DURG 6/16/2021

From Risk to Value:

Examining San Francisco’s COVID19 Shelter in Place (SIP) Hotels

Elizabeth Abbs, MD Addiction Medicine/Internal Medicine

Mason Lai, 4^th Year UCSF Medical Student

Naomi Schoenfeld, PhD, FNP UCSF Faculty Affiliate, Medical Anthropologist

and SFDPH Nurse Practitioner

OCOH TL Group 10/7/2021

Agenda

• Background/Aims:
• What factors led to this housing “experiment”?
• Methods: Mixed Quantitative and Qualitative Design
• Preliminary Results and Discussion
• Demographics
• Medical Complexity
• Substance use
• Health Care Utilization
• Next Steps:

​

​

Virus predicted to pose deadly threat to persons experiencing homelessness (PEH)��Tourism shuts down��Advocates urge The City to use hotel rooms so PEH can SIP. �

An outbreak in the city’s largest homeless shelter in April 2020 🡪 conversion of tourist into “Shelter-in-Place” (SIP) sites.

​

Recruitment aimed to target the “most vulnerable” with a changing list of social and medical criteria

​

March 2020 COVID19 hits SF

Risk Predicted�

“This is going to be disastrous for people experiencing homelessness…AND there’s going to be lots of hotel rooms available and they are going need to put people in hotels. That’s not some great rocket science.

​

Getting [PEH] a chance to shelter in place the way they’re recommending everybody else seemed a matter of justice and equity”

-Dr. Barry Zevin, Medical Director Street Medicine and Shelter Health, Study Interview Feb 18, 2021.

• To PEH
• Hospital Capacity
• Political/Economic Risk

Advocates Applied Pressure

Project Objectives

• Describe the population of PEH temporarily housed in SIP hotels in San Francisco
• Expand the narratives of PEH:
• Examine care utilization patterns before and after SIP
• Consider value of providing temporary housing for PEH
• Consider forms of housing justice
• Illuminate factors enabling the RAPID SCALE UP of housing options BEYOND the pandemic itself including the role of advocacy, social media, and political economic considerations.
• To better design integrated supportive housing (temporary and permanent) that reduce mortality and promote wellness for

​

​

What is a SIP hotel/motel?

• 25 hotels/motels leased by the city
• Some single some double occupancy, all rooms had private bathrooms (and TV)
• Laundry and cleaning services
• Three meals daily delivered to rooms
• Each site associated with a Community Based Organization (CBO)
• Five Keys, Providence House, WeHope, Urban Alchemy, ECS, Mother Brown
• Healthcare staffing at all sites (mostly SFDPH, one site Healthright 360)
• Nursing staff range from 5 days per week to ½ day per week depending on size and complexity
• Medical providers offered a range of services including limited primary and urgent care, addiction medicine, and mental health treatment.
• Behavioral health, physical and occupational therapists were available after initial nursing/medical request
• Limited case management including housing assessment
• Homebridge in-home health worker support as needed

Distribution of SIP Hotels/Motels

Tenderloin

SOMA*

* Includes largest SIP site

Japantown

Cow Hollow

Union Square

Polk Gulch/

Nob Hill

N = 346

What is known about how homelessness impacts care?

• Homelessness is associated with increased sickness and death.
• People experiencing homelessness (PEH) often forced to choose between immediate needs vs. preventative healthcare²
• Housing First models or other long-term shelter/housing can:
• reduce (psych) ED visits³
• increase preventative care utilization^{1, 2}

​

​

¹Kushel et al. (2001). Jama, 285(2), 200-206.

²Weiser et al. (2013). Journal of general internal medicine, 28(1), 91-98.

³Raven, M. C., Niedzwiecki, M. J., & Kushel, M. (2020) Health services research, 55, 797-806.

⁴Kessell et al. (2006) Journal of Urban Health, 83(5), 860-873.

Can temporarily (and at times chaotic) COVID-forced housing provide similar “value”?

Methods

Methods: Quantitative Database

• Coordinated Care Management System (CCMS)*

​

​

• Whole Person Integrated Care (tab linked to Epic records)
• Race/Gender/Age/Medical Diagnoses (limited)
• Number of years experiencing homelessness in SF*
• Healthcare utilization (ever)
• Jail
• Behavioral health services (inpatient and outpatient)
• Respite
• Sobering Center
• Urgent Care

​

• SFDPH Epic electronic medical records
• Primary Care Provider
• Type of housing prior to SIP
• Healthcare utilization (excluding Dignity Health)

*Existed for 24 years; captures all events within San Francisco Co

​

Medical Complexity

• Elixhauser comorbidity index¹
• WPIC tab
• Other diagnoses in Epic (ICD-10, provider notes)
• Number of medications
• Epic last 6 months

​

¹Elixhauser et al(1998). Medical care, 36(1):8-27. 

