Supportive housing didn’t cut deaths
Second phase of study hopes for better health outcomes, care savings
Denver’s initial experiment in housing some of the higher- risk population living on the street didn’t reduce deaths, but the groups running the program’s second phase think they can improve health outcomes while saving taxpayers money.
In 2016, Denver launched a supportive housing program for people who had been homeless for an extended period and had been arrested at least eight times in the previous three years but weren’t facing significant prison time.
A study that followed 275 participants and a similarly sized group that didn’t receive supportive housing over three years found that the housed group received more outpatient care and had fewer visits to emergency rooms, but hospitalizations and deaths remained largely unchanged.
Private investors paid about $ 8.6 million to fund the supportive housing program through what became known as the Social Impact Bond. The city ultimately paid back $ 9.6 million, under a formula based on results. At the time, Denver was looking to reduce spending on jails, emergency shelters and drug detox facilities, and the finding that people went to the emergency room less often was a bonus, said Cathy Alderman, chief communication and public policy officer at the Colorado Coalition for the Homeless.
The second phase of the bond started in 2022, with improving health as a major goal this time. Private investors paid $ 11.75 million to move high- risk people into housing, and the U. S. Department of the Treasury pledged to give Denver $ 5.5 million toward repaying that money if the program saved money for Medicaid and Medicare. The repayment amount depends on how many participants stay stably housed and how much their jail time goes down through the end of 2029.
The Social Impact Bond program is separate from other homelessness initiatives in Denver, including moving people into repurposed hotels, which the city expects to cost $ 40 million this year. It involves a relatively small number of the more than 9,000 people the Metro Denver Homeless Initiative estimated were unhoused when it conducted the annual point- in- time count in late January 2023.
The Colorado Coalition for the Homeless houses about 120 of the current participants, and WellPower, formerly known as the Mental Health Center of Denver, houses 25. Participants receive visits at least once a week from a case manager, social worker, nurse or mental health provider.
This time, the participants not only have a history of homelessness and incarceration but also are relatively frequent emergency room patients, Alderman said.
It’s a high- need group, but also a group the coalition is used to working with through its clinics and street medicine program, she said. Information on outcomes from the first
year of the current program isn’t yet available.
“We’ve identified the individuals, and we’ve gotten them into housing. So we’re on the path,” she said.
In the initial round of the Social Impact Bond, about two- thirds of the people in the housing and control groups had a documented substance- use disorder, which is the medical term for drug or alcohol addiction. More than one- third also had a diagnosed mental health condition.
People who received housing had about 40% fewer arrests over three years than those who didn’t and reduced their time in jail by 38 days on average, which met the project’s goals. They also had fewer stays in detoxification facilities.
They also received about 10 more office visits for mental health treatment and had eight fewer emergency room visits than those who received usual services over three years. The housing group did have two more hospitalizations on average than the usual- care group, but the researchers who studied the program found that the number of hospitalizations was small enough that it could have been due to chance.
Interestingly, people who didn’t receive housing were more likely to be enrolled
in Medicaid, which the researchers attributed to efforts to enroll people in health coverage when they were about to get out of jail. Death rates were similar for both groups.
Three years may not be long enough to see a significant difference in death rates, and having housing doesn’t solve all of a person’s problems, such as being unable to afford healthy food and pay for medications, said Dr. Elizabeth Bowen, an associate professor at State University of New York- Buffalo’s School of Social Work.
People who are homeless have higher rates of most health conditions than housed people and have less ability to manage them, while also facing threats of violence and the effects of extreme cold and heat, which all contribute to them dying younger than the general population, she said.
That said, the lesson isn’t that supportive housing doesn’t work, but that it can achieve more if people move off the street before their health has deteriorated too much, Bowen said.
“People are often coming into supportive housing in really poor health,” she said. Housing “might not solve every problem, but it does help with a lot of issues.”