A Mixed Methods Evaluation of Interventions to Meet the Requirements of California Senate Bill 1152 in the Emergency Departments of a Public Hospital System

Abstract

Policy Points.

• Clarifications to Senate Bill (SB) 1152 are necessary to address the differences between inpatient and emergency department (ED) discharge processes, determine how frequently an ED must deliver the SB 1152 bundle of services to a single patient, and establish expectations for compliance during off‐hours when social services are unavailable.

• Because homelessness cannot be resolved in a single ED visit, the state should provide funding to support housing‐focused case workers that will follow patients experiencing homelessness (PEH) through the transition from temporary shelters to permanent supportive housing. Medi‐Cal could fund the delivery of the SB 1152 bundle of services to defray the costs to public hospitals that provide care for high numbers of PEH.

• California legislators should consider complementary legislation to increase funding for shelters so that sufficient capacity is available to accept PEH from EDs and hospitals, and to fund alternative strategies to prevent poverty and the upstream root causes of homelessness itself.

Context

Prompted by stories of “patient dumping,” California enacted Senate Bill (SB) 1152, which mandates that hospitals offer patients experiencing homelessness (PEH) a set of resources at discharge to ensure safety and prevent dumping.

Methods

To evaluate interventions to meet the requirements of SB 1152 across three emergency departments (EDs) of a Los Angeles County public hospital system with a combined annual census of 260,000 visits, we used an explanatory sequential mixed methods approach, focusing first on quantitative evaluation and then using information from qualitative interviews to explain the quantitative findings.

Findings

In total, 2.9% (1,515/52,607) of encounters involved PEH. Documentation of compliance with the eight required components of SB 1152 was low, ranging from 9.0% to 33.9%. Twenty‐five provider interviews confirmed support for providing assistance to PEH in the ED, but the participants described barriers to compliance, including challenges in implementing universal screening for homelessness, incongruity of the requirements with the ED setting, the complexity of the patients, and the limitations of SB 1152 as a health policy.

Conclusions

Despite operationalizing universal screening for homelessness, we found poor compliance with SB 1152 and identified multiple barriers to implementation.

Homelessness is an enormous issue in the united states, with an estimated 161,548 individuals experiencing homelessness in California on any given day. ¹ Media reports of “patient dumping,” which described hospitals transporting patients experiencing homelessness (PEH) to homeless encampments in hospital gowns in large California cities such as Los Angeles ² and San Francisco, ³ prompted state lawmakers to enact Senate Bill (SB) 1152, which requires a standardized process for PEH discharge planning (the process of transitioning a patient between levels of care) ⁴ in all California hospitals. The Service Employees International Union, the California Police Chiefs Association, and the California Pan‐Ethnic Health Network supported the legislation, arguing that it would encourage hospitals to address the unmet social needs of these patients and might encourage hospitals and social services to partner more effectively. However, the California Hospital Association and California chapter of the American College of Emergency Physicians opposed the bill for being too prescriptive and putting too much pressure on hospitals to meet posthospitalization needs. ⁵ Opponents argued that SB 1152 could also compound overcrowding in emergency departments (EDs). ⁶ On September 30, 2018, Governor Jerry Brown signed SB 1152 signed into law. No funding was provided to the hospitals to meet the requirements. ⁷

In January 2019, the California Department of Public Health issued All Facilities Letter (AFL) 19‐01 to notify hospitals of the requirement to create a discharge planning policy and process specific to PEH. ⁸ The AFL states that hospitals must identify PEH, “prioritize placing the homeless patient at a sheltered location with supportive services,” and identify a postdischarge destination, either with a social service agency or at a dwelling place identified by the PEH. Additionally, hospitals are required to offer the following to PEH ⁹ :

•  Transportation to a location of the patient's choice

•  A meal

•  Weather‐appropriate clothing

•  An adequate supply of medications

•  Appropriate infectious disease screening and vaccinations

•  Referrals to follow‐up medical care

•  Referrals for follow‐up behavioral health care, if necessary

•  Housing resources

It has been suggested that discharge planning may help address homelessness ¹⁰ ; however, given the brief nature of most ED visits, discharge planning in EDs may be minimal, especially in comparison to the inpatient hospital setting. ¹¹ While SB 1152 explicitly mandates discharge planning processes for hospital inpatients, it does not specifically mention EDs. In November 2018, a California Hospital Association guideline stated that SB 1152 requirements ought to be applied to EDs as well. ¹² Before our study, the implications of that guideline for EDs had not been evaluated. We aimed to evaluate the implementation efforts and interventions to comply with SB 1152 across the EDs of a public hospital system.

