Validation of the Unesp-Botucatu pig composite acute pain scale (UPAPS) in piglets undergoing castration

Abstract

To accurately assess pain and support broadly-based analgesic protocols to mitigate swine pain, it is imperative to develop and validate a species-specific pain scale. The objective of this study was to investigate the clinical validity and reliability of an acute pain scale (UPAPS) adapted for newborn piglets undergoing castration. Thirty-nine male piglets (five days of age, 1.62 ± 0.23 kg BW) served as their own control, were enrolled in the study and underwent castration in conjunction with an injectable analgesic administered one-hour post-castration (flunixin meglumine 2.2 mg/kg IM). An additional 10, non-painful female piglets were included to account for the effect of natural behavioral variation by day on pain scale results. Behavior of each piglet was video recorded continuously at four recording periods (24 h pre-castration, 15 min post-castration, 3 and 24 h post-castration). Pre- and post-operative pain was assessed by using a 4-point scale (score 0-3) including the following six behavioral items: posture, interaction and interest in surroundings, activity, attention to the affected area, nursing, and miscellaneous behavior. Behavior was assessed by two trained blinded observers and statistical analysis was performed using R software. Inter-observer agreement was very good (ICC = 0.81). The scale was unidimensional based on the principal component analysis, all items except for nursing were representative (rs ≥ 0.74) and had excellent internal consistency (Cronbach's alpha ≥ 0.85). The sum of scores were higher in castrated piglets post-procedure compared to pre-procedure, and higher than in non-painful female piglets confirming responsiveness and construct validity, respectively. Scale sensitivity was good when piglets were awake (92.9%) and specificity was moderate (78.6%). The scale had excellent discriminatory ability (area under the curve > 0.92) and the optimal cut-off sum for analgesia was 4 out of 15. The UPAPS scale is a valid and reliable clinical tool to assess acute pain in castrated pre-weaned piglets.

Fig 1. Flow of the experimental design.

Fig 2. Timeline of the periods of observations used for pain scale validation.

Fig 3. Distribution of UPAPS items scores (percentage of occurrence) in castrated piglets (n = 39).

Legend: M1 (24h pre-procedure), M2 (15-min post-procedure, before rescue analgesia), M3 (3h post-procedure, after rescue analgesia), M4 (24h post-procedure) GM—data of the grouped moments (M1 + M2 + M3 + M4).

Fig 4. Distribution of UPAPS items scores (percentage of occurrence) in awake castrated piglets (n = 14).

Legend: M1 (24h pre-procedure), M2 (15-min post-procedure, before rescue analgesia); GM—data of the grouped moments (M1 + M2).

Fig 5. Biplot of the principal component analysis of the UPAPS items applied to all castrated piglets (n = 39).

UPAPS–Unesp-Botucatu pig acute composite pain scale. Confidence ellipses indicate perioperative moments and pain scores. Moments: M1 (24h pre-procedure), M2 (15-min post-procedure, before rescue analgesia), M3 (3h post-procedure, after rescue analgesia), M4 (24h post-procedure). Ellipses were constructed according to the assessment. Each circle corresponds to the score from one observer attributed to each piglet at each moment.

Fig 6. Biplot of the principal component analysis of the UPAPS items applied to awake piglets (n = 14 castrated piglets).

UPAPS–Unesp-Botucatu pig acute composite pain scale. Confidence ellipses indicate perioperative moments and pain scores. Moments: M1 (24h pre-procedure), M2 (15-min post-procedure, before rescue analgesia). Ellipses were constructed according to the assessment. Each circle corresponds to the score from one observer attributed to each piglet at each moment.

Fig 7. Violin plot of the total sum (median/amplitude) of the a) UPAPS (all castrated piglets; n = 39) and b) UPAPS (awake castrated piglets; n = 14) before and after castration in piglets.

The violin contour represents the dispersion of data density, with the wider contour representing greater data density. The top and bottom box lines represent the interquartile range (25 to 75%), the line within the box represents the median, the extremes of the whiskers represent the minimum and maximum values, black lozenges (♦) represent the mean, black circles (●) represent outliers. UPAPS: Unesp-Botucatu pig acute composite pain scale. Different letters express significant differences between moments where a > b > c, according to the generalized mixed linear models [27]. M1 (24h pre-procedure), M2 (15-min post-procedure, before rescue analgesia), M3 (3h post-procedure, after rescue analgesia), M4 (24h post-procedure).

Fig 8. Violin plot of the total sum (median/amplitude) of the a) UPAPS for castrated piglets (n = 14) and non-painful female piglets (n = 6) after castration (M2) and before rescue analgesia in piglets awake during the assessment.

The violin contour represents the dispersion of data density, with the wider contour representing greater data density. The top and bottom box lines represent the interquartile range (25 to 75%), the line within the box represents the median, the extremes of the whiskers represent the minimum and maximum values, black lozenges (♦) represent the mean, black circles (●) represent outliers. Sub-UPAPS: Unesp-Botucatu pig acute composite pain scale. Different letters express significant differences between moments where a > b > c, according to the generalized mixed linear models [27]. M2 (15-min post-procedure, before rescue analgesia).

Fig 9. Receiver Operating Characteristic (ROC) curve and area under the curve (AUC) of the UPAPS applied to castrated piglets awake (n = 14).

AUC values above 0.90 represent high discriminatory capacity (accuracy) of the scale [37]. The gray region represents the 95% confidence interval.