Experimental studies that evaluated interventions targeting social isolation and/or loneliness in older people were eligible for inclusion. Studies had to include an inactive control group (usual care, no intervention, attentional). Studies were required to report a clear aim and sufficient outcome data to allow for treatment effect to be obtained.
A variety of participants was targeted in the included studies; most were women. The mean age of participants ranged from 41 to 85 years. Less than half of the interventions were targeted at people who had been assessed as socially isolated or lonely people. More than half of the interventions were delivered in groups, and a third were given one-to-one. Most interventions were delivered weekly or fortnightly, and generally lasted from six weeks to a year. Most programmes were delivered by a health or social care professional. Nearly half were support interventions (discussion, counselling, therapy or education). Other interventions included social or physical activities, Internet training, home visiting or service provision. Most programmes were participatory (requiring active input from participants involving social contact) and were delivered in the USA.
Definitions of outcomes varied, and were classed within three domains: social health, mental health and physical health. Outcome data were collected using a range of validated and non-validated outcome measures.
Titles and abstracts were screened by one reviewer; a second reviewer was consulted when needed. Two reviewers assessed full texts of potentially eligible studies. Disagreements were resolved by consensus.