Health records may not tell the full story of a person’s health, particularly for people experiencing homelessness. Researchers in London are hoping artificial intelligence can help change that. A team led by Stephanie Dixon, a research scientist at the London Health Sciences Centre Research Institute and ICES Western, is examining how existing health data can be used to better understand the needs of people who are unhoused. The project draws on anonymous health records from more than 35,000 Ontarians with a history of homelessness. Dixon says the first step is determining whether the information routinely collected by the health care system reflects the realities of people without stable housing. “Individuals who are assumed housed may interact quite differently with our health care setting than those who are experiencing homelessness,” Dixon said. The research will look at how health records can be interpreted more effectively before using that information to develop tools for patients and their care providers. One potential outcome is a new frailty score, which helps health-care professionals assess a person’s overall health and likelihood of developing serious complications. Current assessment tools may overlook the cumulative effects of homelessness, including trauma, chronic stress, mental health concerns, substance use and difficulties accessing medical care. By working with community organizations, health care providers and people who have experienced homelessness, researchers hope to identify factors that existing assessments miss. Dixon said the team wants to establish clear guidelines for using health data before building tools intended to improve care. “Once we have those rules in place, we can build on those to develop tools that will help this particular community as well as the individuals who are caring for them,” she said. The long-term aim is to help health-care providers identify risks sooner and connect people with appropriate treatment and support. The project is supported by the AMS Healthcare Leadership in Compassion and Artificial Intelligence Fellowship, which supports research into the responsible and compassionate use of AI in health care.