The short form of the Glasgow Composite Measure Pain Scale in post-operative analgesia studies in dogs: a scoping review

Testa, Barbara (2022) The short form of the Glasgow Composite Measure Pain Scale in post-operative analgesia studies in dogs: a scoping review. MVM(R) thesis, University of Glasgow.

Full text available as:
[thumbnail of 2022TestaMVM(R).pdf] PDF
Download (4MB)


The measurement and treatment of acute pain in animals are essential from a welfare perspective. Valid pain-related outcome measures are also crucial for ensuring reliable and translatable findings in veterinary clinical trials. The short form of the Glasgow Composite Measure Pain Scale (GCMPS-SF) is a multi-item behavioural pain assessment tool, developed and validated using a psychometric approach, to measure acute pain in the dog. The psychometric approach refers to a scientific method used to develop tools intended to measure complex and multifaceted constructs like pain. Relevant words and expressions related to pain are collected, refined, and classified into domains and associated categories through a multi-step approach that involves the participation of a large sample of pain sufferers (or observers for non-verbal patients) and a pool of experts in the field. Ultimately, the instrument developed is tested by clinical studies to assess its validity, reliability, and responsiveness. While this approach has been widely adopted to reliably assess pain in humans, the GCMPS-SF is at present the only validated tool to measure acute pain in dogs developed using this methodology.

The GCMPS-SF comprises four sections (section A: observation of resting behaviours from a distance, section B and C: evaluation of interactive behaviours, section D: assessment of the overall attitude of the patient), with instructions for completion provided at the beginning of each section. The questionnaire encompasses two categories within each section, incorporating a total of six behavioural categories. These categories are associated with multiple descriptive expressions of pain, assigned an individual score each and ordered in an increased level of severity within the category.

We conducted a scoping review through systematic search of the literature to identify prospective research studies that have used the GCMPS-SF. We aimed to describe the contexts in which it has been used, verify the correct use of the scale, examine whether these studies are well-designed and adequately powered, and determine whether statistically significant differences in GCMPS-SF scores appear clinically relevant.

We identified 114 eligible studies, indicating widespread use of the scale. We documented a limited number of modifications to the scale and intervention level, which would alter its validity, and a variety of methods to analyse the data derived from the scale.

We also documented many deficiencies in reporting of experimental design in terms of the observers used, the underlying hypothesis of the research, the statement of primary outcome, the use of a priori sample size calculations, blinding and randomisation strategies. These deficiencies in reporting and study design may predispose to both Type I and Type II statistical errors in the small animal pain literature. Results of our analyses also suggest that methodological factors affected study outcomes in our dataset. The probability of finding a statistically significant difference was 7 times higher in studies that used negative control groups, 3 times higher when the GCMPS-SF scores were used as a primary outcome, and 12 times higher if the pain scale was modified.

Finally, we documented a wide range (1.00 to 11.0) of actual effect sizes in GCMPS-SF scores, with approximately 30% of the values below 1.60, and a median largest actual effect size of 2.00 in trials that declared statistical significance. With the consideration that clinical relevance is perhaps more anchored to the intervention level with the GCMPS-SF, rather than to a minimum difference in pain scores, we question whether some of the differences detected, albeit statistically significant, are clinically relevant without accounting for their position on the scale.

Based on our findings, we encourage methodologically sound study design, high quality of reporting, and a more robust use of the scale and derived data to ensure attainment of reliable and translatable outcomes.

Item Type: Thesis (MVM(R))
Qualification Level: Masters
Subjects: S Agriculture > SF Animal culture > SF600 Veterinary Medicine
Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Biodiversity, One Health & Veterinary Medicine
Supervisor's Name: Bell, Dr Andrew and Murison, Professor Pamela J.
Date of Award: 2022
Depositing User: Theses Team
Unique ID: glathesis:2022-83386
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 31 Jan 2023 10:07
Last Modified: 01 Feb 2023 10:31
Thesis DOI: 10.5525/gla.thesis.83386
Related URLs:

Actions (login required)

View Item View Item


Downloads per month over past year