Methods

We used an explanatory sequential mixed methods approach to evaluate the health system's implementation of interventions to achieve compliance with SB 1152 in the ED. We defined compliance as documentation of having addressed all required elements of SB 1152. The explanatory sequential design was chosen for its ability to help us understand the success or failure of implementation efforts as well as unintended consequences of implementation. ¹³ , ¹⁴

The evaluation took place in three stages. First, we queried the electronic health record (EHR) to identify rates of screening for homelessness, ED patients identified as homeless, and documented compliance with each component of SB 1152. Second, we performed a manual chart review of PEH presenting to a single ED to better characterize the patients identified as experiencing homelessness. We then interviewed frontline ED providers to explicate the barriers and facilitators to achieving compliance. Thus, the qualitative interview data built on the quantitative data, following the explanatory sequential design. ¹⁵

Study Context

The Los Angeles County Department of Health Services (LAC DHS) is the second largest municipal health system in the United States and includes four hospitals and 26 health centers. It cares for about 750,000 patients annually, employs 23,000 staff, and has an annual operating budget of $6.2 billion. ¹⁶ There are three EDs within the LAC DHS system; they are located at Olive View–UCLA Medical Center (OVMC), Harbor‐UCLA Medical Center (Harbor), and LAC+USC Medical Center (LAC+USC), and their combined annual census is approximately 260,000 visits. The patient population has high rates of self‐described unmet social needs and interest in obtaining assistance. ¹⁷

The LAC DHS Emergency Department Expected Practice Committee determined the workflow for SB 1152 implementation, including roles of each provider group in the process of ensuring identification of homelessness, SB 1152 compliance, and documentation. First, the SB 1152 requirements were examined to determine which components were already addressed for all patients and which would require new interventions. At baseline, all ED patients receive a medical screening exam, are offered medical follow‐up, are screened for insurance eligibility, and are counseled on enrollment. SB 1152 components that would require new interventions, either because they were not done consistently or because they were not captured in EHR documentation, included offering a meal; providing weather‐appropriate clothing; offering prescriptions for chronic medications; providing infectious disease screening/vaccination; offering behavioral health referrals; sharing housing resources; offering transportation to the destination of the patient's choosing; and documenting the chosen destination.

Interventions to address the new elements included the following: Registration workers were tasked with screening for homelessness by asking “Are you homeless?” during the registration process. A positive screen would trigger the icon “HL” on the ED tracking board to alert providers. The ED provider would consult social work and address the SB 1152 requirements relevant to medical care (vaccinations and need for behavioral health referrals). Anyone in a provider role (nursing, medical provider, or social work) could offer clothing and a meal. Social work would offer housing resources (which could include a list of shelters, referral to permanent supportive housing programs, or both), address transportation needs, and document the discharge destination. (See Figure 1 for workflow and roles.) An ED Homeless Discharge Form consisting of checkboxes for each new element was created to standardize documentation. All providers involved (physicians, nurse practitioners, social workers, and registered nurses) had access to the form. A homeless discharge report was created to capture data from the ED Homeless Discharge Form to demonstrate compliance for regulatory agencies.

Figure 1.

Swimlane Diagram of Emergency Department Workflow to Comply With California Senate Bill 1152 Requirements [Colour figure can be viewed at wileyonlinelibrary.com]

   

Abbreviations: MD, medical doctor; NP, nurse practitioner; RN, registered nurse; SW, social work.

Quantitative Methods

EHR Review

We performed a retrospective EHR query of patients presenting to the three LAC DHS EDs between December 1, 2019, and January 31, 2020. This was six months after the regulatory requirement went into effect but prior to the start of the COVID‐19 pandemic.

Subjects. ED patients were included in our data set if they were identified as PEH via the homeless discharge report.

Measures and Outcomes. We reviewed rates of screening (defined as the number of encounters with an answer to the homelessness question divided by the total number of encounters) and documented rates of PEH and compliance with the eight SB 1152 components that were not previously part of the EHR (infectious disease screening/vaccination, supply of medications, referral to behavioral health care if necessary, housing resources, meals, weather‐appropriate clothing, transportation, and discharge destination). Compliance for each component was defined as any response on the homeless discharge form (offered, patient accepted; offered, patient declined; or not indicated) that indicated that the patient had been offered the required service or good (meal, clothing, housing resource, and so on).

Analysis. We used descriptive statistics for the analysis.

Manual Chart Review Methods

We conducted manual chart review of PEH at a single ED (OVMC) for the same period as the EHR review. The objectives were (a) to determine the rate of SB 1152 compliance when considering all documentation compared with the documentation in the ED Homeless Discharge Form alone, and (b) to detail the complexity of cases and frequency of ED visits for PEH.

