Abstract
Introduction:
People experiencing unsheltered homelessness (PEUH) have higher disease burden yet limited access to healthcare. COVID-19 introduced even greater risk for PEUH aged 65+ years with an underlying chronic health condition and were temporarily housed in hotels/motels for Project RoomKey (PRK). This study aimed to characterize a PRK cohort who received primary care from a street medicine program.
Methods:
This observational case series study included a sample of 35 PRK participants receiving primary care from a street medicine team at a single site from July to September 2020. We used the HOUSED BEDS assessment tool for taking history on PEUH.
Results:
Participants were 63% male, 40% Hispanic/Latino/a, 40% white, 94% English-speaking, and 73% had chronic health conditions. Assessment revealed: average Homelessness (H) of 4 years; 76% had no prior social service Outreach (O); average Utilization (U) was 4 emergency department visits in prior 6-months; 68% received Salary (S) from government income; Food access or Eat (E) was commonly purchased (29%) or donated (26%); clean water to Drink (D) for 59% of participants; 86% had access to a Bathroom (B); Encampment (E) was varied and 38% reported safety concerns; Daily routine (D) showed 76% could access a telephone, 32% received social support from family; 79% reported past or current Substance use (S). No participants contracted COVID-19 during study period.
Conclusions:
This study describes health and demographic characteristics of PRK participants in Southern California. Findings inform policies to continue PRK that includes onsite healthcare such as via street medicine.
Introduction
People experiencing homelessness (PEH) have higher rates of chronic disease, mental illness, and substance use as compared to the general population. 1 Unhoused populations are more likely to have and develop severe health problems, highlighting the unique and complex health and social needs of this cohort.1,2 Notably, PEH most commonly access the healthcare system for acute care, rather than primary care. 3 People experiencing unsheltered homelessness, in particular, face higher mortality risk than sheltered homeless and housed populations, 4 commonly driven by accelerated aging, 3 drug and alcohol related deaths, 5 and suicide (further lowering age of death). 6 Diseases such as cancer, cardiovascular/coronary heart disease, and respiratory disease are also common causes of premature death, potentially preventable with the timely delivery of healthcare services.6,7
PEUH have limited access to healthcare, especially primary care.8 -10 PEUH who are Veterans have historically received even less access to healthcare 11 unless enrolled in the medical home model Homeless Patient Aligned Care Team (HPACT). 12 According to 1 study, PEUH who identify as ethnic minorities (Hispanic/Latino/a/x, Native American/American Indian, Mixed/Other Ethnicity), with physical health needs and health insurance, are more likely to use hospitals for care; the study findings underscore the need for preventive healthcare services for subgroups of PEUH. 13 Another recent study cites differences between unsheltered and sheltered homeless subgroups; the unsheltered cohort, with many chronic health conditions, used more health services overall, including emergency department (ED) and outpatient care. 14 Older adults age 50 years and older who are unhoused are shown to have higher rates of ED use than any other population. 15 A significant number of these visits related to chronic disease or pain-related issues that could have been managed in non-acute settings, further demonstrating the impact of limited access to primary care for older PEUH. 15
Street medicine, defined as health and social services developed specifically to address the unique needs of PEUH, delivered directly in their own environment, 16 is a promising model of primary care for this underserved population. Street medicine differs from traditional office-based medicine or mobile (van or recreational vehicle) medicine in that providers travel on foot to meet with and treat patients in their encampment. This practice has 2 purposes: First, it intentionally flips the patient-provider power-dynamic that occurs when patients must enter a medical space. Second, it reduces barriers to accessing care such as physical distance, travel, having to leave possessions to attend a medical visit, and competing priorities, among others. 16 Recent research has highlighted the impressive breadth and scope of services provided by street medicine teams, 16 a model of care that is now proliferating across the U.S. and worldwide. 17
At the beginning of the COVID-19 pandemic, older PEUH with chronic health conditions were believed to be at heightened risk of hospitalization, critical care, and death if they were to become infected with COVID-19. In response, these individuals were offered temporary housing in hotels and motels as part of the nation-wide initiative, Project RoomKey (PRK).18 -20 On the streets, PEUH were unable to abide by COVID-19 public health recommendations, such as safer-at-home orders, self-isolation, and hand hygiene practices. 19 Given higher risk for adverse consequences for these more vulnerable unhoused groups, PEUH age 65 years and older, experiencing underlying chronic health conditions, and/or who were medically compromised were temporarily sheltered in PRK to prevent the spread of COVID-19 and potential for severe illness and death.18,20
Eligible individuals who were unhoused and selected for the PRK program received temporary shelter, during which time healthcare needs arose or were identified by PRK staff and medical care was requested onsite. As such, 1 PRK site partnered with a street medicine program to provide onsite medical care for participants. The objective of this study was to characterize the cohort of PRK participants who received primary care services from street medicine at a single PRK site using the HOUSED BEDS assessment tool. 21
Methods
We conducted an observational, case series study among a convenience sample of individuals participating in the federal PRK program. Participants elected to receive primary care from a university-affiliated street medicine program at a single PRK site in Los Angeles County from July to September 2020. This study was reviewed and approved by the University of Southern California Institutional Review Board as exempt research. Participants provided signed consent to engage in medical care with USC Street Medicine and have deidentified data (reported here) analyzed for research and quality improvement. This report adheres to the strengthening and reporting of observational studies in epidemiology (STROBE) checklist for cohort studies.
