Abdominal pain assessment in rabbits: using the CANCRS to recognize pain and testing its internal validity over time

A composite scale for pain assessment in rabbits has been previously designed and tested (CANCRS). The present study describes the refinement of the scale and the evaluation of its ability to detect pain variations over time. Furthermore, a comparison between the CANCRS and the Visual Analogue Scale (VAS) has been performed, to underline the differences between an objective (CANCRS) and a subjective (VAS) assessment of abdominal pain. In the first part of the study, 86 rabbits (n=47 heathy patients and n=39 patients with gastrointestinal stasis syndrome) underwent pain assessments with the VAS and the CANCRS. Thirty-two patients with gastrointestinal stasis syndrome participated to the second part of the study. These patients underwent four pain assessments with the CANCRS. The first assessment took place before meloxicam administration and the others after 30, 60 and 90 minutes. The CANCRS showed differences between healthy and diseased rabbits (P = 0.0001), median scores were 5 (IQR 4 - 6) and 9 (IQR 7 - 11) respectively. The VAS showed differences between healthy and diseased rabbits (P = 0.02), the median scores were 4 (IQR 2 - 5.35) and 5.3 (IQR 2.65 - 6.45) respectively. The cut-off scores for the CANCRS and for the VAS for differentiation between healthy and diseased patients were 7 (Sp 89%, Se 79%) and 4.4 (Sp 59%, Se 69%) respectively. Sensitivity and specificity for each parameter of the CANCRS were calculated, in order to obtain weighting factors. Accordingly, the evaluation of respiratory pattern and vocalizations should be excluded from the CANCRS, since their performances in pain evaluation are poor. Internal validity of the CANCRS was tested assessing pain before and after the analgesic treatment and the results showed significancy at each time point. The CANCRS showed better performances than the VAS and its responsiveness to pain variations has been verified.

80 subjective assessments for pain in rabbits. This led to the decision of testing the VAS, which is a 81 tool that allows a subjective assessment of pain, with numerical results. Specifically, the objectives 102 The owners of 7 of the 39 patients belonging to the RGIS group refused hospitalization. Therefore, 103 only 32 rabbits of the RGIS group were included in the second part of the study. Furthermore, 104 rabbits requiring nasogastric tube were excluded from the study because of its interference with the 105 use of facial expression for pain evaluation.
106 The first part of the study took place at admission: all the rabbits underwent a pain assessment with 107 both the CANCRS and the VAS. Regardless of the attending clinician, the same veterinarian 108 performed all the assessments using the VAS and the CANCRS at the end of the clinical 109 examination. The VAS score was recorded prior to the assessment of the patients with the 110 CANCRS, in order to avoid bias deriving from clinical data collection.
111 Pain assessments for the first part of the study always happened in the clinical examination room.
112 The patients were taken out form the carrier and positioned on a table for clinical examination and 113 pain assessment.

114
The second part of the study included thirty-two of the patients included in the RGIS group. These 115 patients underwent four pain evaluation with the CANCRS during the hospitalization. A pain 116 assessment was performed just before the administration of the anti-inflammatory drug (meloxicam 117 1 mg/kg SQ q24). Afterwards, three assessment were performed 30 (T30), 60 (T60), and 90 (T90) 118 minutes after the administration of the anti-inflammatory drug. Pain assessments for the second part 119 of the study always happened in the hospitalization room. The patients were taken out from the cage 120 and positioned on a table for clinical examination and pain assessment.
121 Environmental conditions, including light, noise, and temperature were stable through the process. 127 As for what concern the CANCRS, the scores were given as stated in our previous study (4) and the 128 scale is presented in Table 1. It includes five facial action units (orbital tightening, cheek flattening, 129 nostril shape, whisker position, and ear position) and each one was scored according to whether it 130 was not present (score 0), moderately present (score 1) and obviously present (score 2).
131 Furthermore, it includes some clinical parameters. Pupillary dilation was scored as 1 if present or 132 scored as 0 if not present. For heart rate, increases lower than 20% of the physiological value (250 133 bpm)(16) were given the score of 0, increases higher than 20% were given the score of 1, and 134 increases higher than 50% were given the score of 2. Heart rate was measured with a stethoscope 135 for 15 seconds and the result was multiplied by four, in order to obtain beats per minute.
136 The average respiratory rate varies from 30 to 60 bpm (16); for rates of 60 bpm or less a score 0 137 was given, for rates between 61 and 72 a score 1 was given, for rates between 73 and 90 a score of 138 2 was given, and for rates higher than 90 bpm a score of 3 was given.
140 Since the palpation of the suspected painful area is a parameter included in the CANCRS, 141 abdominal palpation was delicately performed and the reaction of the patient was classified as no 142 reaction (score 0), reaction during the palpation (score 1) and reaction before the palpation (score 143 2).
144 Absence of vocalization was scored as 0, vocalization when touched was scored as 1, intermittent 145 vocalization without any contact with the operator was scored as 2, whereas continuous 146 vocalization was scored as 3.