Wiseman-Orr, Margaret Lesley
The development of an instrument to measure chronic pain in dogs.
PhD thesis, University of Glasgow.
Full text available as:
This research applied the psychometric approaches of proxy human chronic pain and HRQL instrument development to the development of an instrument to measure chronic pain in the dog, using the owner to provide a proxy report. The development of the instrument followed established steps designed to ensure an instrument’s validity: the identification of all domains relevant to the measurement of interest; generation of a pool of potential instrument items; selection of instrument items from the item pool, and validation of that selection; design and pre-testing of the prototype instrument; field-testing of the instrument to establish its psychometric properties.
Domain identification was carried out through interviews with 47 owners of dogs suffering chronic pain. Potential items (descriptive terms) were generated using questionnaires completed by 165 dog owners. These domains and the items selected were subsequently validated by 12 veterinary practitioners and by 10 owners of dogs suffering chronic pain. The validated list of items was incorporated into a structured questionnaire, and this instrument was pre-tested using 26 dog owners. The finished instrument was then field-tested using the owners of 155 dogs who completed a total of 390 questionnaires prior to and during treatment at the University of Glasgow Small Animal Hospital and at a local Veterinary Practice, a majority of which dogs were suffering from chronic degenerative joint disease (DJD).
Factor analysis of the instrument responses for dogs suffering DJD revealed an interpretable 12-factor model, in which factors were interpreted as domains of canine HRQL: ‘vitality’, ‘physical limitation’, ‘lethargy’, ‘anxiety’, ‘aggression’, ‘emotional upset’, ‘appetite’, ‘consistency of behaviour’, ‘mental disturbance’, ‘attention-seeking’, ‘sadness’ and ‘acceptance’. This analysis provided evidence for the construct (factorial) validity of the instrument, since responses to instrument items revealed an underlying structure that reflected the construct upon which the instrument was developed.
Scores were calculated for each of the 12 domains of HRQL identified by the factor analysis, and these were able to discriminate between dogs with chronic pain and healthy dogs on >86% of occasions.
Profiles of HRQL scores obtained for dogs with chronic pain were compared with those obtained for healthy dogs in a control group, and differences in these profiles were observed. An examination of changes in HRQL domain scores over time for individual dogs revealed that these scores tended to reflect clinical change in those individuals.
The process described here offers a novel approach to the development of chronic pain and HRQL instruments for a range of animal species, and may have relevance for human chronic pain and HRQL instrument development.
Actions (login required)