During each observation, the number of times each resident manifested each of 20 identified agitated behaviors was recorded. Presence or absence of restraints during each observation was also recorded. To determine the relationship between agitation and use of restraints, the frequency of each agitated behavior and the total number of agitated behaviors exhibited by each resident were examined during the following four conditions: (1) unrestrained (excluding the hour preceding the application of restraints); (2) unrestrained during the hour preceding the use of restraints; (3) restrained during the
hour immediately following restraint use; and (4) restrained (excluding the aforementioned hour). An analysis of variance (ANOVA) indicated that significantly more strange movements (F = 2.78, P < .05), strange noises (F = 5.40, P < .Ol), and total agitated behaviors (F = 4.89, P < .Ol) were manifested when residents were restrained as com- pared with when they were not restrained. Higher rates of agitated behaviors were observed immedi- ately after application of restraints but not imme- diately before application. This study suggests that use of restraints may contribute to agitated behav- ior rather than decrease agitated behavior.