RESUMEN
BACKGROUND: People with dementia experience a decline in language skills required to self-report pain; researchers thus recommend the use of nonverbal behaviors to assess pain. Although multiple instruments exist for assessing nonverbal pain behaviors, psychometric data are lacking for African American nursing home residents with dementia. AIMS AND DESIGN: The purpose of this methodological study was to describe the development and testing of the Pain Assessment Tool in Cognitively Impaired Elders (PATCIE) in African American and Caucasian nursing home residents with dementia. SETTINGS/PARTICIPANTS: The convenience sample included 56 African American and 69 Caucasian residents with dementia in multiple nursing homes from three states. The research staff completed the pain assessments when the nursing home staff transferred the residents. RESULTS/CONCLUSIONS: Initially, 15 nonverbal pain behaviors were evaluated. Based on the alpha scores and additional literature review, the 15 nonverbal pain behaviors were expanded to 28 behaviors. The PATCIE had a Cronbach's alpha of .73 during movement. Construct validity for the pain behaviors was demonstrated because higher scores were noted during movement, and scores before movement were significantly higher than those obtained after movement. For movement over time, there was a significant difference in the PATCIE score, regardless of ethnicity or time (p < .0001). There were no significant differences found between ethnic groups, either overall or in change over time between movements or between the categories of cognitive function. African Americans were more likely to display frowning, and Caucasians to display irritability. The PATCIE demonstrates preliminary reliability and validity in assessing pain in African American and Caucasian nursing home residents with dementia.