Introduction:  Studies show that housing instability, and homelessness in particular, are associated with poorer health. (Dormer D, 2018)  The relationship is complex and multi-directional. Housing instability, along with challenges to other social determinants of health (SDOH) such as food, education, employment and criminal justice, contributes to poorer health, and, conversely, poor health contributes to increased anxiety about SDOH including housing.  While health objectives, along with other non-housing social objectives, have long been considerations for housing programs, such consideration has been for the primary purpose of enhancing the goal of eliminating homelessness, less for the alternative purposes of improving health outcomes or lowering cost.  However, over the last few decades, and particularly since 2017, there has been a policy shift towards an integrated continuum of care that spans physical health, behavioral health and social supports.  A key component of that shift has been an increased focus on housing interventions for the express purpose of improving health outcomes and lower cost.  The most common program model for providing housing to chronically homeless people is permanent supportive housing (PSH) following the Housing First practice model.   

Although some studies have shown that some permanent supportive housing programs that follow the Housing First model can lead to a significant improvement in health service utilization, health outcomes and lower cost, a comprehensive review of the evidence by the National Academies of Sciences, Engineering, Medicine published in July, 2018 “…found no substantial evidence that PSH contributes to improved health outcomes, notwithstanding the intuitive logic that it should do so…” (National Academies of Sciences, Engineering and Medicine, 2018)

On the other hand, in April, 2019, Manatt Health, in a study for the Robert Wood Johnson Foundation, summarized the trend from the healthcare perspective towards an integrated continuum of care that includes SDOH: “States are entering a new phase of work related to SDOH:  Moving beyond simply referring people to social supports and looking to ensure that people receive the services they need; as well as engaging Medicaid managed care companies more fully as partners in addressing social factors.  Increasingly, SDOH interventions will be more closely integrated to the delivery of care.” (Manatt Health, 2019) [Emphasis added.] 

These somewhat inconsistent if not conflicting perspectives raise interesting and important questions:  If the evidence does not show that PSH contributes to improved health outcomes, does that mean that the underlying Housing First model is flawed and doesn’t consistently achieve the stated results or does it mean that the HF model is valid as an evidence-based practice but is not being consistently adhered to by PSH programs?  Or does it mean, simply, that the wrong metrics are being tracked?  Also, if policy makers, healthcare providers and payers, particularly Medicaid and Medicaid’s affiliated managed care organizations, see a trend towards a more fully integrated continuum of care, how does this trend look from the perspective of the PSH providers?  Do housing service providers share the same view of an integrated continuum of care, with the same goals and coordinated EBP as healthcare providers and payers? And, finally, what role does technology play in achieving the goals of the integrated continuum of care at the intersection of housing and health?

To begin to answer these important questions, we conducted a short national survey of permanent supportive housing programs that follow the Housing First model.  This report summarizes the results of that survey.

Objectives:  To better understand the extent to which housing providers that offer permanent supportive housing services view their objectives as aligned with healthcare providers and payers, as a part of an integrated continuum of care.  Also, to better understand how permanent supportive housing services in the real world align with Housing First best practices.  Finally, to better understand the role that technology plays in meeting combined housing and health objectives as well as the perceived barriers to the use of technology in support of the Housing First permanent supportive housing model.  We hope this improved understanding can inform policy makers, payers and providers across the care continuum leading to improved outcomes for both housing and health. 

Methods:  We conducted a Google search using the phrases “permanent supportive housing” and “housing first” to identify agencies that provide PSH services adhering to the HF model.  In addition, knowing that in November, 2018, California passed Prop 2, the No Place Like Home Act of 2018, providing funds for every county in California to offer permanent supportive housing services targeting homelessness and mental health based on the Housing First model, we searched specifically for PSH housing programs in every county in California. We prepared a 10 question survey using Survey Monkey which was sent to a representative of every organization that met our search criteria.  The survey was open from Monday, April 15 to Friday April 19, 2019. We performed an analysis of the results to address four questions:

• Are the metrics that housing providers use to measure participation and outcomes consistent with those used by healthcare providers and payers?
• Are the services included in permanent supportive housing programs consistent with Housing First best practices?
• What is the present state of technology use by PSH programs and what are the perceived benefits and potential barriers to the use of technology to achieve their objectives?
• Are housing organizations aware of the funding opportunities afforded by the SUPPORT Act of 2018, and are they pursuing a strategy to leverage this opportunity?

