Bos indicus (Nelore) bulls after castration. METHODS: Ten Nelore and nine Angus bulls were anaesthetised with xylazine-ketamine-diazepam-isoflurane-flunixin meglumine. Three-minute videos were recorded at -48 h, preoperative, after surgery, after rescue analgesia and at 24 h. Two evaluators assessed 95 randomised videos twice one month apart. RESULTS: There were no significant differences in the pain scores between breeds. Intra and inter-rater reliability varied from good (>0.70) to very good (>0.81) for all scales. The criterion validity showed a strong correlation (0.76-0.78) between the numerical rating scale and VAS versus UCAPS and CPS, and between UCAPS and CPS (0.76). The UCAPS and CPS were responsive; all items and total scores increased after surgery. Both scales were specific (81-85%) and sensitive (82-87%). The cut-off point for rescue analgesia was >4 for UCAPS and >3 for CPS. CONCLUSIONS: The UCAPS and CPS are valid and reliable to assess postoperative pain in Bos taurus and Bos indicus bulls.

Automated recognition of pain in cats.

Facial expressions in non-human animals are closely linked to their internal affective states, with the majority of empirical work focusing on facial shape changes associated with pain. However, existing tools for facial expression analysis are prone to human subjectivity and bias, and in many cases also require special expertise and training. This paper presents the first comparative study of two different paths towards automatizing pain recognition in facial images of domestic short haired cats (n = 29), captured during ovariohysterectomy at different time points corresponding to varying intensities of pain. One approach is based on convolutional neural networks (ResNet50), while the other-on machine learning models based on geometric landmarks analysis inspired by species specific Facial Action Coding Systems (i.e. catFACS). Both types of approaches reach comparable accuracy of above 72%, indicating their potential usefulness as a basis for automating cat pain detection from images.

Multilingual validation of the short form of the Unesp-Botucatu Feline Pain Scale (UFEPS-SF).

Background: Pain is the leading cause of animal suffering, hence the importance of validated tools to ensure its appropriate evaluation and treatment. We aimed to test the psychometric properties of the short form of the Unesp-Botucatu Feline Pain Scale (UFEPS-SF) in eight languages. Methods: The original scale was condensed from ten to four items. The content validation was performed by five specialists in veterinary anesthesia and analgesia. The English version of the scale was translated and back-translated into Chinese, French, German, Italian, Japanese, Portuguese and Spanish by fluent English and native speaker translators. Videos of the perioperative period of 30 cats submitted to ovariohysterectomy (preoperative, after surgery, after rescue analgesia and 24 h after surgery) were randomly evaluated twice (one-month interval) by one evaluator for each language unaware of the pain condition. After watching each video, the evaluators scored the unidimensional, UFEPS-SF and Glasgow composite multidimensional feline pain scales. Statistical analyses were carried out using R software for intra and interobserver reliability, principal component analysis, criteria concurrent and predictive validities, construct validity, item-total correlation, internal consistency, specificity, sensitivity, the definition of the intervention score for rescue analgesia and diagnostic uncertainty zone, according to the receiver operating characteristic (ROC) curve. Results: UFEPS-SF intra- and inter-observer reliability were ≥0.92 and 0.84, respectively, for all observers. According to the principal component analysis, UFEPS-SF is a unidimensional scale. Concurrent criterion validity was confirmed by the high correlation between UFEPS-SF and all other scales (≥0.9). The total score and all items of UFEPS-SF increased after surgery (pain), decreased to baseline after analgesia and were intermediate at 24 h after surgery (moderate pain), confirming responsiveness and construct validity. Item total correlation of each item (0.68-0.83) confirmed that the items contributed homogeneously to the total score. Internal consistency was excellent (≥0.9) for all items. Both specificity (baseline) and sensitivity (after surgery) based on the Youden index was 99% (97-100%). The suggestive cut-off score for the administration of analgesia according to the ROC curve was ≥4 out of 12. The diagnostic uncertainty zone ranged from 3 to 4. The area under the curve of 0.99 indicated excellent discriminatory capacity of UFEPS-SF. Conclusions: The UFEPS-SF and its items, assessed by experienced evaluators, demonstrated very good repeatability and reproducibility, content, criterion and construct validities, item-total correlation, internal consistency, excellent sensitivity and specificity and a cut-off point indicating the need for rescue analgesia in Chinese, French, English, German, Italian, Japanese, Portuguese and Spanish.

Clinical validation of the short and long UNESP-Botucatu scales for feline pain assessment.

