Social isolation and loneliness worsen older peoples' quality of life, risk of dementia, and contributes to 45,000 deaths/year in Canada - as much as smoking. Isolated people use the health care system more often, but have worse outcomes. Effective, inexpensive interventions exist but unfortunately they have not been implemented in Canada.

We partnered with the Australian developer of HOW R U?, an effective and feasible intervention that uses specially trained, older, hospital volunteers to provide peer support to combat isolation and loneliness in isolated older peers.

Little is known about older people's preferences for virtual care (telephone vs. video) nor their relative effectiveness. Thus we will compare two ways of delivering HOW R U: telephone support and a tested, secure user-friendly video conferencing app, aTouch Away® to a common control arm.

We also partnered with Emergency Medicine, Family Medicine, Geriatrics and Psychiatry to identify people who will benefit from peer support; and with Volunteer Services to recruit volunteers.

Inclusion Criteria:

• Any community-dwelling person 70 years of age and older receiving care from the ED, Family Medicine, or Geriatric clinics at our two participating sites (MSH and NYGH) will be eligible.

Exclusion Criteria:

• age less than 70 years; communication problems (critically ill, unconscious, language barrier, speech impairment or otherwise unable to provide consent), or admission to a hospital for > 72 hours. Patients with severe cognitive impairment or those living in nursing homes who are dependent on others for their activity of daily living will be excluded. To target the intervention for older people with loneliness a baseline de Jonge loneliness scores of 2.0 will be required for participation in the trial.

We will use intention-to-treat analysis for the primary outcome to further protect against selection bias. It is not feasible to blind participants to the intervention. Primary outcomes will be assessed after 12 weeks by a research assistant who has had no previous contact with participants and is blinded to the intervention group.