Methods: Qualitative

• Participant Observation (field notes - ethnographer worked as nurse practitioner in SIP placement and as a regular medical provider at three sites)    
• Interviews (verbal consent, $25 gift card reimbursement for participation)
• SIP hotel residents (n = 27)
• In-depth interview
• Option for in-depth EMR chart review (N = 9)
• SIP medical and nursing staff (N = 9) 
• SIP CBO staff (N= 4, from two CBOs)
• Focus Group (n = 5)
• RNs with SFDPH: 3 with experience in shelter health prior to SIP
• Media analysis (in process)

​

​

RESULTS � and Discussion

Results: Quantitative Demographics (n = 346)

• Age: mean 52, median 54 years
• Sex: 69% male
• Gender Identity (n = 329): transmale (1, 0.3%), transfemale (9, 3%), non-binary (3, 0.9%)
• Ethnicity: 85% not Hispanic or Latino
• Race: 43% black , 35% white, 7% mixed/other, 6% Asian
• Language (n = 335): 90% English, 7% Spanish
• Years homeless (n = 345): mean/median 10 years (range 0-24*)
• Housing prior to SIP (n = 342):
• 32% street/car, 21% shelter, 5% hospital/respite/sobering, 3% co-living, 38% unknown (not in chart)

​

​

​

Young, new experience of homelessness

​

​

​

Older, decades of experiencing homelessness

​

​

​

​

Engages in Care

​

​

​

​

Declines all care

​

​

​

​

Prior engagement in care (most shelter health or street medicine)

​

​

​

​

Newly engaging in care

​

​

​

​

No notes in system

​

​

18% (61) experiencing homeless less than one year*

*in San Francisco

SIP residents tended to fall into one of two groups:

Comparison to larger population of PEH 10/2020 in San Francisco (per CCMS data)

Medical Complexity (n = 305)

12% (41) no known PMH

​

Top 5:

1. Drug use d/o (59%)
2. Depression (47%)
3. HTN (37%)
4. Psychosis (30%)
5. Alcohol use d/o (27%)

​

Mean number of medications = 5 (SD 4.7)

​

Results: Types of System Utilization

• PCP in chart (n = 342): 58%

​

Care Utilization: 7% (23) with no care utilization

Of 323 with care utilization:

• 78% urgent care, (includes 50 Ivy Tom Waddell and Family Health Center)
• 56% jail,
• 50% behavioral/mental health services,
• 25% PES,
• 18% mental health crisis day program,
• 16% Respite,
• 11% Sobering center

Difference in ED visits (6 months prior vs 6 months after SIP): p-value = <0.0001

Difference in hospitalizations (6 months prior vs 6 months after SIP): p-value = 0.0056

Connection to Outpatient Care

*N/A = not in SIP long enough at time of data entry or exited from SIP by date

Takeaway: best opportunity for high engagement in first few months

Recap of Quantitative Data 8/2/21

• SIP hotel residents included two distinct groups:
• 1) older, sicker persons with decades of homelessness
• 2) younger, healthier* persons newer to homelessness
• SIP hotel stay was associated with an interval decrease in ED use and inpatient use
• SIP hotel stay was associated with increased connection to outpatient care in the first 6 months after placement
• Top 5 most common medical diagnoses among SIP hotel residents were:
1. Drug use d/o (59%)
2. Depression (47%)
3. HTN (37%)
4. Psychosis (30%)
5. Alcohol use d/o (27%)

​

Selected emerging themes from resident interviews �

​

• From surviving to living
• SIP hotels are housing AND they can sometimes remind people of their experiences being incarcerated.
• Decreased substance use (unplanned)
• Uncertain SIP hotel end dates cause distress
• Benefits of private bathrooms
• Benefits of regular connection to nursing staff

​

​

“This is not a COVID hotel”

“This is not a COVID hotel. This is a hotel where they're trying to help you get back into society and live life on life's terms, if this is what you want.”

-Mara, March 16, 2021

Leon: From Surviving to Living

“I know how to survive. I know how to do that well. I want to live. Surviving and living are two different things… Surviving is when you're sitting and you're scrapping and everything. No babes, nothing, just try to eat through this, that, and the other. Living's making yourself healthy, giving somebody help in life, not trying to take…. Living takes work. When you get better and better at it, it's like a muscle. It gets bigger and bigger, it gets stronger and stronger.”  

​

Clarence: A chance to get things straight.

It afforded me that opportunity to get all of my medical appointments…This has been good to me in that regard, it's been great experience that it's given me. I had the chance to get things straight, keep thing straight, and organize, keep appointments... It's hard to keep appointments, and be on time, and do things when you're homeless. It would be definitely a struggle. I might not have accomplished some of the things I have. (Clarence,  4/21).

