An Eye Tracking Investigation of Pain Decoding Based on Older and Younger Adults\u2019 Facial Expressions

Rhonda J N Stopyn,Jeff Loucks,Thomas Hadjistavropoulos

doi:10.1007/s10919-020-00344-0

Rhonda J N Stopyn, Jeff Loucks + Show 1 more

Open Access

https://doi.org/10.1007/s10919-020-00344-0

Copy DOI

Abstract

Nonverbal pain cues such as facial expressions, are useful in the systematic assessment of pain in people with dementia who have severe limitations in their ability to communicate. Nonetheless, the extent to which observers rely on specific pain-related facial responses (e.g., eye movements, frowning) when judging pain remains unclear. Observers viewed three types of videos of patients expressing pain (younger patients, older patients without dementia, older patients with dementia) while wearing an eye tracker device that recorded their viewing behaviors. They provided pain ratings for each patient in the videos. These observers assigned higher pain ratings to older adults compared to younger adults and the highest pain ratings to patients with dementia. Pain ratings assigned to younger adults showed greater correspondence to objectively coded facial reactions compared to older adults. The correspondence of observer ratings was not affected by the cognitive status of target patients as there were no differences between the ratings assigned to older adults with and without dementia. Observers’ percentage of total dwell time (amount of time that an observer glances or fixates within a defined visual area of interest) across specific facial areas did not predict the correspondence of observers’ pain ratings to objective coding of facial responses. Our results demonstrate that patient characteristics such as age and cognitive status impact the pain decoding process by observers when viewing facial expressions of pain in others.

Highlights

Observers tend to decode another’s pain experience by interpreting verbal and nonverbal cues (Hadjistavropoulos et al 2011; Prkachin and Craig 1995)
Following examination of the interrater reliability of our facial action coding system (FACS) coding for our specific stimulus video set, we conducted a 2 x 3 analysis of variance (ANOVA) with FACS-based scores as the dependent variable
While observers tended to look more often at the eyes and nose interest areas across all conditions, they looked less at the facial pain cues of patients with dementia compared to individuals without dementia

Summary

Introduction

Observers tend to decode another’s pain experience by interpreting verbal and nonverbal cues (Hadjistavropoulos et al 2011; Prkachin and Craig 1995). Our goal was to examine the way in which observers decode nonverbal expressive behaviors of pain in older adults with and without dementia while using eye tracker technology. While self-report of pain is seen as the gold standard, the ability to report pain deteriorates as dementia progresses (Hobson 2008) Nonverbal pain cues, such as facial expressions, can facilitate valid pain assessments as these cues tend to be relatively preserved even after linguistic and cognitive abilities deteriorate (Hadjistavropoulos and Craig 2002; Hadjistavropoulos et al 2011; Lints-Martindale et al 2012). Several facial movements have been proposed to be directly associated with pain expression including: brow lowering, orbit tightening, levator contraction, and eye closure (Prkachin 1992), as demonstrated in numerous studies (e.g., Ashraf et al 2009; Gallant and Hadjistavropoulos 2017; Hadjistavropoulos et al 2018; Prkachin 2005; Prkachin and Solomon 2008; Rash et al 2019; Rocha et al 2003)

Objectives

Methods

Results

Discussion

Conclusion