The Integrated Wellness in Supportive Housing Model (IWISH) Shows Promise in Supporting Aging in Place

The publication of the first phase of the evaluation of HUD's IWISH demonstration provides a comprehensive but not yet complete examination of the federal government's demonstration to support aging in place among older, low-income adults living in federally assisted multifamily housing. The IWISH demonstration tests a supportive, place-based housing model that delivers enhanced service coordination to residents onsite through a full-time wellness director and a part-time wellness nurse. The demonstration involves 40 treatment sites and 84 control sites in a cluster-randomized controlled trial to determine the model's impacts on the health and housing stability of adults aged 62 years and older who reside in multifamily housing. IWISH marks the first time a HUD demonstration has included the cost of a part-time nurse at the project site. The first phase report draws on two previous interim implementation reports and a new impact analysis to summarize the model's impacts at 3 years.

Although the demonstration period is ongoing, the report nevertheless provides an in-depth analysis of resident and staff experiences with the program, offering a window into the challenges that residents faced at treatment and control sites. Because this demonstration has a clear health focus, the report provides considerable insight into the health of the older adults that reside in multifamily assisted housing--the report confirms the prevalence of multiple chronic health conditions among the study population and the need for supportive assistance. It also describes important mitigating factors that complicated the analysis of the first phase impacts and that will be addressed more fully in the subsequent and final report. Importantly, this report lays the groundwork for discussing how to support aging in place in multifamily housing once the demonstration concludes.

What is the IWISH Model, What are its Goals, and How is Success Measured?

Older households are a growing share of HUD's housing assisted population. Most older adults prefer to age in their homes but require additional supports. Congress supported launching the IWISH demonstration to test whether an enhanced service coordination and wellness model delivered at multifamily assisted housing sites could help residents age in place, increase the housing tenure of older residents and delay the transition to long-term care. The demonstration also was designed to examine whether the evidence-based health and wellness programming that the Resident Wellness Director and Wellness Nurse were expected to coordinate at the housing site, and the services they were required to provide, would increase residents' use of planned primary care and reduce their use of unplanned emergency care such as unplanned hospitalizations, outpatient emergency department visits, and ambulance events.

The IWISH model includes a full-time Resident Wellness Director and a half-time, nonclinical Wellness Nurse onsite at the property to coordinate health and social services and implement six core components:

• Proactive engagement with residents to maximize participation and ensure that residents understand the benefits of IWISH.
• Standardized resident health and wellness assessments at enrollment and annually thereafter so that Resident Wellness Directors and Wellness Nurses gain a comprehensive understanding of resident needs and priorities.
• Individual and community plans for healthy aging to help residents identify and meet wellness goals.
• A web-based data platform to record information about IWISH participants, assessment data, healthy aging plans, and the use of programming and service coordination.
• Partnerships with health and social service providers to coordinate services and enhance opportunities and resources for residents.
• Evidence-based health and wellness programming to address the needs of residents. Each IWISH property received $15 per unit each month to help meet IWISH participants' health and wellness needs and support wellness programming.

Researchers examined IWISH's effects on residents in four areas: housing exits and resident tenure, transitions to long-term institutional care, the use of Medicare- and Medicaid-covered primary care and other nonacute healthcare services, and the use of Medicare- and Medicaid-covered unplanned hospitalizations and other acute care. For each of these areas, researchers selected primary and secondary outcomes. For example, the total number of days that residents were admitted to a hospital for unplanned acute care services during the demonstration period was used as a confirmatory measure of whether the IWISH model affected the use of acute medical care. Researchers also incorporated four secondary measures related to the use of acute care, including unplanned hospital admissions and unplanned hospital readmissions within 30 days of a previous hospital discharge. The latter measure, the authors note, is an important indicator of the overall quality of care for patients discharged from a hospital; fewer hospital readmissions would reflect both improvements in the quality of discharge planning at the hospital and more coordinated care that residents of IWISH properties received after they were discharged. Researchers also expected that reductions in the number of emergency department visits and ambulance trips could signal the IWISH model's longer-term potential to limit adverse outcomes among residents and reduce spending on high-cost healthcare services. The confirmatory measure for housing tenure is the end of residency for any reason.

The researchers assessed the model's impacts on resident housing and healthcare outcomes by analyzing quantitative and qualitative data, including matched housing and healthcare administrative data as well as interviews with residents and staff at treatment and control properties. Researchers also constructed a rating system to score how effectively IWISH properties implemented the model during the first phase. The study team used HUD administrative data and Medicaid and Medicare claims data to compare housing tenure and healthcare use outcomes between residents of IWISH and control properties. The team measured the impacts of IWISH on healthcare utilization and resident tenure by linking individual-level HUD Tenant Rental Assistance Certification System data with Medicare and Medicaid fee-for-service claims data.

