2-15-18 final SSD Supporting Statement B

Supporting Statement for Paperwork Reduction Act Submissions

HUD Supportive Services Demonstration/Integrated Wellness in Supportive Housing (IWISH)

OMB Number 2528-New

Part B- Collections of Information Employing Statistical Methods

1. Respondent Universe and Sampling Methods

The Fiscal Year (FY) 2014 Consolidated Appropriations Act gave the U.S. Department of Housing and Urban Development (HUD) authority to develop a demonstration to test a model of housing and supportive services for low-income elderly residents in HUD-assisted housing. In FY 2015, HUD announced the availability of a funding opportunity under the Supportive Services Demonstration that will provide grants to property owners to participate in the demonstration.

This information request is for the Integrated Wellness in Supportive Housing (IWISH) Resident Assessment. The IWISH Resident Assessment will collect self-reported information from low-income elderly¹ tenants residing in HUD properties selected to participate in the Supportive Services Demonstration (SSD). Forty properties across seven states were selected for the treatment group of the demonstration (see Table 1 for breakdown of properties by state). This information collection will affect approximately 4,249 individuals residing in units of 40 funded demonstration sites. All residents enrolled in the SSD will be asked to participate in the program and sign the informed consent. All those enrolling will complete a IWISH Resident Assessment within 60 days of consent in the program. HUD expects a response rate to the assessment of nearly 100% of residents who sign the informed consent to participate in the demonstration. This survey will comply with all requirements to be accessible for both persons with disabilities, in addition to respondents with limited English proficiency

The IWISH Resident Assessment is a critical element of the demonstration. Typical housing properties do not conduct standardized assessments on all residents and are not aware of the health/functional status, current service use and service gaps of these individuals. This lack of information has historically stymied staff efforts to design an appropriate package of programs and to assist in linking residents to needed community-based services. IWISH Resident Assessment data will provide critical information that the enhanced service coordinator/wellness nurse team will use to decide on the most appropriate external partners and will help the team identify various at-risk groups within the property that need service and wellness coordination and that are most likely to be of interest to potential health care and social services collaborators. Information from resident assessments will also feed into the development of healthy aging and service and wellness coordination plans that will guide which services are needed and how these services are most efficiently delivered, both internally and with external partners. These data will also be used by SSD Implementation Team, contracted by HUD to monitor program implementation quality and fidelity to the demonstration design.

Since the results of the IWISH Resident Assessment will be used to by staff to determine resident health and social services needs and appropriate services, we expect that nearly all participants will complete the assessment. However, residents are not obligated to participate in the Demonstration or receive services, and their tenancy or receipt of services is not dependent on participating. The SSD Implementation Team will not employ any statistical methodology for stratification or sample selections as a selection of demonstration sites is being conducted by HUD. All responses will be reviewed by property staff to inform their activities and analyzed by The SSD Implementation Team, HUD’s contracted implementation vendor to monitor program performance and inform technical assistance efforts.

Table 1

State	Number of Properties	Number of Residents
California	15	1,434
Illinois	5	821
Massachusetts	6	582
Maryland	2	131
Michigan	7	773
New Jersey	3	194
South Carolina	2	314
Total	40	4,249

2. Procedures for the Collection of Information

On January 20, 2016, HUD announced the availability of a funding opportunity (Notice of Funding Availability or NOFA) under the Supportive Services Demonstration for Elderly Households in HUD-Assisted Multifamily Housing, which made available approximately $15 million in three- year grants to owners of multifamily properties to implement the demonstration.

Eligible applicants were owners of elderly designated or restricted Multifamily properties with at least 50 assisted housing units occupied by eligible tenants (households consisting of one or more persons of whom one is at least 62 years of age or older); up to 10% of the units could be occupied by a person with a disability under the age of 62. Properties with and without current resident service coordinators could apply. Eligible HUD-assisted housing types included housing assisted under Section 202 of the Housing Act of 1959, housing with project based Section 8 assistance (including USDA Section 515 rural housing projects), housing insured under Section 221(d)(3), and housing assisted under Section 236 of the National Housing Act. HUD received over 700 applications.