Results: We received 40 responses from 184 invitations, covering 16 of 23 states.  Following is a summary of the results:

• 75% of organizations have case managers but fewer than 33% of organizations have other positions identified in the Housing First EBP such as staff for mental health, employment, addiction or offender reentry.
• 75% of organizations align with other mental health and food or nutrition programs, about 67% align with employment, addiction and physical health service providers and 50% align with offender reentry service providers.
• 87% measure participation in their programs by the number of people in the program at any time. 50% measure person-days over time.  Fewer than 33% use other participation measures, such as those related to addiction, employment, education or offender re-entry.
• 82% measure success by the number who transition to permanent housing outside of the PSH program, 54% measure success by the rate of recidivism. 27% measure success by the change in emergency department visits and outpatient health services utilization, and 7% measured cost.
• 67% of organizations use little or no technology in their programs, while 33% use technology more extensively.
• 26% of organizations indicated that adopting technology for consumer engagement and care collaboration was of little or no value, while 44% said they would be interested if the revenue models supported the cost and effort, and 30% said they are currently looking for or developing such technologies.
• The primary barriers to adopting new technologies for consumer engagement and care collaboration are: 1)  high cost versus low offsetting revenue, 2)  complexity and uncertainty about requirements and standards.
• 72% of organizations had little or no familiarity with the SUPPORT Act and the implications it might have for their programs, while 28% said they were interested or already working on a strategy to address the opportunities presented by the SUPPORT Act.

 Discussion:  All of the respondents listed themselves as permanent supportive housing organizations, which means that they are designed to provide housing supports without a specific expiration period, so it is interesting that the most common measure of success was the number of people who exited the program.  At least on the surface, this appears to be inconsistent with the goals of the Housing First model for PSH, at least as expressed in some of the Housing First guides, which prioritize on-going participation in housing and supporting services programsn without a focus on exiting the program.  Also, the second most common measure of success, at 54% of the organizations, was the rate of recidivism.  However,  only 12.5% track participation in reentry programs as a measure of engagement and only 17.5 percent have someone on staff focused on offender reentry.  With respect to collaboration with healthcare providers, only 27.5% of agencies measure success by the change in health services utilization, such as emergency department visits or in-patient days, and only 17.5% measured the cost of services as an indicator of success.  Finally, given the financial pressure that most of these organizations face, it is significant that the levels of awareness or interest in the SUPPORT Act, that could provide financial incentives for innovative services and may be the catalyst to new models for sustainability,were quite low.

Conclusion:  These results suggest a significant gap exists between how healthcare providers view integration with social service providers generally and PSH providers in particular, and how PSH providers view progress towards integration with healthcare providers.  Although the Housing First model has been in practice since the early 1990s, in some ways the PSH market is still in its formative stage, much as healthcare IT was before HITECH and the Affordable Care Act that mandated adherence to evidence-based practice and the adoption of technology in ways that other industries have accomplished organically.  As a broad observation, we conclude that the Housing First model operates more as a conceptual framework than as a best practice guide, with many PSH organizations paying less attention to supportive services broadly than is suggested in the literature.  This may account for the divergent results described in the National Academies study.  It is likely that the gap between PSH in practice and the Housing First model as expressed in the guides reflects the traditional funding sources which largely came from the Department of Housing and Urban Development, with its more limited mission that does not include the same level of focus health outcomes or cost compared to the DHHS and SAMHSA vision for Housing First and PSH.  We conclude that the integration of housing, along with other SDOH services, is inevitable, but that closing the gap between healthcare and SDOH services like PSH will require a combination of policy initiatives with funding and mandates (perhaps as seen in the SUPPORT Act), combined with technologies that simplify and address the concerns over complexity and risk, and leadership, most likely coming from the Managed Care Organizations acting on behalf of state Medicaid agencies.