Study Sample
Individuals were eligible for PRK if they were (1) unhoused and age 65 years and older, or if they (2) had underlying chronic medical conditions and/or were medically compromised. 18 Study participants who met criteria were temporarily housed at a single PRK site in Los Angeles County. They were either referred to street medicine by a PRK staff member or had personally requested to meet with street medicine. Individuals referred to street medicine were notified of the referral and were given the choice of engaging in care or not, without penalty or impact on their participation in PRK. Those electing to engage in care completed consent documentation with a member of the street medicine team. PRK participants who were not referred to the street medicine team were not included in this study.
Procedures and Measurement
As part of usual care intake activities, 2 street medicine team members engaged with participants and/or conducted physical examinations multiple times (range: 0-3) over the study period. Team members used the HOUSED BEDS assessment tool 21 : a clinical tool for taking history on unsheltered homeless patients. The acronym stands for: Homeless history (H), Outreach (O), Utilization (U), Salary (S), Eat (E), Drink (D), Bathroom (B), Encampment (E), Daily routine (D), and Substance use (S). 21 The following details the measures in each letter of the acronym:
Participant demographic and clinical data collected during intake was supplemented by electronic medical record documentation (both the university’s health system and LA General Medical Center) to verify unclear data or obtain missing data.
Statistical Analysis
Univariate statistics were used to descriptively analyze all demographic and HOUSED BEDS data points (eg, frequency, percent, mean, median, standard deviation, and range). Data were analyzed using SPSS statistical software version 28. 22
Results
Participant Demographics
Street medicine team members conducted intake assessments with 35 PRK participants and 34 provided responses to the HOUSED BEDS questions (97% response rate). On average, participants were 49 years old (range = 19-69 years) and most identified as male (63%), Hispanic/Latino/a (40%) or white (40%), and English-speaking (94%). A handful of participants were veterans (11%) (Table 1).
Summary of Demographic Characteristics of PRK Patients.
All participants reported at least 1 chronic medical condition (
Summary of Clinical Characteristics of PRK Patients.
Housed Beds
Participants were asked to respond to questions about their lives prior to entering PRK in the summer of 2020. The main results from the HOUSED BEDS assessment tool are reported here:
Homeless history (H): Participants reported having experienced, on average, 4 years of homelessness and the majority (79%) reported experiencing chronic homelessness (defined as 12 or more months homeless). 23 Housing status was mixed; 45% reported living on the street and 34% described a combination of living on the street, in a car, and/or in a shelter; 7% reported living in their car, shelter, or “other” site, respectively.
Outreach (O): Most (76%) participants had not previously engaged with outreach for physical health, mental health, or housing services.
Utilization (U): On average, participants reported having had 4 ED visits in the past 6 months. Six individuals (18%) did not report visiting the ED in the past 6 months.
Salary (S): Over half of the participants reported receiving government income (68%) which was most commonly Supplemental Security Income (SSI; 29%); other forms of government income including EBT (food stamps) (26%). Personal income from recycling was acquired by 14 (41%).
Eat (E): While living unsheltered, 10 (29%) participants reported having access to food by buying it and 9 (26%) received food given to them by an organization. Before arriving to PRK, 74% of participants were receiving at least 7 meals per week.
Drink (D): For 24% of participants, “other” water sources, such as water fountains and restroom faucets were their primary source. Bottled water was the second most common source (21%) and a combination of bottled water, donated water, and/or other type of water source was reported by 6 (18%).
Bathroom (B): Nineteen (86%) participants reported having some type of access to a bathroom although not necessarily 24-h access. Sources included bathrooms inside of businesses or located at an organization, park restroom, or a combination of the above sources.
Encampment (E): Participants commonly reported sleeping outside on the street (50%), in a tent (11%), a combination of encampment locations (21%); including inside a tent, on the street, or inside a vehicle), or “other” situation (18%) such as a friend’s house or trailer.
Daily routine (D): Most participants (76%) had access to a telephone and 11 (32%) received social support from family members; 8 (24%) reported social support from a combination of their partner, family members, and/or friends.
Substance use (S): Twenty-seven participants (79%) reported ever using substances in their lifetime, including: alcohol: 17 (50%), tobacco: 12 (35%), cannabis: 12 (35%), methamphetamine: 9 (26%), and heroin: 5 (15%) (Table 3).
Summary of PRK Participant Responses to HOUSED BEDS Assessment Tool.