Data Abstraction. We used a prespecified data abstraction method and a trained abstractor. We reviewed all notes for the index ED visit, used broad definitions of the eight components, and assumed providers were compliant if any of the components were mentioned.

Measures Abstracted. Demographic characteristics of the PEH including age, gender, race, ethnicity, preferred language, health insurance status, and comorbid conditions were recorded.

Qualitative Methods

Theoretical Framework

We chose the consolidated framework for implementation research (CFIR) to evaluate the implementation of the SB 1152 requirements because it is a determinant framework that helps describe what works and why across multiple contexts. ¹⁸ While an implementation science approach was not used to inform the initial operational changes made in response to SB 1152, the use of CFIR as an evaluation tool aids in the recognition and classifications of themes across settings.

Sampling Strategy and Participants

Qualitative interviews were conducted as part of a larger project focused on the integration of social care and took place between June 2020 and March 2021. We used purposeful sampling to identify key informants through the LAC DHS Emergency Department Expected Practice Committee and a successive snowball sampling scheme. ¹⁹ We ensured that providers of various patient‐facing roles (physicians, nurse practitioners, registered nurses, social workers, and registration workers) across each of the three EDs were included. Interviews ended when thematic saturation was reached.

Data Collection Procedures

Interview participants participated in a one‐time video or telephone‐based semistructured interview, with interviewers using a guide developed using the CFIR. ¹⁸ Interviews were recorded, transcribed, and checked for accuracy by a trained research assistant.

Analysis

We used ATLAS.ti 8 software for Mac and employed a combination of inductive and deductive approaches to consider how the data related to the CFIR, and to ensure that no emergent codes were lost because of lack of fit with the prespecified framework. ²⁰ , ²¹ We generated an initial codebook based on the CFIR, open‐coded the initial four transcripts as a group, and added to the codebook by consensus. One member of the qualitative team, BRT, coded the remainder of the transcripts and discrepancies between their coding and prior coding were discussed at weekly team meetings to achieve consensus. We used analytic memos to record our reflections and emerging ideas. We then analyzed the coded text and all memos to generate and categorize themes. We focused on salience of the material rather than frequency. ²²

Techniques to Enhance Trustworthiness

In order to ensure reliability of the data and face validity of the emergent themes, we used a system of member checking in which we presented preliminary findings and collected feedback from a group of LAC DHS frontline staff and administrators. ²³ In addition, we used an audit trail through memo creation and analyzed deviant cases as a team to achieve consensus on their meaning. ²⁰ , ²⁴

Researcher Characteristics and Reflexivity

The qualitative team consisted of two emergency physicians/implementation science researchers employed by the health system (BRT and KY), a social epidemiologist/health policy researcher (RTS), three research assistants (AA, LS, and GT), and a medical student (LGP). As a team, we reflected on how our backgrounds and perspectives influenced the interpretation.

Reporting Standards

This manuscript adheres to the good reporting of a mixed methods study standard for mixed methods research (see the checklist in the supplementary material). ²⁵

Human Subjects Protections

Research was approved by the Olive View–UCLA Education Research Institute Institutional Review Board prior to the commencement of research activity.

Data Integration

After qualitative interviews were analyzed, the team revisited the quantitative data and used consensus discussions to map qualitative themes to the quantitative data. ²⁶ , ²⁷ Integrated presentation of results was accomplished by displaying exemplary quotes next to the quantitative findings by institution, SB 1152 component, and provider role.

Results

For the two‐month study period, 2.9% (1,515/52,607) of encounters across the three EDs had a designation of homelessness (see Table 1 for hospital‐specific percentages), and rates of documented compliance with all components of SB 1152 were low (see Table 2).

Table 1.

Unique Attributes of Each ED Create Differing Barriers and Facilitators by Location

