Abstract
OBJECTIVES
We describe the validation of an informant-based pain assessment for persons suffering from dementia called Pain Assessment in Noncommunicative Elderly persons (PAINE) using 2 different samples.
METHODS
In the first study, the participants included 80 residents from one nursing home. We assessed internal consistency using Cronbach alpha, interrater and test-retest reliability using Pearson correlations, and validity using receiver operating characteristic curve analyses, comparing PAINE to these criteria on the basis of reports from physicians, nurses, relatives, and the residents themselves. In the second study, the participants included 91 residents from 2 different nursing homes. We assessed validity by correlating scores on PAINE with those from other assessments designed to detect pain.
RESULTS
PAINE shows adequate internal consistency and both interrater and test-retest reliability. It also shows adequate receiver operating characteristic curve results and reasonable correlations with the existing measures of pain in persons with dementia.
DISCUSSION
PAINE has the advantage of using a comprehensive list of pain symptoms on the basis of systematic questioning of direct caregivers from several institutions. The validity results suggest that this assessment could be a useful tool in detecting pain in persons with dementia.
Collapse