Table 5.

Summary of key recommendations for technology interventions.

Category	Key recommendations	Certainty of evidence
Group vs one-to-one	Studies should be designed as group-based interventions, as they appear to better facilitate social connectedness than one-to-one interventions.	Very low
Effectiveness of technology interventions	Certain types of technologies (information and communications technology and videoconferencing) are particularly suitable as interventions for social isolation and loneliness. For best results, studies should be designed to strengthen existing bonds, especially the connections between family members (eg, grandchildren).	Moderate low
Frequency of use	Frequency of use is encouraged (greater use increases the effect size).	Very low
Training	Training, especially in the use of technology, is encouraged as it improves the success of the study.	Very low
Duration	Shorter-duration studies are recommended (shorter studies achieve better results than longer-duration studies).	Low
Outcome measures	The impact of intervention is stronger on social isolation than on loneliness, and studies should be designed to look further on how to impact loneliness. Use of standardized measures such as University of California Los Angeles Loneliness Scale and the Lubben Social Network Scale is recommended.	Low
Mechanisms	Mechanisms by which interventions reduce social isolation through the design of studies, including the gaining of social support, engagement in activities of interest, the making of new connections, and search for new information, should be clearly defined at the outset.	Very low
Usability	Intervention studies should adopt standard measures of usability (eg, System Usability Scale) because the adoption of technology by older adults largely depends on learnability and perceived difficulty of use. These barriers often prevent technology from reducing loneliness in older adults.	Very low