HUD designed a rigorous, cluster-randomized controlled evaluation of the Supportive Services Demonstration. The prospective design was included in the demonstration NOFA, such that the implementation roll-out will accommodate a robust and simultaneous evaluation. Eligible properties were randomized to one of three groups: a treatment group or one of two control groups—active and passive. The treatment group will implement the demonstration model in a standardized way across all sites. Properties in the treatment and active control groups are required to participate in the evaluation as a condition of award. The active control group properties will know that they are participating in the evaluation and will be paid a participation fee of $5,000 per property. The first payment of $2,500 was paid upon the execution of Cooperative Agreements in September 2017. The second and final payment of $2,500 will be paid at the end of the Demonstration upon the completion of the grantee performing milestones of site visits, telephone interviews, and focus groups that are outlined in the cooperative agreement. As detailed in the Cooperative Agreement, allowable program costs are those costs which are necessary and reasonable for the Grantee’s participation in the Control Group and include, but are not limited to, staff hours devoted to preparation and/or participation in site visits, interviews, or resident focus groups. The passive control group properties will not know that they are participating and data for residents will be obtained from administrative sources.

Forty properties were selected into the treatment group chosen not to exceed the NOFA budget. Eighty properties were selected for the control group, evenly split between active and passive controls. A few additional properties were drawn into waitlists for inclusion in the treatment and control groups to account for post-implementation attrition due to ineligibility or refusal.

Properties were randomized to the different groups using the following methodology:

1. A total of 756 applications were submitted for funding consideration. As noted in the NOFA, HUD expected to constrain the demonstration to a select number of states based on the pool of applicants and property characteristics. After screening all applications for eligibility, HUD ranked states for the demonstration based on three factors: The number of properties with and without a current service coordinator; the rate of county-level fee-for-service enrollment among Medicare beneficiaries; and the number of Centers for Medicare and Medicaid Services (CMS)-funded initiatives improving care coordination for elderly persons. Based on the rankings, 185 properties across seven states (California, Illinois, Massachusetts, Maryland, Michigan, New Jersey, and South Carolina) were eligible for randomization into the three groups.

2. Prior to randomization, CBSAs were used as a geographic unit to stratify properties. CBSAs across the country vary in numerous ways, including in measures of access to and cost of healthcare services and access to social services. Stratification on CBSAs was adopted to help ensure that the treatment and control groups are balanced, not just on the observed characteristics for which balance can be checked, but also for unobserved characteristics of the locality. CBSAs with too few properties to treat as independent strata were combined within states to form one larger stratum. Three properties were considered outliers and dropped before randomization, leaving 185 properties across 14 strata.

3. A set of importance weights were created to account for desired characteristics—such as higher fee-for-service Medicare penetration weights—of the properties selected for the demonstration.

4. Random sampling was executed using the default random-number generator in Stata 14. An unequal probability sampling without replacement algorithm was implemented using the random systematic sampling technique developed by Hartley and Rao (1962). For each of the 14 strata, 3x44/182*(number of properties in stratum) properties using the importance weights and the unequal probability sampling without replacement algorithm were drawn. Simple random sampling was then used to allocate selected properties in each stratum into treatment, active control and passive control groups.

5. Forty properties were randomized to the treatment group, 40 to the active control group, and 40 to the passive control group which also serves as the active control group waitlist. Three properties were randomized to the treatment group waitlist. Properties on the waitlists were assigned a rank to dictate the order in which any waitlist properties would be drawn into the treatment and active control groups.

Post-randomization, three properties from the treatment group and one property from the active control group were either deemed ineligible or declined participation. Sites were replaced from properties from the ranked wait lists.

Forty properties across seven states were selected for the treatment group of the demonstration, the residents of which will be asked to participate in the Resident Assessment.

All respondents will be presented with an IRB approved informed consent form prior to participation in the demonstration. It is estimated that approximately 4,249 residents will be impacted. Upon consent, participants will be asked to complete a IWISH Resident Assessment within 60 days of enrollment in the demonstration. The assessment is expected to last an average of 90 minutes. Trained Resident Wellness Directors and Wellness Nurses (property staff) will conduct the IWISH Resident Assessment face-to-face in a private setting.

Data will be tracked in a demonstration specific web-based platform that allows information to be updated periodically, as sentinel events or changes in health status occur, or during the annual re-assessment. All data collection and analysis will be performed in compliance with OMB and Privacy Act requirements.

3. Methods to Maximize Response Rates and Deal with Nonresponse

Participation in the demonstration is contingent upon an individual’s signed consent. The SSD Implementation Team expects a 90-100% assessment response rate, with a target rate of at least 80% at each property, for each year a demonstration site receives demonstration funds, or 3,399 individuals to consent to the program and complete the IWISH Resident Assessment.

All demonstration sites were required to agree to complete a IWISH Resident Assessment with all interested participants within their cooperative agreements.

Property staff will input IWISH Resident Assessment results directly in the web-based platform. The platform will alert staff when an assessment is not completed with 45 days. It will also provide property staff alerts for incomplete assessments. Staff will continue to receive alerts until the assessment is completed. Additionally, the Implementation Team will monitor response rates and completion monthly and work with property staff, as necessary to achieve the required response rate.