Two factors significantly affected the analysis of the first phase: first, researchers did not have access to some Medicare data, including data for residents enrolled in managed care plans under Medicare Part C as well as Medicare Advantage; second, the onset of the COVID-19 pandemic during the third year of the demonstration significantly disrupted the model's implementation and influenced the delivery of healthcare services.

Summary of Findings After the First 3 Years

The results of the impact analysis are reported below, but the report's authors caution that the 2-year extension period for the demonstration is not yet incorporated, among other mediating factors.

The report states, “The impact study did not find evidence that IWISH prolonged resident tenure, reduced transitions to long term care, increased primary care, reduced unplanned hospitalizations, or affected other types of healthcare use... it found no significant impacts of the IWISH model across any confirmatory and secondary outcome measure during the first 3 years of the demonstration.” IWISH did, however, reduce the rates at which residents aged 62 to 64 and 85 or older used acute or emergency care services. (It had no impact on this for residents aged 65 to 84.) The authors note that interpretation of the impact findings of the First Phase evaluation is limited by several factors:

1. “The statistical power of the analyses of IWISH impact on healthcare use was diminished because the study team was unable to access Medicare data for about one-half of the residents in the demonstration sample, those who were enrolled in managed care plans under Medicare Part C.”

2. “The impact findings are limited to the first 3 years of implementation of the IWISH model, which was curtailed by a 6-month delayed start of resident enrollment, and which coincided with the onset of the COVID-19 public health emergency, which had a fundamental impact on healthcare delivery and use.”

3. “The information collected on model fidelity showed that not all properties implemented all components of the IWISH model or implemented them fully and that several properties had vacancies in one or both IWISH positions for several months during the demonstration period.”

4. “Many of the same health and wellness services offered by IWISH were also available via service coordination and wellness programming at the active control properties and, presumably, the passive control properties. Therefore, residents at those control properties may have experienced some benefits similar to those at the IWISH properties.”

The authors add that although the impact study found no clear evidence that the IWISH model had positive impacts on health care use, the qualitative data suggested that staff and residents who participated in the program benefited from the model and that IWISH's impact could increase over the long term. For example, the researchers reported that, based on interviews with Wellness staff, IWISH staff helped residents during health emergencies, supported residents both during and after emergencies, and taught residents how to prevent emergencies through earlier disease prevention and treatment. They also reported helping residents who were frequent users of emergency care services to seek more appropriate care.

Despite the lack of statistically significant impacts, researchers found evidence of positive effects that were exploratory in nature but for which causation cannot be assigned. (The evaluation was not designed to test statistically for IWISH's impacts on small subgroups.) The study's exploratory subgroup analyses show that relatively younger and older residents tended to benefit more from IWISH than did other age groups, suggesting that IWISH benefited those with greater healthcare needs. The exploratory analyses also found positive correlations between some outcome measures and two specific aspects of the IWISH wellness model: 1) working individually with residents to meet their health and wellness goals and 2) providing transitional care to residents returning home from a stay in a hospital or long-term care facility. Residents of IWISH properties that were rated (in the fidelity analysis) as having higher levels of working individually with residents to meet their health and wellness goals tended to spend more days in the community (meaning that they were residing in their property and not admitted to a hospital, a long-term care institution or an emergency department) than did residents living in IWISH properties with low levels of working on individual resident goals. And residents of IWISH properties that were rated as having higher levels of providing transition care to residents returning home from a hospital or long-term care stay tended to have fewer unplanned hospitalizations, fewer outpatient emergency department visits and ambulance events, and more primary care visits than did residents of lower-rated IWISH properties, suggesting that the way the site implemented the full model mattered.

Implications and Further Research

The impact analysis findings of the first phase didn't find evidence that IWISH prolonged resident tenure; residents of IWISH and control properties were equally likely to end their residency during the first 3 years of the demonstration. On average, 27 percent of residents in both groups exited their property during the 3 years, including 10 percent who died, 2 percent in both groups who exited to long-term care, and 15 percent who moved out for other reasons (such as eviction or a move to another HUD-assisted property or a family member's home). The researchers also did not find that IWISH reduced transitions to long-term care, increased the use of primary care, reduced unplanned hospitalizations, or affected other types of health care use. The authors mention that the IWISH model could benefit residents in ways that are not evident in the administrative data or that might become significant over a longer evaluation period. The full set of research questions will be addressed in the second phase of the evaluation, when data for the entire study population, including the portion dropped from the first phase because healthcare data were not available, will be included in the analysis. The final evaluation report, expected in 2026, will assess the impact of the full 6-year demonstration. The report will include Medicare Advantage data, which will enlarge the analysis sample for health care use of residents in IWISH properties and the control group. In the meantime, this report can and should support discussions about what actions should be taken when the IWISH demonstration draws to a close.