Background: Evidence for improved outcomes for people with dementia through provision of person-centred care and dementia-care mapping is largely observational. We aimed to do a large, randomised comparison of person-centred care, dementia-care mapping, and usual care.

Methods: In a cluster randomised controlled trial, urban residential sites were randomly assigned to person-centred care, dementia-care mapping, or usual care. Carers received training and support in either intervention or continued usual care. Treatment allocation was masked to assessors. The primary outcome was agitation measured with the Cohen-Mansfield agitation inventory (CMAI). Secondary outcomes included psychiatric symptoms including hallucinations, neuropsychological status, quality of life, falls, and cost of treatment. Outcome measures were assessed before and directly after 4 months of intervention, and at 4 months of follow-up. Hierarchical linear models were used to test treatment and time effects. Analysis was by intention to treat. This trial is registered with the Australia and New Zealand Clinical Trials Registry, number ACTRN12608000084381.

Findings: 15 care sites with 289 residents were randomly assigned. Pairwise contrasts revealed that at follow-up, and relative to usual care, CMAI score was lower in sites providing mapping (mean difference 10.9, 95% CI 0.7-21.1; p=0.04) and person-centred care (13.6, 3.3-23.9; p=0.01). Compared with usual care, fewer falls were recorded in sites that used mapping (0.24, 0.08-0.40; p=0.02) but there were more falls with person-centred care (0.15, 0.02-0.28; p=0.03). There were no other significant effects.

Interpretation: Person-centred care and dementia-care mapping both seem to reduce agitation in people with dementia in residential care.