Validity of Pain Behaviors in Persons With Cognitive Impairment
Validity of Pain Behaviors in Persons With Cognitive Impairment
Objectives: To evaluate the validity of traditional pain behaviors (guarding, bracing, rubbing, grimacing, and sighing) in persons with and without cognitive impairment and chronic low back pain (CLBP).
Design: Prospective observational study.
Setting: Outpatient clinics.
Participants: Thirty-seven cognitively intact and 40 cognitively impaired participants with and without CLBP.
Measurements: Frequency of traditional pain behaviors.
Results: Forty-six of the participants were pain free, and 31 had CLBP. The internal consistency reliability coefficient of the five pain behaviors was 0.32, suggesting that a unidimensional scale did not exist. Multivariate analysis of variance analysis according to the independent variables pain status (pain free vs CLBP) and cognitive status (intact vs impaired) with the dependent variable frequency of pain behaviors found significant differences according to pain status (F[5,61]=3.06, P=.02) and cognitive status (F[5,61]=5.41, P<.001) but without evidence of an interaction (F[5,61]=1.14, P=.35). Participants with CLBP exhibited significantly higher levels of grimacing (P<.001) and guarding (P=.02) than pain-free participants. Intact subjects exhibited fewer guarding (P=.02) and rubbing behaviors (P<.001) but a higher number of bracing behaviors (P=.03) than cognitively impaired participants.
Conclusion: These results support the utility of facial grimacing in assessing pain in patients with mild to moderate cognitive impairment and call into question the validity of guarding and rubbing in assessing pain in persons with mild to moderate cognitive impairment.
Older age is the greatest risk factor for the development of cognitive impairment. Given demographic shifts in the U.S. population, the number of persons with Alzheimer's disease or related dementias is expected to increase from 5 million today to 8 million by 2030. Dementia is defined as an acquired loss of memory that is substantial enough to interfere with everyday functioning and is usually associated with behavioral and psychological symptoms. At the same time, persons with cognitive impairment often have multiple coexisting morbidities such as arthritis, heart disease, diabetes mellitus, and peripheral vascular disease that result in additional physical symptoms such as pain. Research suggests that pain represents a particularly common physical symptom in persons with mild to moderate cognitive impairment.
As with cognitive impairment, chronic pain increases in frequency with age, and the presence of pain has been associated with depression, impaired cognitive function, impaired physical function, sleep disturbance, agitation, and decreased socialization. The prevalence of pain in persons with cognitive impairment is typically reported to be 50% or greater but varies with site (nursing home, assisted living, community-dwelling, and hospital), population (mild, moderate, and severe cognitive impairment), and assessment method (self-report, proxy report, and observational scales). Moreover, studies continue to find that persons with cognitive impairment remain at higher risk for being undertreated for their pain than cognitively intact persons, which negatively affects quality of life.
A necessary first step to improve pain management is better detection methods. Given that pain is a subjective experience, self-report is considered the criterion standard detection method. Studies demonstrate that several reliable and valid self-report scales exist for use in persons with mild to moderate cognitive impairment, but self-report is not always feasible, especially as memory worsens and language skills decline. For people unable to report pain, healthcare providers rely on surrogate observations of behavioral indicators, yet the specificity of many of these behaviors (e.g., guarding, bracing) is unknown. That is, in addition to being an expression of pain, it is possible that, for people with cognitive impairment, these behaviors represent anxiety, fear, depression, or nonspecific distress. Even though many behavioral scales have been developed and found reliable (acceptable internal consistency, interrater reliability, and test-retest reliability), the validity (their ability to distinguish between cognitively impaired pain-free and pain-ridden older adults) of pain behaviors in determing pain in persons with cognitive impairment has not been evaluated.
The overarching purpose of this study was to examine the validity of behavioral indicators of pain in persons with mild to moderate cognitive impairment. Specifically, it was desired to examine how well the expression of five classic pain behaviors (grimacing, guarding, bracing, rubbing, and sighing) could be used to discriminate between four groups of community-dwelling older adults: cognitively intact and pain free, cognitively intact with chronic low back pain (CLBP), cognitively impaired and pain free, and cognitively impaired with CLBP.