BACKGROUND: The UNESP-Botucatu multidimensional feline pain assessment scale (UFEPS) is a valid and reliable instrument for acute pain assessment in cats. However, its limitations are that responsiveness was not tested using a negative control group, it was validated only for ovariohysterectomy, and it can be time-consuming. We aimed to evaluate the construct and criterion validity, reliability, sensitivity, and specificity of the UFEPS and its novel short form (SF) in various clinical or painful surgical conditions. METHODS: Ten client-owned healthy controls (CG) and 40 client-owned cats requiring pain management for clinical or surgical care (20 clinical and 20 surgery group (12 orthopedic and eight soft tissue surgeries) were recruited. Three evaluators assessed pain, in real-time, in clinical cases before and 20 min after rescue analgesia and in surgical cases before and up to 6.5 hours postoperatively, by using the visual analog, numerical ratio, and a simple descriptive scale, in this order, followed by the UFEPS-SF, UFEPS and Glasgow multidimensional feline pain (Glasgow CMPS-Feline) in random order. For the surgical group, rescue analgesia (methadone 0.2 mg/kg IM or IV and/or dipyrone 12.5 mg/kg IV) was performed when the UFEPS-SF score was ≥4 or exceptionally according to clinical judgement. If a third interventional analgesia was required, methadone (0.1-0.2 mg/kg IM) and ketamine (1 mg/kg IM) were administered. For the clinical group, all cats received rescue analgesia (methadone 0.1-0.2 mg/kg IM or IV or nalbuphine 0.5 mg/kg IM or IV), according to the clinician in charge, regardless of pain scores. Construct (1-comparison of scores in cats undergoing pain vs pain-free control cats by unpaired Wilcoxon-test and 2-responsiveness to analgesia by paired Wilcoxon test) and concurrent criterion validity (Spearman correlation of the total score among scales), inter-rater reliability, specificity and sensitivity were calculated for each scale (α = 0.05). RESULTS: Reliability ranged between moderate and good for the UFEPS and UFEPS-SF (confidence intervals of intraclass coefficients = 0.73-0.86 and 0.63-0.82 respectively). The Spearman correlation between UFEPS and UFEPS-SF was 0.85, and their correlation with Glasgow CMPS-Feline was strong (0.79 and 0.78 respectively), confirming criterion validity. All scales showed construct validity or responsiveness (higher scores of cats with clinical and postoperative pain vs healthy controls, and the reduction in scores after rescue analgesia). The sensitivity and specificity of the UFEPS, UFEPS-SF and Glasgow CMPS-Feline were moderate (sensitivity 83.25, 78.60% and 74.28%; specificity 72.00, 84.67 and 70.00%, respectively). CONCLUSIONS: Both UFEPS and UFEPS-SF showed appropriate concurrent validity, responsiveness, reliability, sensitivity, and specificity for feline acute pain assessment in cats with various clinical and orthopedic and soft tissue surgical conditions.

Post-operative pain behaviour associated with surgical castration in donkeys (Equus asinus).

BACKGROUND: Recognising pain in donkeys is challenging because they are stoic. OBJECTIVES: To identify the responses of donkeys before and after surgical pain. STUDY DESIGN: Prospective, short-term longitudinal pre- and post-intervention observations. METHODS: Forty adult donkeys underwent surgical castration after sedation with intravenous (IV) xylazine, induction with guaiphenesin/thiopental IV and maintenance of anaesthesia with isoflurane and local anaesthetic blockade. Four hours after recovery from anaesthesia, flunixin meglumine 1.1 mg/kg, dipyrone 10 mg/kg and morphine 0.2 mg/kg IV were administered. Behavioural responses exhibited by the animals housed in individual stalls were recorded in four 30-min videos: before castration (M0), and 3.5-4.0 hours (M1), 5.5-6.0 hours (M2) and 23.5-24.0 hours after recovery from anaesthesia (M3). To exclude the influence of insects, the behaviour of six apparently pain-free donkeys was compared with and without the presence of faeces and urine in the stall. RESULTS: When compared with presurgical baseline behaviours (M0), after surgery (M1) donkeys raised their pelvic limbs more (P = .003). When compared with M1, after analgesia (M2), the median frequencies of ear movements (44 vs 16; P < .001), head shaking (7 vs 1; P < .001), head turning (5 vs 0; P < .001) and lifting of the both limbs (7 vs 0; P = .008) decreased; feeding (0 vs 29; P < .001) and water intake (0 vs 0, range 0-1 vs 0-7; P = .05) increased. The dirty stall increased tail (53 vs 80; P = .03), head (16 vs 30; P = .03) and ear movements (50 vs 78; P = .04). MAIN LIMITATIONS: The dirty stall and presence of insects possibly contributed to the expression of behaviours unrelated to pain. CONCLUSION: Lifting the pelvic limbs was the only specific pain behaviour after castration in donkeys. Analgesia restored appetite and water intake and reduced the frequency of head shaking and turning, ear movement and lifting the limbs. Tail, head and ear movements are unspecific responses related both to pain and a dirty stall, and are confounding factors when pain is assessed in donkeys in the presence of insects.

Postoperative Analgesic Effects of Laserpuncture and Meloxicam in Bitches Submitted to Ovariohysterectomy.

BACKGROUND: This prospective, randomised and blind study investigated the efficacy of laserpuncture for postoperative pain management in dogs. METHOD: Sixteen bitches were sedated with acepromazine and randomly treated before ovariohysterectomy with meloxicam 0.2 mg·kg-1 intramuscular or laserpuncture (wavelength 904 mm, frequency 124 Hz, potency 10 Joules, 100 s in each acupoint). Anaesthesia was performed with propofol, isoflurane/O2, and fentanyl. The Glasgow Composite Measure Pain Scale (GCMPS) and Dynamic Interactive Visual Analog Scale (DIVAS) were used to evaluate postoperative pain before and for 24 h after surgery. Morphine was administrated as rescue analgesia when pain scores were ≥3.33 (GCMPS). Differences between treatments, time points, and amount of rescue analgesia between groups were investigated by the Mann-Whitney test and the area under the curve (AUC) for GCMPS, Friedman, and Chi-squared tests, respectively (p < 0.05). RESULTS: Dogs treated with laserpuncture presented lower GCMPS AUC for 24 h and lower GCMPS scores at 2 and 4 h postoperatively (p = 0.04). Three dogs treated with meloxicam required postoperatively rescue analgesia against none treated with laserpuncture. CONCLUSIONS: In this preliminary study, laserpuncture mitigated postoperative pain in dogs following ovariohysterectomy, and the technique is a promising adjunct to perioperative pain management in dogs undergoing soft tissue surgery.