Wandering and physically non-aggressive agitation: Are they equivalent?

Abstract

Background: This study examined equivalence of two dementia-compromised behaviors: wandering and physically non-aggressive agitated behavior (PNAB). Methods: Using a cross-sectional correlational design, ambulatory residents with dementia (N = 181) from 22 NH and 6 ALF were studied. Each was videotaped an average of nine times during randomly selected 20-minute periods spread over two days, separated by a 48-hour interval. Videotapes were coded by trained raters using Noldus Observer 5.0 software configured for wandering patterns (random, lapping, pacing) and for behaviors on the physically non-aggressive agitated behavior (PNAB) subscale of the Cohen-Mansfield Agitation Inventory. Results: Wandering was observed in 73.5% of participants, PNAB in 100%. Principal components factor analyses yielded a one-factor solution for wandering and a two-factor solution for PNAB (explained variance = 43.66%, 53.45% respectively). Overall wandering rate correlated significantly with PNAB Factor 1 (r =.68, p < .001) and Factor 2 (r =.26, p < .01). At p < .01, random wandering correlated with both Factors (r = .63, .27 respectively); lapping and pacing wandering patterns correlated only with Factor 1 (r = .35, .17). Internal consistency of the CMAI-PNAB calculated with observational data did not achieve previously reported levels calculated with caregiver report data. Conclusions: Findings depict wandering and the PNAB sub-type as overlapping, but non-equivalent phenomena. Further conceptual and empirical development of the PNAB sub-type is warranted. Results have implications for accuracy in scientific and clinical detection and labeling of wandering and agitation.