4. Tests of Procedures or Methods to be undertaken

Approximately 33% of the assessment questions included in the IWISH Resident Assessment were taken from previously validated instruments.

The following table details the number of questions in the IWISH Resident Assessment pulled from validated instruments.

Survey Instrument	Developing Entity	Validated Instrument	Section	Description of Instrument and Our Use
Morisky 8-Item Medication Adherence Questionnaire2			Medications	The Morisky 8-item Medication Adherence Questionnaire is a widely used scale to measure medication-taking behavior. We will use this scale to assess medication adherence.
PHQ-23	Drs. Robert L. Spitzer, Janet B.W. Williams, Kurt Kroenke, and colleagues		General Health Assessment (Mental Health and Substance Use)	The PHQ-2, comprising the first two items of the PHQ-9, inquiries about the degree to which an individual has experienced depressed mood and anhedonia over the past two weeks. The purpose of the PHQ-2 is not to establish a final diagnosis or to monitor depression severity, but rather to screen for depression. Patients who screen positive should be further evaluated with the PHQ-9 to determine whether they meet criteria for a depressive disorder. We will use this instrument as an initial screen for depression which will trigger a full screen (PHQ-9 or the GDS-S, based on age) as needed.
GAD-24	Drs. Robert L. Spitzer, Janet B.W. Williams, Kurt Kroenke, and colleagues		General Health Assessment (Mental Health and Substance Use)	The GAD-2, comprising the first two items of the GAD-7, is a short and efficient screening tool for detecting generalized anxiety disorders. We will use this instrument to screen participants for anxiety, which will trigger the full screen (GAD-7) as needed.
MDS 3.0 Section I5	Center for Medicare and Medicaid		Diagnosis	The Minimum Data Set (MDS) is a component of the IWISH Resident Assessment Instrument (RAI) which forms the foundation of a comprehensive assessment for all residents of nursing homes certified to participate in Medicare or Medicaid. Items in Section I are intended to code diseases that have a relationship to the resident’s current functional status, cognitive status, mood or behavior status, medical treatments, nursing monitoring, or risk of death. We will use this tool to generate an updated, accurate picture of the resident’s health status.
Pain Assessment Scale6	Margo McCaffery, Chris Pasero		Vitals	The Pain Scale is a diagnostic tool used to address patients’ pain in all settings and age groups. We will use the pain scale to assess the severity, quality, and location of pain experienced by participants.
Physical Self-Maintenance Scale (PSMS)7	Lawton MP, Brody EM		Activities of Daily Living (ADLs) & Instrumental Activities of Daily Living (IADLs)	The PSMS is a disability measure used for planning and evaluation treatment for elderly individuals. It includes both ADL and IADL items and focuses on observable behaviors. We will use the tool to detect problems in performing activities of daily living and to support care planning and participant goal development accordingly.
PHQ-98	Drs. Robert L. Spitzer, Janet B.W. Williams, Kurt Kroenke, and colleagues		Appendix A	The PHQ-9 is a nine-item depression scale of the patient health questionnaire. It is one of the most validated tools in mental health and can be a powerful tool to assist clinicians with diagnosing depression and monitoring treatment response. The nine items of the PHQ-9 are based directly on the nine diagnostic criteria for major depressive disorder in the DSM-IV. We will use this form to identify the presence of depression in participant under age 65 years.
Geriatric Depression Scale –Short Form (GDS-S)9	J.A. Yesavage and J.I. Sheikh		Appendix A	The Geriatric Depression Scale Short Form (GDS-S) is comprised of 15 items chosen from the Geriatric Depression Scale-Long Form (GDS-L). It is a self-report measure of depression in older adults. The shortened form were chosen because of their high correlation with depressive symptoms in previous validation studies. We will use this form to identify the presence of depression in participants over 65 years of age.
The DETERMINE Checklist10	The Nutrition Screening Initiative		Appendix B	The Nutrition Screening Initiative (NSI) is a national collaborative developed to address the prevalence of malnutrition among older adults. The NSI DETERMINE questionnaire is used by professionals working with elders in order to assess their risk for poor nutritional status or malnutrition. We will use this instrument to identify warning signs for poor nutrition such as disease, economic hardship, reduced social contact, multiple medicines, and involuntary weight loss/gain.
STEADI Fall Risk Questionnaire11	Greater Los Angeles VA Geriatric Research Education Clinical Center and affiliates		Appendix C	The STEADI Fall Risk Questionnaire is a fall risk self-assessment that is based on both evidence and clinical acceptability and has been initially validated with clinical examination data. We will use this questionnaire to assess participants’ level of risk for falling, which will support property staff in triaging participants as high-risk/high-need.
UCLA Loneliness Scale12	University of California, Los Angeles (UCLA)		Appendix D	The UCLA Loneliness scale was originally a 20-item scale designed to measure one’s subjective feelings of loneliness as well as feelings of social isolation. We will use the three-item scale, which has shown to be highly reliable, to assess dimensions of loneliness including relational connectedness, social connectedness, and perceived isolation.
The Mini-Cog13	Dr. Soo Borson, University of Washington		Appendix E	The Mini-Cog, a composite of three-item recall and clock drawing, was developed as a brief test for discriminating demented from non-demented persons in a community sample of culturally, linguistically, and educationally heterogeneous older adults. We will use this instrument to screen participants for cognitive impairment, which will inform property staff of ongoing care coordination needs.
GAD-714	Drs. Robert L. Spitzer, Janet B.W. Williams, Kurt Kroenke, and colleagues		Appendix F	The Generalized Anxiety Disorder (GAD-7) questionnaire is a seven-item, self-reported anxiety questionnaire designed to assess the patient’s health status during the previous two weeks. The GAD-7 asks about the degree to which the patient has been bothered by feeling nervous, anxious or on edge, and has not been able to stop or control worrying. We will use this instrument to further screen participants at risk for anxiety, based on GAD-2 results.
Short Michigan Alcoholism Screening Test- Geriatric Version (SMAST-G)15	Blow F.C., Brower K.J., Schulenberg J.E., et al.		Appendix G	The SMAST-G is a modified version of the evidence-based Short Michigan Alcoholism Screening Test. Because older adults may show signs of drinking problems that are different than other age groups, this screening tool asks specific questions related to an older population. We will use this instrument to detect "at-risk" alcohol use, alcohol abuse, or alcoholism in older adult participants.

