The Measurement of Pain in Dogs

Holton, Lynne Louise (2000) The Measurement of Pain in Dogs. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 13818967.pdf] PDF
Download (5MB)


The ability to measure pain is a key issue in veterinary medicine for two reasons. Firstly, adequate pain management can only be provided if the animal's pain can be recorded accurately. Secondly, pain research demands a reliable and valid measurement method if the mechanisms of pain and analgesics are to be explored scientifically. The development of pain measurement scales used in veterinary medicine has followed a similar path to that seen in human medicine. However, this is limited by the lack of effective communication between the patient and care provider. The work undertaken in this thesis is aimed at exploring the pain measurement scales commonly used in veterinary medicine and developing a novel composite measurement pain scale specifically designed for use in dogs, in a clinical setting. The consistency of the visual analogue (VAS), numerical rating (NRS) and simple descriptive (SDS) scales when used by a number of observers and over time, and the relationships between the scales were explored. The results indicated that the VAS and NRS were not adequately generalizable when used by more than one observer (generalizability coefficients between 0.27 and 0.53). The generalizability over a long period, i.e. from the day of surgery to the following day, was also low (generalizability between 0.42 and 0.45), however, the generalizability within a relatively short time period was reasonable (generalizability coefficient between 0.69 to 0.73). When using the SDS the agreement between the observers was not acceptable (Kappa statistics between 0.23 and 0.37). Thus, pain measurements made using the VAS, NRS and SDS were not consistent when used by more then one observer or over time. Investigation of the correspondence between the VAS and NRS demonstrated that a strong relationship existed, but this was dependent on the observer. The relationship of the VAS and NRS to the SDS was shown to be consistent across observers, although each category of the SDS corresponded to a wide range of NRS and VAS scores. Thus, when used in a clinical setting, these scales should not be used interchangeably since there is no unique relationship between them. Pain measurement in human medicine has progressed from the simple subjective rating scales to composite measurement scales such as the McGill Pain Questionnaire (MPQ). Since the subjective rating scales used in veterinary medicine have shown inadequate generalizability when used in a clinical setting, a composite measurement pain scale (CMPS) was developed. The construction of this scale followed the methods used in the development of the MPQ. A bank of behaviours and physiological signs were gathered from a group of practicing veterinary surgeons. This information was then rationalised and categorised to form a list of 47 behaviours and signs that were allocated into 8 behaviour categories and one category of physiological signs. Information on the pain intensity associated with each item was collected by consulting another group of veterinary surgeons. Each person assigned a pain intensity score to each item using a VAS (100mm). This information allowed the relationships between the items to be explored and the results of this investigation were reported to a focus group. This group of experts in animal pain refined the items and defined an examination procedure, which constituted the CMPS. Before the scale could be further explored, the items included were defined by consulting with a panel of 16 veterinary surgeons with specific interest in the measurement of pain. Weights were then assigned to each item using Thurstone's paired comparisons model, which provides interval level measurement. The performance of the CMPS was assessed by carrying out two studies. In the first study 5 veterinary surgeons examined 4 groups of 20 animals each. The group to which an animal belonged was defined by the reason for hospitalisation (orthopaedic surgery, soft tissue surgery and medical cases) a fourth group of clinically sound animals was also included, this comprised the control group. The pain measurement scores collected were used to investigate the validity and reliability of the CMPS. The second study involved 4 veterinary surgeons who had not previously used the CMPS. Each person watched a vide recording of examinations carried out on 12 dogs and assessed the pain each animal was experiencing, using the CMPS and a NRS. This exercise was repeated between two and four weeks later and the generalizability of the pain scores over observers and time was examined. The generalizability of the CMPS scores over the observers was improved compared to the previous study and was comparable to the NRS (generalizability coefficients of 0.61 and 0.66 for the CMPS and NRS respectively).

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Adviser: Marian Scott
Keywords: Veterinary science
Date of Award: 2000
Depositing User: Enlighten Team
Unique ID: glathesis:2000-76155
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 19 Nov 2019 16:34
Last Modified: 19 Nov 2019 16:34

Actions (login required)

View Item View Item


Downloads per month over past year