Randomised controlled trials (RCTs) that assessed non-pharmacological interventions for agitation in patients with dementia were eligible for inclusion. Eligible trials had to use a published scale to assess agitation as an outcome and report sufficient information to determine the effects of therapy. Both parallel group and cross-over trials were eligible.
Interventions assessed were sensory interventions (aromatherapy, thermal bath, calming music and hand massage), social contact (simulated presence, pet therapy), activities (rocking chair therapy, therapeutic recreational activities), environmental modification (morning bright light therapy), carer/caregiver training (behaviour management techniques, abilities focused morning care), combination therapy (stimulation retreat program) and behaviour therapy (activities of daily living intervention and way finding intervention). Comparator interventions included placebo, usual care and no intervention. Intervention duration ranged from 10 minutes to one year. The majority of participants were women resident in nursing homes or other care facilities.
Two reviewers independently assessed studies for inclusion; disagreements were resolved through consensus between three reviewers.