Institution	Documented % of ED Patients Experiencing Homelessness	Representative Quotations About Barriers and Facilitators (Interview Participant Type)
OVMC	2.6% (307/11,866)	Barriers: “Because typically, we know, there are limited to no resources in the San Fernando Valley for shelters. And if there are, the screening process is a little difficult to navigate around with the type of clientele we have. And so if it says, ‘Well, you know, you have to come in two days,’ well, first, you have to come in for an appointment. No, we need some place now. Can't come in tomorrow for an appointment or in two days for an appointment.” (SW)
		Facilitators: “We actually implemented social workers that are readily present in the emergency department. About four years ago when I arrived at Olive View, we really did not have any social worker presence. There was social work present Monday to Friday, 7a‐3p; that has expanded to 24/7 social coverage in the department. So that has changed significantly from what was started four years ago to now.” (MD)
Harbor	2.4% (401/16,680)	Barriers: “Even with the law, some of the providers, they're just not getting it, not all of them, but some of them, you know, okay, there's no other medical needs. So, let's discharge them. And discharge people who are really still fragile, who need something like housing for health, or, you know, shelters are not giving you shelter beds, so let social work see them and see what we can do. In some cases, we can't do anything.” (SW)
		Facilitators: “Our social work team has a partnership with PATH and LAHSA. They have a special team called the homeless task force that focuses on housing and they sort of look at all the different options they're aware of and for people based on what they understand to be available and what the criteria for different organizations have.” (MD)
LAC+USC	3.4% (807/24,061)	Barriers: “A huge, a very big population that comes in is, you know, they are homeless, or a lot of them are not being able to be seen whether it could be for other reasons, a lot of our patients actually they leave without being seen, let's say they are about to get discharged, or they do get discharged, and they're waiting on social work. But for one reason or another, social workers are unable to see them or they leave without being seen. But we, unfortunately, are not able to touch every homeless individual that comes into the ER.” (SW)
		Facilitators: “You know, the county, I thought we did a pretty good job before (SB 1152); you know, if we identified somebody who's homeless, we would help them with their meds, and food and things like that. That has kind of been our mission at County. But a private hospital, you know, I'm pretty sure that their goals are different in order to get the patient out as quickly as possible.” (NP)

Abbreviations: ED, emergency department; ER, emergency room; Harbor, Harbor‐UCLA Medical Center; LAC+USC, LAC+USC Medical Center; LAHSA, Los Angeles Homeless Services Authority; MD, medical doctor; NP, nurse practitioner; OVMC, Olive View–UCLA Medical Center; PATH, People Assisting the Homeless; RN, registered nurse; SB, Senate Bill; SW, social work.

Table 2.

The Eight Requirements of SB 1152 Differ in Their Level of Difficulty to Accomplish in the ED Setting

SB 1152 Component	Documentation Completed for PEH	Representative Quotations (Interview Participant Type)
Infectious disease screening/vaccination	9.3% (141/1,515)	“So, we don't do the routine, just wellness vaccines. So pneumococcal vaccine, shingles vaccines, those are vaccines that are preventive measures for elderly populations. We don't normally do that.” (RN)
Prescriptions	9.0% (136/1,515)	“They're coming for seizures, and they're taking, suppose; they're supposed to take seizure medication. And we would ask, are you still taking? Are you taking your seizure medication? And they say, well, I ran out of it, we would prescribe that. But if they have a bunch of other conditions, we usually don't ask, do you have the medications for everything else? Let me refill your medications for everything else. It's more visit specific condition that we're interested and we'll address and we provide results for.” (RN)
Behavioral health needs	9.2% (139/1,515)	“Yeah, so we have a mental health urgent care just down the street from us. Their hours seem to change frequently. So, it becomes difficult and again, I don't even know how they're impacted by COVID, I don't think that's ever been looked at. But, we say we have flyers for their department and their services. And we say, ‘If you have mental health needs, here's their telephone number; they are, you know, a quarter mile down the street, you can walk down there.’” (RN)
Social work consult for housing resources	33.9% (513/1,515)	“And, you know, we give them what we've got, you know, the majority of our, our homeless people are pretty, you know—at least the ones that we kind of know and see on a regular basis, again, we can say, ‘Oh, this is so‐and‐so,’ we already know that he's got clothing, he doesn't want shelter.” (RN)
Meal offered	21.1% (319/1,515)	“We, we run out of food all the time, our fridge is filled at six at night. And then it's not filled until eight in the morning. You still got a lobby of 50 people, not all 50 people are going to eat but you're going to have a lot of people in there that are hungry, that need to be, you know that, you kind of you can kind of tell the ER has become a food kitchen for a lot of these patients. So there have been plenty of times where I can barely like find like a cracker and a milk for a patient, because all our resources have left.” (RN)
Weather‐appropriate clothing offered	19.9% (302/1,515)	“One of the things that we had a problem here with was with clothing. The hospital doesn't really provide the clothing. And it's run by the volunteer office. And it's not really that well organized, it's just kind of a big pile of used clothes. And it's kind of left up to the nursing staff to actually go down there and try to dig up something that fits this patient. So that was kind of a problem.” (NP)
Transportation offered	24.8% (375/1,515)	“We had issues before, just like having the ethical dilemma on a patient, feeling like a little bit unsafe to go on the bus, but or like being late at night, in that having the bus and then my understanding was that we cannot arrange an ambulance or any kind of other means of transportation, taxi or anything like that, because it's considered dumping. So that's the case. Also, taxi vouchers are extremely limited. I think my understanding was that they have a certain number that they get in the beginning of the month. And if they run out of them towards the beginning, then they are not going to get it. Most of the time it's bus, some kind of bus pass.” (RN)
Discharge destination documented	27.1% (410/1,515)	“You must call and get a shelter bed. But you can't get no shelter bed, nobody's just gonna give you a bed. Every once in a while, we are able to get assisted living where we can say that patients are going the x, y and z, but if you call shelters they don't really hold beds. So why would you put that in the law? That you need to know the patient's destination? That's not…you know, if it had come with some resources, that would have been so much better, but it didn't, so we have to just make our way.” (SW)