​

Clarence: between homelessness, hotels and prison

“I spent years of homelessness, being on state parole, you didn't know what was your home, by definition. Those transitional hotels, that's not permeant housing. I was always homeless. You come out prison, you're on parole. You wind up in one of those transitional hotels. You might be good for one week, but suffer the money when that next week, and you're back out on the street hustling until you get some rent money again. I've experienced homelessness for all my life because I've been on parole for quite a long time. It took me 20 years to just discharge my number from prison. I got some PTSD from my experience in prison, [this hotel room is] no bigger than a prison cell…

I used to smoke about $50 per day of crystal. I don’t know where I got the money but that’s how much I smoked. Now it varies, sometimes what I buy lasts me 2, 3 even 4 days. Sometimes I don’t have money and I don’t smoke any. (Norma, 35, 4/29/21)

I used to wake up with the shakes every morning. Now…I can go a week and a half or so without having a drink… I just finished my last beer this morning that I had saved up for two days …Basically, I'm not out there drinking as much as I used to drink. I used to drink from 7:30 in the morning to 2:00 AM at night...It's a real big change (Jacqueline, 4/21/21).

SIP Hotel led to cutting down…

perhaps unintentionally (Harm Reduction)

Edwin: “But you know, I use drugs, right. Sometimes Crack-cocaine. Sometimes fentanyl. I'm trying to get off of it real bad, because I've been doing pretty good since I've been here…When I was on the street, there was no telling, and it was so bad. I think wellness checks is good because you have got to make sure you're alive.”

Lester: The healing power of your own bathroom

When I got here, first day here, I took a shower for four hours. Yes, I sat in there for four hours. I had to cleanse myself. Then came out, knock on the door and it was food. "Oh, what is this?" and, "Oh, yes, we feed you." "Oh, okay." I ate, I must have slept for two days.

William: The anxiety of uncertain end dates

The thing is the pressure that they put on a lot of people right now. They closing this. People wondering, "What I'm going to do if I'm going to be homeless?" Where I'm going to go?

​

Some of the nicest staff that I've ever seen. It just hurt me to see that they ain't going to have no job, we ain't going to have no house. I'm worried that I don't end up on the street. Especially sick like I am, I need all this medicine to be put somewhere. I don't need to be in no shelter…The shelter, you burden them with that. How are you going to keep stuff in the refrigerator? Rather than that, [blood sugar] going to shoot sky-high. I'm going to be dead. They ain't got to worry about me because I'm going to be gone.  

Recap of Select Themes from Interviews with SIP Residents

​

• From surviving to living, a chance to pause and plan for the future, attend appointments, etc…
• SIP hotels are housing AND they can sometimes feel connected to people’s experiences with being incarcerated. COVID protocols worsened this. Architectural design and visiting protocols contribute. Staff training and turnover rates can make this better or worse. This is complex and layered.
• Decreased substance use (unplanned)
• Private bathrooms are healing
• Labor of looking for housing, anxiety of housing anxiety, instability even in SIP due to uncertain end dates
• Regular connection to consistent nursing staff helps residents feel safe and cared for

​

​

​

Efforts paid off!

How We Plan to Use This Feedback

• Qualitative Data Validation
• Do these themes resonate or not with folks who have lived experience or who have worked closely with those with lived experience?
• Refine to present to policy/decision-makers
• e.g. in setting of Delta-variant, our evidence would support continuing to backfill SIPs, not shut them down.
• In support of expansion of different forms of housing, especially individual rooms with private bathrooms
• To advocate for changes to make SIP hotels feel less like jails and more like homes

​

​

Thank You!

​

To our team:

​

Barry Zevin, MD

Josh Bamberger, MD

Trevor Lee, MD

Stephanie Chang, MD

Sara Zhu, MDc

Hamzah Yusuf, MDc

Shannon Satterwhite, MD, PhD

Consultants and Partners

Faithful Fools

Code Tenderloin

Skywatchers

UCSF Center for Community Engagement

Kelly Knight, PhD

​

​

This study was partially funded by the Social Science Research Council

References

• Culhane, D., Treglia, D., Steif, K., Kuhn, R., & Byrne, T. (2020). Estimated Emergency and Observational/Quarantine Capacity Need for the US Homeless Population Related to COVID-19 Exposure by County; Projected Hospitalizations, Intensive Care Units and Mortality. UCLA: Campuswide Homelessness

​

• Raven/Kushel et al (2020): randomizied control study of providing perm supportive housing in Santa Clara Valley reduces psych ED visits and increased outpatient care (less impact on medical ED/hospitalizations) amg high healthcare utilizers; 86% remained in housing

​

• Kessell, E. R., Bhatia, R., Bamberger, J. D., & Kushel, M. B. (2006). Public health care utilization in a cohort of homeless adult applicants to a supportive housing program. Journal of Urban Health, 83(5), 860-873.

​