Abstract and Introduction
Abstract
Objectives: To evaluate the validity of traditional pain behaviors (guarding, bracing, rubbing, grimacing, and sighing) in persons with and without cognitive impairment and chronic low back pain (CLBP).
Design: Prospective observational study.
Setting: Outpatient clinics.
Participants: Thirty-seven cognitively intact and 40 cognitively impaired participants with and without CLBP.
Measurements: Frequency of traditional pain behaviors.
Results: Forty-six of the participants were pain free, and 31 had CLBP. The internal consistency reliability coefficient of the five pain behaviors was 0.32, suggesting that a unidimensional scale did not exist. Multivariate analysis of variance analysis according to the independent variables pain status (pain free vs CLBP) and cognitive status (intact vs impaired) with the dependent variable frequency of pain behaviors found significant differences according to pain status (F[5,61]=3.06, P=.02) and cognitive status (F[5,61]=5.41, P<.001) but without evidence of an interaction (F[5,61]=1.14, P=.35). Participants with CLBP exhibited significantly higher levels of grimacing (P<.001) and guarding (P=.02) than pain-free participants. Intact subjects exhibited fewer guarding (P=.02) and rubbing behaviors (P<.001) but a higher number of bracing behaviors (P=.03) than cognitively impaired participants.
Conclusion: These results support the utility of facial grimacing in assessing pain in patients with mild to moderate cognitive impairment and call into question the validity of guarding and rubbing in assessing pain in persons with mild to moderate cognitive impairment.
Introduction
Older age is the greatest risk factor for the development of cognitive impairment. Given demographic shifts in the U.S. population, the number of persons with Alzheimer's disease or related dementias is expected to increase from 5 million today to 8 million by 2030. Dementia is defined as an acquired loss of memory that is substantial enough to interfere with everyday functioning and is usually associated with behavioral and psychological symptoms. At the same time, persons with cognitive impairment often have multiple coexisting morbidities such as arthritis, heart disease, diabetes mellitus, and peripheral vascular disease that result in additional physical symptoms such as pain. Research suggests that pain represents a particularly common physical symptom in persons with mild to moderate cognitive impairment.
As with cognitive impairment, chronic pain increases in frequency with age, and the presence of pain has been associated with depression, impaired cognitive function, impaired physical function, sleep disturbance, agitation, and decreased socialization. The prevalence of pain in persons with cognitive impairment is typically reported to be 50% or greater but varies with site (nursing home, assisted living, community-dwelling, and hospital), population (mild, moderate, and severe cognitive impairment), and assessment method (self-report, proxy report, and observational scales). Moreover, studies continue to find that persons with cognitive impairment remain at higher risk for being undertreated for their pain than cognitively intact persons, which negatively affects quality of life.
A necessary first step to improve pain management is better detection methods. Given that pain is a subjective experience, self-report is considered the criterion standard detection method. Studies demonstrate that several reliable and valid self-report scales exist for use in persons with mild to moderate cognitive impairment, but self-report is not always feasible, especially as memory worsens and language skills decline. For people unable to report pain, healthcare providers rely on surrogate observations of behavioral indicators, yet the specificity of many of these behaviors (e.g., guarding, bracing) is unknown. That is, in addition to being an expression of pain, it is possible that, for people with cognitive impairment, these behaviors represent anxiety, fear, depression, or nonspecific distress. Even though many behavioral scales have been developed and found reliable (acceptable internal consistency, interrater reliability, and test-retest reliability), the validity (their ability to distinguish between cognitively impaired pain-free and pain-ridden older adults) of pain behaviors in determing pain in persons with cognitive impairment has not been evaluated.
The overarching purpose of this study was to examine the validity of behavioral indicators of pain in persons with mild to moderate cognitive impairment. Specifically, it was desired to examine how well the expression of five classic pain behaviors (grimacing, guarding, bracing, rubbing, and sighing) could be used to discriminate between four groups of community-dwelling older adults: cognitively intact and pain free, cognitively intact with chronic low back pain (CLBP), cognitively impaired and pain free, and cognitively impaired with CLBP.
