Read our COVID-19 research and news.

A new report questions the impact of support services that go beyond providing stable, safe shelter for the homeless.

Ted S. Warren/AP Photo

Does ‘supportive housing’ help the homeless with medical needs? Not clear, study says

A new report finds no evidence that efforts to provide support for the chronically homeless improve their health or reduce the cost of their care. But that doesn’t mean the U.S. government should stop funding programs that offer permanent supportive housing (PSH), say the authors of the study, conducted by a panel convened by the National Academies of Sciences, Engineering, and Medicine. Instead, the report calls for more analyses of what types of interventions work best for which populations.

“As a former clinician and public health official it would seem logical, if not obvious, that if you can keep people safe with housing they should have better health outcomes,” says the report’s chair, Kenneth Kizer, who leads the Institute for Population Health Improvement at the University of California, Davis. “Indeed, one of our recommendations is that we need to increase the supply of permanent supportive housing.”

The chronically homeless represent less than 20% of the 550,000 homeless people in the United States. But that group tends to require more social and medical services than those with even occasional access to safe and stable housing environments do.

PSH was designed to address such “housing sensitive conditions” as ensuring that diabetics have a refrigerator to keep their insulin or that cancer patients receiving chemotherapy have ready access to a bathroom. The committee, which issued its report on 11 July, cited studies showing that PSH is effective in serving the medical needs of those with HIV/AIDS. However, it says similar evidence doesn’t exist for other conditions. In fact, Kizer and his colleagues were sorely disturbed by the spotty state of the research.

“The committee was disappointed to find that the existing literature lacks information on the type, intensity, frequency, or length of the needed services,” they write. “The lack of data about these things effectively precluded generalizing to who among individuals experiencing homelessness are most likely to benefit from the services and different models of PSH.”