Although frail older patients are likely to have painful conditions, managing pain in elderly patients whose cognitive impairment prevents them from communicating well about their pain may be difficult. Krulewitch and associates conducted a prospective, observational study to determine how nonprofessional caregivers recognize pain in the confused elderly patient and to compare their reports with those of the patient. Specifically, they sought to compare standard pain assessment instruments (e.g., a visual analog scale, numerical scale).

Researchers enrolled community-dwelling patients with Alzheimer's disease, vascular dementia and other forms of dementia. The caregiver (defined as the person who provided the preponderance of daytime care) also was included in the study. Patients with Mini-Mental State Examination (MMSE) scores of 27 and above were excluded from participation.

A research assistant interviewed the caregiver and patient in the patient's home, and administered to both the Faces Pain Scale (“faces”), involving seven pictures of faces, smiling to grimacing; the Nonverbal Visual Analog Scale (“line”), in which the respondent marks a line between the extremes of “no pain” and “worst pain”; and the Pain Intensity Scale (PIS), a six-item questionnaire. Caregivers also completed the Cornell Scale for Depression in Dementia. The research assistant completed the Hospice Approach Discomfort Scale, which categorizes behaviors based on expressions and body language and is designed for use in patients with an MMSE score of less than 5. The goal was to classify the patient's worst pain in the past week. If fewer than one half of the items on each tool were completed, no score was assigned. If at least one half but fewer than all items were answered, the score was extrapolated using the available answers.

A total of 156 patient/caregiver pairs participated. The age range of the patients was 65 to 98 years, the mean MMSE score was 15.7 and most patients had advanced functional impairment. One third of the patients were unable to complete any of the instruments, and 42 percent were able to complete all of them. Of the patients who could only complete one or two scales, the PIS was most likely to be completed (with 79 percent of the patients in this group completing it). Although patients with higher MMSE scores were more likely to be able to complete more instruments, the reported pain levels did not differ based on the level of cognitive impairment. There was statistically significant agreement between patient and caregiver on all three pain instruments used, with the highest correlation occurring on the PIS.

Twenty-four of the pairs reported no pain or minimal pain, and 46 pairs reported at least moderate pain. In 16 pairs, the patient reported moderate to severe pain while the caregiver reported minimal or no pain. The Hospice Approach Discomfort Scale had low correlation with the other tools, but there were few patients in this study with extremely low MMSE scores.

The authors conclude that the PIS is the instrument most likely to determine the existence and magnitude of pain in a cognitively impaired older adult. A cognitively impaired patient is also more likely to be able to complete this instrument.