Discussion
This study aimed to describe the health and demographic characteristics of a sample of PRK participants who received onsite primary care from a street medicine team. Our results corroborate those by Donesky et al. who identified a need for medical care among participants at a PRK site in Northern California. 24 Results are consistent with prior studies among general homeless populations showing high prevalence of chronic disease, mental illness, and substance use.1,3 Current study participants were older, sicker, and less diverse compared to the general Los Angeles homeless population.25,26 Study results highlight the complexity of health needs for this group 2 and underscore the importance of finding solutions to address the ongoing healthcare and social service needs of PEUH following the conclusion of the temporary PRK program. These results also raise questions about PRK program recruitment methods where PEUH had to be engaged with a social service organization and be directly referred to the program (ie, no walk-ins). Assertive models of care requiring prior participant engagement, such as with PRK, have the potential to introduce selection bias and recruiting participants who are not necessarily representative of the unsheltered population. Nevertheless, study results suggest that the participants at this PRK site were appropriate for the program as they were found statistically to be at high risk for serious adverse events and/or hospitalization if they contracted COVID-19.
High response rates by PRK participants to the HOUSED BEDS assessment tool support the acceptability and feasibility of the tool for use in street medicine practice and by healthcare providers at a PRK-type program. Application of the HOUSED BEDS assessment among a currently sheltered yet formerly unsheltered population was unique and not the intended setting when the tool was originally developed. Study results underscore the importance of using the HOUSED BEDS assessment tool to illuminate the reality and unique experience of PEUH 21 even if temporarily sheltered. Future research efforts should focus on widespread use of this assessment tool to better describe the health and circumstances of PEUH across the country and further quantify PEUH access to basic needs such as toilets, food, and clean water. A subsequent study to validate the HOUSED BEDS assessment tool among a larger sample of PEUH is warranted and underway as no valid or reliability-tested instruments currently exist for the unsheltered homeless population.
The spread of COVID-19 was not the focus of the present study; however, it was the primary aim of the PRK program. We did observe that none of the participants in this study died from COVID-19 or contracted it, suggesting that the PRK program was successful in its primary aim at this particular site. The secondary aim of PRK was to transition participants into permanent housing which was reported in November 2022 to have been accomplished for over 4,800 out of over 10,000 individuals who entered the program. 27 Given the poor health status of participants in our sample, permanent supportive housing appears to be a critically important benefit to these individuals which likely holds implications for their health and wellbeing. Returning PRK participants to the street at the conclusion of the program could place them at risk, once again, for requiring acute medical services and experiencing adverse health outcomes. Other research efforts exploring the health outcomes of PRK participants who left the program for the street or permanent supportive housing is warranted to better inform this program and others with similar aims.
Future housing initiatives, including current programs such as PRK and Project Homekey, 28 can predict that healthcare services for PEUH will be needed. Embedding a street medicine team, well-versed in the reality of street living, at the site of temporary housing allowed for more comprehensive assessment in a setting the participants felt comfortable (eg, personal living space) as distinguished from services delivered in a clinical setting where PEUH traditionally have felt marginalized or unwelcome. 29 Another study described implementation and services provided by a learner and licensed clinician/preceptor model of care provided at a PRK site in Northern California. 24 That approach differed considerably from the present study model by namely, providing medical care in a designated room at the PRK site (not in participants’ own/preferred space), and staff (volunteers and learners) facilitated services, rather than a full-time employed, licensed, interdisciplinary street medicine team. Future research can explore the characteristics of people experiencing sheltered versus unsheltered homelessness, their motivations and preferences for receiving supports, and the characteristics and biases of helping organizations.
Limitations
This study is limited by the single site, sample size, reliance on self-reported data, and chart review limited to unclear or missing data. However, the response rate was at least 60% for each item on the HOUSED BEDS assessment tool. Potential bias from self-reported data may include issues with recall from participants. The small sample was also a convenience sample leading to possible bias in how participants were selected. There is the possibility of interviewer bias due to how the interviews using the HOUSED BEDS assessment tool were conducted, including data collection over multiple visits and by more than 1 interviewer. Thus, study results may not be generalizable to other areas.
Conclusions
This study sheds light on the health and demographic characteristics of PRK participants. Findings suggest the importance of healthcare and social supports for individuals transitionally housed at a PRK site. Additionally, findings may inform practice and policy for PEUH by highlighting the role of tailored, person-centered models of care, such as street medicine, and need for continuation of the PRK program with inclusion of onsite primary care services.
Footnotes
Acknowledgements
We would like to thank the research participants for giving their time and for agreeing to participate in this study. Additionally, we would like to thank the street medicine team members and student trainees for their efforts and work that contributed to this study.
Authors’ Note
Prior Presentations: North American Primary Care Research Group (NAPCRG) 49th Annual Meeting, November 19 to 23, 2021.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a grant from the United States Health Resources and Services Administration, grant number D57HP32746. No financial disclosures were reported by the authors of this paper.
IRB
UP-20-00853.
