Related Experiment Video
Updated: Aug 8, 2025
Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli
Published on: April 5, 2019
Assessing pain using facial recognition software among Aboriginal aged care residents with cognitive impairment: A
Chris Rissel1, Nicole Tate2, Leigh Moore1
1Rural and Remote Health Northern Territory, Flinders University, Darwin, Northern Territory, Australia.
Objective:
To assess the observed pain behaviours of Aboriginal residents with cognitive impairment in aged care facilities and compare these results with a matched national sample of non-Aboriginal residents.
Methods:
Observed pain behaviours of Aboriginal residents (N = 87) with cognitive impairment in aged care facilities across the Northern Territory of Australia were assessed using PainChek® Adult and compared with data from a matched national sample of non-Aboriginal residents (N = 420). Pain scores were derived from inbuilt automated facial recognition and analysis software plus a series of digital checklists requiring manual input by care staff.
Results:
The median total pain score for the Aboriginal residents was 2 (IQR 1-4) and for the matched external residents was 3 (IQR 2-5). In a multivariable negative binomial regression model, this difference in total pain score was statistically significant (p < 0.001). The pain score derived from the automated facial recognition and analysis component of the PainChek® Adult app was not statistically different between the two groups when adjusted for multiple observations and context of observation (odds ratio = 1.06, 95% confidence interval 0.97-1.16, p = 0.169).
Conclusions:
We found under-reporting of observed pain signs and behaviours for Aboriginal aged care residents by assessors. Further training in the assessment of pain in Aboriginal and Torres Strait Islander aged care residents may be necessary and a continuing shift in clinical practice to using technology and point-of-care assessment.