Abbreviations: ER, emergency room; NP, nurse practitioner; PEH, patients experiencing homelessness; RN, registered nurse; SB, Senate Bill; SW, social work.

To explain these findings, we interviewed 25 frontline providers across the three hospitals: nine physicians, seven nursing representatives (four registered nurses, one nurse practitioner, and two nursing administrators), three clinical social workers, and six registration workers (see Table 3 for selected quotations by role). In the following sections, four themes are used to group the integrated results pairing the quantitative findings with explanatory observations and quotations from the providers: challenges in implementing universal screening for homelessness, incongruity of the intervention with the ED setting, the complexity of the patients who were the intervention recipients, and the limitations of SB 1152 as a health policy.

Table 3.

Comparison of Perspectives on California SB 1152 Requirements by ED Provider Role

Provider Role	Representative Quotations (Interview Participant Type)
Multiple roles	“And then the SB 1152 was a huge undertaking, a couple years ago that involved social work, you know, people who were in charge of the EHR, because there's a lot of documentation pieces that need to happen, providers, nurses, you know, hospital administration, there's a lot of things that went into a pharmacy, so it was a really big multidisciplinary effort.” (MD)
Registration	“The problem is, is that sometimes when people come into the hospital, the financial person who goes interview them, they might not be able to get that information, because, for example, the patient might come in a completely psychotic, they're too agitated and too aggressive, to get information from them. Right, so that's the difficulties with them.” (Registration) “Patient is homeless or concerned about losing their house, and so, yeah, a lot of times, the patient may tell us ‘No.’ They don't want us in their business. So, they feel more comfortable talking to social work or the provider.” (Registration)
Social work	“The issue I had with SB 1152 was more of the long term, more long‐term resources, again, like if I helped somebody apply for, for, for permanent housing, how do they, how do they get connected without a phone? And that was a lot of the drawback that I found with SB 1152 was, yes, we found them shelter for the night, we found them food and clothing. But how do we, how do we prevent readmissions? How do we? How do we increase a person's quality of life by giving them something more long term and something more stable versus something that was just overnight and I, over the time I've worked here, I just found that not having a phone or not having ID was a real drawback that SB 1152 didn't really touch upon—but how do we ensure that they have a permanent housing application?” (SW)
Nursing	“So, for me, if you're a good nurse, you would have always done all this stuff anyway. But again, not everybody's a good nurse, not everybody goes to the lengths that they should for, for certain cases. I mean, of course, even if I wasn't mandated to do this, if I saw that somebody had, you know, inappropriate clothing, or they were hungry, or they, you know, I would, I would have done all those things anyway, you know, I would have done all of them.” (RN) “However, it takes up a lot of time from nursing standpoint, when you also have when you're in a trauma bay, and you also have like two other patients you're handling and they could be dying from a stab wound or something like that; your least concern is trying to make sure that person has enough resources.” (RN)
MD	“In general, I tend to think of SB 1152 as a documentation issue in the sense than, like provision with the goal of influencing patient outcomes. So, it's been an interesting thing to watch DHS roll this out, in the sense that it seems to have produced a mandate to document rather than a focus on delivering services that are improving patient outcomes.” (MD)

Abbreviations: DHS, Department of Health Services; ED, emergency department; EHR, electronic health record; MD, medical doctor; RN, registered nurse; SB, Senate Bill; SW, social work.

Universal Screening for Homelessness

Across the three EDs, universal screening for homelessness was achieved—all encounter records had an answer to the question about homelessness. This field was mandated in the EHR, which ensured that “homelessness status” was recorded for every ED patient. Providers, however, described concerns about the accuracy of the data. Patients make a conscious choice about whether to disclose homelessness, and interview participants suspected that some PEH did not answer “yes” when the screening question “Are you homeless?” was asked during registration. One registration representative stated,

I am also asking financial questions, so we don't always get the right answer. That question—if I ask it, it is different than if the doctor asks or nurse asks. They know the nurse and doctor, they're there to help. I am there to collect money. So, they can give me wrong answers. But the way they look at the nurses and doctors is different. They're there to help. (Registration)

Interview participants also suggested that more emphasis was placed on this question during the registration process if the patient “looked homeless.” One registration worker said, “Now if the patient looks well kept, doesn't look homeless, most of us don't ask. If he looks homeless, then, yes, then we would ask.”