5. Individuals Collecting and Analyzing Data

• Elizabeth Martin, The Lewin Group, (703) 269-5625

• Kristina Rerucha-Azeem, The Lewin Group, (703) 269-5562

• Christina Wu, The Lewin Group, (703) 269-5728

• Joy Oguntimein, The Lewin Group, (703) 269 -5967

• Resident Wellness Directors and Wellness Nurses working at each demonstration site

1 A respondent is considered elderly if they are age 62 or older according to the Fair Housing Act

2 Morisky, Donald, Ang, Alfonso, Krousel-Wood, Marie, and Ward, Harry. “Predictive Validity of A Medication Adherence Measure in an Outpatient Setting.” Journal of Clinical Hypertension 10(5) (2008): 348–354.

3 Kroenke, Kurt, Robert L. Spitzer, and Janet BW Williams. "The Patient Health Questionnaire-2: validity of a two-item depression screener." Medical care 41.11 (2003): 1284-1292.

4 Kroenke K, Spitzer RL, Williams JB, et al. “Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection.” Ann Intern Med 146 (2007): 317–25.

5 Centers for Medicare & Medicaid Services. “Long-Term Care Facility IWISH Resident Assessment Instrument 3.0 User’s Manual, Version 1.14.” (October 2016): Web.

6 McCaffery, Margo, and Chris Pasero. Pain: Clinical Manual, p.16. St. Louis: Mosby, 1999. Print.

7 Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist 1969; 9:179-86

8 Kroenke, Kurt, Robert L. Spitzer, and Janet BW Williams. "The Phq‐9." Journal of general internal medicine 16.9 (2001): 606-613.

9 Sheikh, J. I. and Yesavage, J. A. “Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version.” Clinical Gerontologist, 5 (1986): 165-173.

10 White JV, Dwyer JT, Posner BM, Ham RJ, Lipschitz DA, Wellman NS. “Nutrition screening initiative: development and implementation of the public awareness checklist and screening tools.” J Am Diet Assoc. 1992 Feb;92(2):163–167

11 Rubenstein et al. J Safety Res; 2011:42(6)493-499

12 Russell, D , Peplau, L. A.. & Ferguson, M. L. (1978). Developing a measure of loneliness. Journal of Personality Assessment, 42, 290-294.

13 Borson, Soo, et al. "The Mini-Cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly." International journal of geriatric psychiatry 15.11 (2000): 1021-1027.

14 Spitzer, Robert L., et al. "A brief measure for assessing generalized anxiety disorder: the GAD-7." Archives of internal medicine 166.10 (2006): 1092-1097.

15 Blow F.C., Brower K.J., Schulenberg J.E., et al. “The Michigan Alcoholism Screening Test- Geriatric Version (MAST-G): a new elderly-specific screening instrument.” Alcoholism Clin Exp Res, 16 (1992): 372.

3