Reproducibility and diagnostic accuracy of the Unesp-Botucatu Feline Pain Scale and the visual analog scale among laypeople, cat owners, students and experienced veterinarians

Abstract

OBJECTIVE: To investigate the reproducibility and diagnostic accuracy of the Unesp-Botucatu Feline Pain Scale short form (UFEPS-SF) and the Visual Analogue Scale (VAS) among untrained laypersons, cat owners, veterinary students and experienced veterinarians. STUDY DESIGN: Blind and randomized. ANIMALS: Videos of 10 female cats. METHODS: Three self-declared female and three male raters were included per group of experienced veterinarians, students, cat owners and laypersons (n = 24). Participants assessed 40 videos of female cats that underwent ovariohysterectomy: preoperative, postoperative, post-rescue analgesia, and 24 hours postoperatively. A multilevel generalized linear model followed by Bonferroni post hoc test was used to evaluate differences in UFEPS-SF scores between groups. Interobserver agreement was calculated by intraclass correlation coefficient (ICC) between pairs of observers within each group. Area under the curve (AUC) for the receiver operating characteristic curves, specificity (no pain) and sensitivity (pain) were calculated. RESULTS: ICC for UFEPS-SF was higher for veterinarians and students than for laypersons and cat owners, although all results were very good (≥0.81). Reliability of the VAS was very good for experienced observers only. Females assigned higher median scores than males only at 24 hours postoperatively. UFEPS-SF specificity [confidence interval (CI)] and sensitivity (CI) were 0.83-1 (0.7-1) and 0.98-1 (0.91-1), respectively, and diagnostic accuracy [AUC (CI)] was excellent across all groups [0.99-1 (0.97-1)], with no significant differences between groups. The predictive capacity (AUC) of the VAS followed the order: experienced > students > cat owners = laypersons. VAS AUC values were lower than those of UFEPS-SF. CONCLUSIONS AND CLINICAL RELEVANCE: The UFEPS-SF can be reliably used by untrained laypersons, cat owners and students with no experience to identify acute postoperative pain in cats. The reproducibility, sensitivity and diagnostic accuracy of the VAS were suboptimal and inferior to those of the UFEPS-SF, except among experienced observers.