The lack of formally planned implementation processes, including training protocols for screening for homelessness, left providers uncertain about their ability to discuss homelessness with patients. A registered nurse stated:

So, whenever we think of changes, we have to communicate and educate the staff that's going to be doing that so they can kind of get a better understanding as to why they're doing it. So, like I said, the biggest barrier was just asking those questions, and the staff not being comfortable and feeling comfortable. (RN)

Lack of training may explain both why some providers avoided the question and why some patients who gave a “no” response to the screening question were identified as PEH later in their visits.

Interview participants believed that helping patients with housing needs, including the universal screening process, aligns with the LA DHS mission (“To advance the health of our patients and our communities by providing extraordinary care”), but they also noted the difficulties of accomplishing this goal in the setting of many competing priorities. A registered nurse explained as follows:

It's unfortunate that [SB 1152] has to be a mandate. But without being a mandate, it doesn't become a priority. Honestly, it's been a few years so I don't really remember the whole political thing other than the TV cameras following people around and publicizing kind of shaming hospitals into what you perceive on the outside to be the right thing to do. Most of the time, I perceive it to be the right thing to do. (RN)

Available resources, however, varied among the EDs, and thus the standardized processes were adapted to the local settings and resources (see Table 1). Notable differences in resources among the EDs included the presence or absence of a homeless task force within the hospital, the geographic locations (closer to/farther from shelters, community partners, and public transit), staffing of the social work team (especially outside of business hours), and the extent to which clothing was available from the hospital or auxiliary organizations. The local environment and resources of the individual facilities were noted by participants to have particular influence over implementation strategies because of the lack of additional resources provided by the bill. “It is an unfunded mandate, as they say,” stated a physician.

Congruity of SB 1152 Components With the ED Setting

Documentation rates of compliance with the eight components of SB 1152 that we tracked were low: 9.3% (141/1,515) for infectious disease screening/vaccines, 9.0% (136/1,515) for prescriptions, 27.1% (410/1,515) for documented discharge destination, 9.2% (139/1,515) for behavioral health resources, 33.9% (513/1,515) for social work consult for housing resources, 21.1% (319/1,515) for meal offered, 19.9% (302/1,515) for weather‐appropriate clothing, and 24.8% (375/1,515) for transportation offered. Barriers differed by component; however, a unifying theme was the misalignment of the ED setting with required elements of SB 1152. For example, providing clothing (other than paper scrubs) for patients is not part of the normal operations of an ED. In some cases, staff donated their own clothes to fill the need. According to one registered nurse, “It's just very difficult practice to have when you don't have the resources to just do these basic needs that you're trying to fulfill for these people.” Another example is vaccination, which is more typically done in primary care settings. The nonemergent vaccinations mandated as a component of SB 1152 to be offered to PEH fall outside of typical ED workflow.

ED recidivism was common, posing an additional challenge to SB 1152 compliance. The 216 encounters captured in the PEH charts that we manually reviewed represented 116 unique individuals. The number of visits per patient ranged from 1 to 58. Return visits were often prompted by the unresolved housing issues. A registered nurse said,

But sometimes it doesn't even say on the chart that they're homeless, but the patient will say like, “Oh, I don't have a place to go. And then can I stay in the hospital? Can I stay for a day or two?” That's why they want to be admitted, just because they don't have a place to stay. But usually we can't do that. (RN)

An extreme example involved a patient who checked into the ED 58 times during the two months of the study. Because the text of the law does not address the issue of multiple visits, providers attempt to comply with each of the SB 1152 components at every visit.

Providers also noted the challenges of connecting patients to services from the ED setting. One physician stated:

If the patient is interested in shelter or other placements, those people don't answer the phone in the middle of night, you have to wait till the next morning. So, it does lead to a number of patients who would normally be medically stable for discharge, waiting until morning hours to complete the referrals. (MD)

Several providers noted that PEH stay in the ED for prolonged periods as they wait to see a social worker or wait for the social worker to assess shelter or housing options. Such prolonged encounters are at odds with the goal of ED operations, which is to maintain an efficient patient flow through the evaluation process to ensure there are open beds for new patients presenting with life‐threatening emergencies.

