Assessment of pain in cognitively impaired older adults: a comparison of pain assessment tools and their use by nonprofessional caregivers.

Abstract

To compare self-reporting of pain by cognitively impaired (CI) older adults living in the community with reports of their caregivers; to identify the factors associated with differences in agreement; and to identify those pain assessment tools that are most useful to nonprofessional caregivers. Prospective observational cohort of patients with dementia and their caregivers. In-home assessments of community-dwelling CI older adults and family members, adult foster care providers, or aides in residential care facilities. All CI subjects were enrolled in a Program for All-inclusive Care of the Elderly program (ElderPlace) in Portland, Oregon. A total of 156 CI adults and their caregivers. Standardized pain assessment instruments were administered to the CI subjects and their caregivers in the subject's home by trained research assistants between June and October 1998. CI subjects were assessed with the Nonverbal Visual Analog scale ("line"), the Faces Pain Scale ("faces"), and the Philadelphia Pain Intensity Scale (PIS). Research assistants observed pain behaviors in the CI subjects using the Hospice Approach Discomfort Scale. The CI subject's caregivers completed the "line," "faces," PIS, and Cornell Scale for Depression in Dementia. A baseline Mini-Mental State Exam, Functional Assessment Stage Test, and Geriatric Depression Scale were performed on all CI subjects by ElderPlace medical, nursing, or social work staff. The mean age of the CI subjects was 83, and 83% were women. The mean Mini-Mental State Exam score was 15.7. One-third of the CI subjects were unable to complete any of the three pain assessment tools. Of the 104 subjects completing at least one tool, 13 (12.5%) reported no pain and 91 (87.5%) reported some pain. Sixteen (10%) of the subjects were depressed as measured with the GDS or Cornell instrument. In 70 of the 104 subjects (67%) able to complete any tool, the caregiver and CI subject agreed as to the level of pain experienced by the CI subject. The number of tools completed by the CI subjects decreased with increased cognitive impairment. The Pain Intensity Scale was the tool most likely to be completed by both CI subjects and caregivers. The means of test scores were not significantly different for the paired groups of CI subject and caregiver, and the nonparametric correlation of each tool was significant: "faces" (Spearman's rho (p) = .417); "line" (p = .420); and PIS (p = .452). The Hospice Approach Discomfort Scale did not correlate well with other tools. The Pain Intensity Scale seems to be more useful than other pain assessment tools in assessing pain in cognitively impaired patients and can be used by nonprofessional caregivers in a community-based care setting.