Table 2.

Patient-reported quantitative outcomes

Trial	Patient quality of life	Patient symptom severity	Patient self-esteem	Depression	Anxiety	Access to and ability to navigate services
Cheston et al. [18]	–	–	–	Significant decrease p = 0.034 (Cornell scale)	Borderline-significant decrease p = 0.05 (RAID scores)	–
				No significant change p = 0.241 (HADS depression)	No significant change p = 0.071 (HADS-anxiety)
Cheston and Jones [19]	–	–	–	No significant change when adjusted for pre-intervention scores (Cornell scale, BASDEC)	No significant change when adjusted for pre-intervention scores (RAID, BAI)	–
Marshall et al. [20]	Non-significant evidence of improvement (QOL-AD)	Non-significant evidence of a reduction in cognitive functioning (MMSE)	Non-significant evidence of improvement (Rosenberg scale)	No change (Cornell scale)	–	Average contacts with NHS increased then decreased
						Use of social groups and day care increased
Livingston et al. [21]	Non-significant increase (QOL-AD)	–	–	–	–	–
Livingston et al. [22]	No change (QOL-AD)	–	–	–	–	–

HADS Hospital Anxiety and Depression Score, RAID rating of anxiety in dementia, BASDEC the Brief Assessment Schedule Depression Cards, BAI the Beck Anxiety Inventory, QOL-AD quality of life in Alzheimer’s disease, MMSE Mini-Mental State Examination. − denotes that the outcome was not measured by this study