According to our chart review, social work consultation for housing resources was offered in 71% of cases and, of these cases, patients declined social work consultations in 27% of the encounters. ED social workers do not follow patients over time until their social needs are resolved, and the social workers typically assist the patient during the ED visit by sharing a list of housing resources, which is unlikely to resolve housing needs. Providers noted this paradox. The superficial nature of available interventions (for example, a handout with housing resources) was an important barrier to staff buy‐in, and the interview participants expressed frustration about being unable to provide what they considered meaningful assistance. One physician said,

We need more case management. We need more for the patient, like the ability [to] give out phones. They have to be engaged in real time in the emergency room by someone who can, like, walk the walk with them, not just hand them a list of detox places. We don't do that to a patient having an MI [myocardial infarction]—we don't tell them, “Here's a list of cardiologists and cath labs; you're having a heart attack, go find your own doctor.” (MD)

Complex Patients

The manual chart review conducted at OVMC suggests that many patients identified as PEH in this study had multiple comorbidities in the biomedical, social, and behavioral realms, making them a particularly complex group of patients. Common comorbidities documented in the 116 unique patients identified in our manual chart review were hypertension (44/116 [37.9%]), diabetes (24/116 [20.7%]), substance use disorder (59/116 [50.1%]), and psychiatric disorders (47/116 [40.5%]). Ten patients were in treatment for substance use disorder, and 11 were receiving mental health care. Regarding access to care, 81.9% (95/116) of the patients whose charts were manually reviewed had insurance, 32/116 (27.6%) had LAC DHS as their designated health system, and 23/116 (19.8%) had assigned primary care providers within the LAC DHS system. Of the 32 patients assigned to LAC DHS, 14 had a primary care visit within the prior six months and 15 had a primary care visit after the index ED visit. A physician noted,

Those patients that are in the unscheduled care settings are often the most complex social needs patients. So, if your solution . . . to get those patients’ social needs addressed is just, “Well, show up to your primary care appointment, and they can connect you,” that might not be the best strategy. (MD)

Providers noted that while health systems want to engage complex patients in resources through their primary care clinic visits, many of those patients do not have telephones or transportation to scheduled visits, making unscheduled care such as ED and urgent care visits a more feasible option and suggesting the benefits of increasing the social resources available in unscheduled care settings.

Limitations of SB 1152 as a Health Policy

Providers expressed dismay that lawmakers did not consider how SB 1152 implementation would affect ED care. Many of the required interventions for SB 1152 compliance are either additions to or at odds with the ED's usual scope of care. One physician said,

Oftentimes, you create a legislation. You know, now you have to do this, and you have to do that. But then, do they give you the resources to address this? Right? Like, they didn't come together with a piece of legislation that built more shelters. I don't know that they did. I mean, it basically, it kind of puts the responsibility on the hospital. But then we're dealing with the same kind of resources that we've always been doing. So now you're going to be part of compliance, but did they really give you what you needed? (MD)

The lack of funding to support SB 1152 compliance taxed the system at multiple levels: at the institutional level, health systems lacked funds to resource the mandate; at the frontline level, providers in the ED were left to fill the gaps when resources were unavailable; and at the community level, outside social service agencies were presumably also affected by demands associated with the implementation of universal screening for homelessness in all of the EDs. “In terms of shelters, and like, just general shelters, I don't think there was much of a coordination, let's say, like, ‘Hey, you know, we have this new piece of legislation, I need your help,’” said a physician.

Communication infrastructure, funding to expand hours for social service agencies, and funding to expand capacity are prerequisites for the success of SB 1152 that are beyond the control of the health system. A physician described how the lack of sufficient support undermined the legislation's intent as follows:

The SB 1152 program basically was, you know, like, let's try to provide services for patients experiencing homelessness. But when you don't have those services, all you end up doing is, like, this endless kind of administrative documentation process without actually deploying any resources that influence patient outcomes. And that is not the goal of any social behavioral determinant screen, right? (MD)

Finally, ED providers described several unintended consequences of SB 1152. ED staff began to hear from patients that they were seeking care in the ED to obtain a meal or clothing. Some patients reported to the providers that public transit workers told them to get off the bus at the ED if they wanted a place to sleep. Another unintended consequence for the EDs in our study was confusion between SB 1152, which mandates that patients be transported to a location of their choice, and a seemingly contradictory 2008 Los Angeles City ordinance that makes it illegal for hospitals to transport patients experiencing homelessness to a homeless encampment without the patient's written consent. ²⁸

Discussion

Encounters with PEH are common in the ED, and homelessness is a well‐established predictor of poor health outcomes and early mortality. ²⁹ , ³⁰ Hospitals increasingly recognize that awareness of homelessness is critical to the delivery of appropriate medical care in the ED and sometimes a missed opportunity to intervene on unmet social needs that contribute to poor health. ²⁹ , ³¹

Samuels‐Kalow and colleagues suggest that EDs should consider universal screening for unmet social needs that affect health, including homelessness. ³² Universal screening for homelessness was not a stated objective of SB 1152 (the exact wording of the law calls on the hospital to “inquire about a patient's housing status during the discharge process” ⁹ ). However, because the EHR field for homelessness status was completed for all patients, the EDs in our study effectively achieved universal screening, though the quality of the data was questioned by participants.

