A longitudinal examination of agitation and resident characteristics in the nursing home

Purpose: Agitation frequently accompanies cognitive decline among nursing home residents. This study used cross-sectional and longitudinal (up to 18 months) methods to examine agitation among profoundly and moderately impaired residents using both staff report and direct observation methods. Design and Methods: The study included participants (N = 78) from a larger study who completed either 12 or 18 months of data collection. There were four measurement points, each separated by 6-month intervals. We recorded agitation using two measures: (a) the computer-assisted behavioral observation system (CABOS) and (b) the staff-completed Nursing Home Behavior Problem Scale. We used longitudinal hierarchical linear modeling to capture the dynamic nature of behavior change as a function of individual resident characteristics and time. Results: The profoundly cognitively impaired residents (M Mini-Mental State Examination [MMSE] score = 1.7) displayed more agitation than the moderately impaired group (M MMSE score = 15.4) at Epoch 1 (cross-sectional analyses). Longitudinal analysis found a significant linear and quadratic trend only with the profoundly impaired residents using the CABOS. These residents showed slight improvements in agitation up to 12 months, with agitation increasing significantly from 12 to 18 months. Implications: With nursing homes receiving an increasing number of profoundly cognitively impaired residents, these results have potential cost and policy implications. Measuring agitation over time by using both staff report and observation measures presents various problems, and we present an alternative measure that may help to avoid these difficulties. Key Words: Agitation, Nursing home, Longitudinal methods, Direct observation