Twenty-five publications assessing 22 programmes were included in the final analysis. The total number of participants is unclear, owing to multiple publications from what appear to be the same study or intervention programme.
The majority of the studies were non-randomised, but most included a control group and/or compared interventions with other interventions; only one study was uncontrolled. The quality scores ranged from 4 to 9 and 8 studies were rated as good quality.
Effects on caregivers.
The majority of studies assessing depressive symptoms reported heterogeneous results. Three of the 4 studies assessing general mental health found significant improvements associated with the intervention. Two of the 3 studies assessing overall well-being found no significant effects. The majority of studies assessing subjective burden found no significant effects. The majority of studies assessing the effects of the intervention on the competence of the caregiver reported heterogeneous results.
Effects on patients with dementia.
Three out of 5 studies reported significant improvements associated with the intervention for a variety of mental health aspects. The 5 studies assessing cognitive function reported conflicting results. Only one of the 9 studies assessing behavioural problems reported that the intervention group was associated with positive effects; the remainder reported no significant effects or heterogeneous effects. The majority of studies assessing behavioural problems and physical functioning reported no significant effects or heterogeneous effects. The majority of studies (8 out of 12) assessing admission to long-stay facilities (usually a nursing home) favoured the intervention group compared with the control group. Evidence on survival was reported in only 2 studies, with conflicting results.