In our study, 2.9% of ED patients were identified as experiencing homelessness. This rate was higher than both the rate reported by Oates and coauthors³³ in 2009 and the overall homelessness rates within the population of Los Angeles County. ³⁴ Nevertheless, the rate in our study is likely an underestimate because frontline workers are reluctant to ask the screening question and some patients who are asked do not disclose their homeless status. The operational decision on the part of the health system not to use a validated screening tool for homelessness further clouds the question of the accuracy of screening. Before implementing universal screening for homelessness, health systems must prepare their workforce to have sensitive, trauma‐informed conversations around homelessness.

We observed ample confusion about SB 1152's intent. ED providers described the intent to be an effort to resolve homelessness; however, the text of the legislation focuses narrowly on safe discharge and the prevention of patient dumping. Whereas legislators may believe the law's requirements are helpful, ED providers viewed them as superficial and a sign of the lawmakers’ lack of understanding about the level of resources needed to assist PEH. Good intentions can be misaligned with the reality of complex problems, especially in the setting of limited resources. ³⁵

Providers noted that an ED's ability to comply with the requirements of SB 1152 was shaped by the existing resources within the ED and in its local external environment. Because no funds were allocated to pay for implementation of SB 1152 requirements, hospitals and EDs used available resources to implement the intervention. The level of difficulty in achieving implementation and compliance may vary based on the hospital's existing resources and the geographic proximity of the hospital to other social services.

These are all important lessons related to the large‐scale integration of social care in clinical settings that may benefit other health systems. The National Academy of Medicine model of social care integration describes the levels as awareness of the social need, adjustment of clinical care considering the social need, assistance with the social need, alignment of the health care delivery system given the prevalence of social needs, and advocacy on the health system level for resources to meet the population's social needs. ³⁶ In this case, SB 1152 prompted universal screening that led the health system to achieve awareness of homelessness. SB 1152 also mandated, in some respects, adjustment of clinical care, as providers were prompted to take measures to ensure that PEH were given additional resources such as a supply of their chronic medications. Where SB 1152, as an unfunded mandate, fails is at the level of assistance. The inability to address patients’ housing needs was a source of dissatisfaction with SB 1152 legislation and a source of moral injury for providers wanting to provide meaningful assistance.

Policy Recommendations

Based on our interpretation of the study data, we recommend the following:

• Clarifications to SB 1152 are necessary to address the differences between inpatient and ED discharge processes, determine how frequently an ED must deliver the SB 1152 bundle of services to a single patient, and establish expectations for compliance during off‐hours when social services are unavailable.

• Because homelessness cannot be resolved in a single ED visit, the state should provide funding to support housing‐focused case workers that will follow PEH through the transition from temporary shelters to permanent supportive housing. Medi‐Cal could fund the delivery of the SB 1152 bundle of services to defray the costs to public hospitals that provide care for high numbers of PEH; this Medi‐Cal initiative could be accomplished through the waiver program which allows for additional services to be provided to certain groups of individuals.

• California legislators should consider complementary legislation to increase funding for shelters so that sufficient capacity is available to accept PEH from EDs and hospitals, and to fund alternative strategies to prevent poverty and the upstream root causes of homelessness itself.

Limitations

Given COVID‐19‐related research restrictions during the time of the study, we were not able to interview patients who received the SB 1152 intervention. This limitation to our study is particularly important because patients who screen positive for social risk (an individual‐level adverse social determinant of health) do not always perceive a social need. ³⁷ , ³⁸

Another potential limitation of the study is that the compliance rates found in the two‐month study period may not be representative of rates for other periods. Finally, the interviews themselves may have affected the rates of compliance, perhaps leading to greater compliance if they serve as a point for self‐reflection.

Conclusion

In response to SB 1152, we implemented universal screening for homelessness in a public hospital system with limited resources. SB 1152 has complex requirements and is difficult to comply with, even in a mission‐aligned system with provider buy‐in. California policymakers would be well served to critically evaluate the current law, clarify its intent, and consider Medi‐Cal funding to ensure that SB 1152 has a meaningful positive impact for PEH.

Funding/Support: Authors Aleman, Taira, and Tucker‐Seeley received grants from the California Health Care Foundation during the conduct of this study.

Conflict of Interest Disclosures: All authors completed the ICMJE Form for Disclosure of Potential Conflicts of Interest. No conflicts were reported.