Objective: To assess the impact of treatment with divalproex sodium (DVPX) on behavioral, mood, and cognitive measures in a population of elderly nursing home residents with a history of behavior problems.
Design: Retrospective, naturalistic pre-post treatment design in which DVPX was initiated in 461 residents at individually determined points over a one-year observation period (November, 2000 through October, 2001). Patients were residents in the Beverly Enterprises Long Term Care System and were identified through patient and pharmacy databases. Minimum Data Set (MDS) items relating to problems of behavior (e.g., wandering, physical or verbal abuse of others), cognition (e.g., disordered thinking, delirium, memory), and mood were collected during on- and off-DVPX portions of the observation period.
Analysis: Two-phase generalized linear regression, with fixed intersections at the time of DVPX initiation, was used to estimate trends in each measure during the on- and off-DVPX periods. Tests for different slopes in each period were performed at alpha=0.05 (two-tailed).
Results: Prior to initiation of DVPX, an aggregate measure of frequency of behavioral symptoms showed a progressive worsening over time. After initiation of DVPX, this trend was reversed. Expressions of verbal distress, indicators of sad, apathetic, and anxious behaviors, and sleep-cycle problems were also significantly reduced during DVPX treatment. Favorable but statistically insignificant changes were noted on measures of memory and cognitive skills for daily decision-making.
Conclusions: Although obtained in an uncontrolled, natural setting, these data indicate that DVPX may have multiple beneficial effects on elderly nursing home residents with a history of behavioral problems. Prospective, randomized trials of DVPX in this population appear to be warranted.
Back to PD Thursday Poster Sessions
Back to The Eleventh International Congress