Barriers to Healthcare for American Indians Experiencing Homelessness

Abstract

Members of American Indian (AI) communities face many barriers to receiving both mental and physical healthcare. These barriers can have a negative effect on overall health. Barriers are compounded for AIs who are also experiencing homelessness, and AI make up a disproportionate percentage of the homeless population nationwide. In-depth semi-structured interviews were conducted with 12 service providers and 16 homeless participants in a mid-size Midwestern city to identify barriers to care for homeless participants. Key barriers identified in this study for homeless participants were: transportation, phone accessibility, discrimination, and cold and bureaucratic cultures of healthcare systems. Major barriers identified by service providers were: access to care, discrimination and mistrust, and restrictive policies. Given generally higher disease prevalence within the homeless population and health disparities within the AI community, steps should be taken to reduce barriers to healthcare.

American Indians and Alaska Natives (AI/AN) comprise a disproportionate percentage of the 209,000 people who were counted as “homeless” in the 2010 Census (U.S. Census Bureau, 2010). Among AIs in the northern Midwest, over one-third had experienced homelessness at some point during their lifetimes (Whitbeck, Crawford, & Sittner Hartshorn, 2012), and homelessness is a key determinant of poor health outcomes (Wilkinson & Marmot, 2003). Nonetheless, research on the intersection of homelessness and race is limited. In this study, we explored barriers to both medical and mental healthcare for homeless AIs in the northern Midwest United States (U.S.) via qualitative interviews with homeless AI adults and AI-serving health and human service workers. Unless specifically identified, the use of the terms healthcare or barriers throughout this paper refers to general medical and mental healthcare or barriers to both medical and mental healthcare.

Barriers to Healthcare for Individuals who are Homeless

Homelessness entails a daily struggle for the essentials of life. Barriers to healthcare among homeless populations in the U.S. include such factors as discrimination and stigma (Wen, Hudak, & Hwang, 2007), limited clinic hours (Ensign & Panke, 2002), and transportation problems (Gelberg, Browner, Lejano, & Arangus, 2004). In one study examining hospitalizations, homeless individuals were admitted to hospitals five times more frequently than the general population (Martell et al., 1992), and for most homeless individuals living in the U.S., lack of health insurance is a major obstacle to obtaining medical, mental, and preventive healthcare (Stark & Jahiel, 1992).

Similar to other marginalized populations, homeless women reported seeking advice from non-professionals and trying self-care interventions prior to seeking help from a clinic, in part due to a lack of trust in service providers (Ensign & Panke, 2002). In a qualitative study, a majority of the homeless participants cited experiences of not being welcomed in clinics as overt acts of discrimination (Wen, Hudak, & Hwang, 2007). Gelberg and colleagues (1997) also found that individuals struggled with unattainable health recommendations regarding rest and changes in diet when providers did not take into consideration their homeless status.

Pragmatic barriers also impede access to healthcare. Transportation was a special hardship for homeless women living in shelters with required check-in times. This limited ability to attend medical appointments and arrange transport within the constraints of clinic and shelter hours (Gelberg et al., 2004). Not all homeless needs arise during business hours, and shelters with shutdown hours limit the availability of those using these emergency services (Ensign & Panke, 2002). Beyond these issues, there is also the potential for competing priorities such as finding a home, a job, and getting care for children thus leaving healthcare needs to be viewed as discretionary (Gelberg, Gallagher, Andersen, & Koegel, 1997).

Barriers to Healthcare for AI/AN

For AI populations in general, there is a high proportion of unmet physical and mental health needs (Duran et al., 2000; Beals et al., 2005; Whitbeck, Johnson, Hoyt, & Walls, 2006; Indian Health Service, 2014). AI/AN were more likely to report discrimination as a barrier to care than other racial/ethnic minority groups (Johansson, Jacobsen, & Buchwald, 2005). Similarly, in a study of over 50,000 participants, the only racial category that was a positive predictor of reporting discrimination was AI/AN (Trivedi & Ayanian, 2006).

In a study of enrollees in public healthcare programs, AIs reported experiencing a higher percentage of barriers than those of other ethnic backgrounds. These barriers included discrimination, access, and transportation difficulties (Call et al., 2006). Sixty-eight percent of participants from a Western tribe stated that they had no transportation to get to their appointments (Duran et al., 2000). Even when individuals can physically access Indian Health Services (IHS), their needs may not be met. In the IHS 2014 Fiscal year, an estimated $761 million was denied for services to eligible AI/ANs due to underfunding from Congress (Purchased/Referred Care, 2014), resulting in many being left to utilize urgent or emergency care (Kaiser Family Foundation, 2014). Geographic barriers also exist. Although 71% of AI live in urban areas (Health Disparities, 2015), Urban Indian Health Organization primary care services are funded at 22% of the level of need (IHS, 2016). In some instances, the nearest clinic may be located outside the city on a reservation (Marrone, 2007).

Differences in communication styles between some AI individuals and healthcare providers have been identified as a further barrier to quality care (Tripp-Reimer, Choi, Kelley, & Enslein, 2001). Differences in patient/provider values and experiences can influence how patients discuss symptoms, what they feel comfortable sharing, and their preferences for communication (Marrone, 2007). Microaggressions – a more subtle, but frequently occurring form of discrimination - also can inadvertently be introduced into the interaction (Walls, Gonzalez, Gladney, & Onello, 2014).

Study Context and Aims

The experience of being both homeless and AI can create compounding barriers to care. Despite considerable research on barriers to healthcare for homeless individuals and barriers to healthcare for AIs, the overlap between the two groups has been minimally explored. In addition, studies have generally examined only providers’ perspectives or perspectives of homeless individuals; this study interviewed both. Qualitative interviews of AI homeless adults and homeless-serving health and human service providers to examine wellness and barriers to care.

Methods

Participants

Data for this study were collected in Minnesota, in a northern city that experiences harsh winter weather conditions dangerous for homeless individuals. One tribal reservation is in close proximity to the city and other reservations are several hours away. Within this city, 29% of homeless individuals were AI in 2011, compared to the 2.5% of AI in the city’s overall population (U.S. Census Bureau, 2010; Wilder Research, 2011).

The study included 28 participants; 16 self-identified as experiencing homelessness and 12 were service providers (e.g., social service workers, case managers) working in agencies that served those of lower socioeconomic status. The mean age of homeless participants was 37 years (range = 19–57; SD = 13.1 years) with an equal ratio of males to females (n = 8 each; 50%). The mean age of the service providers was 45 years (range = 26–57; SD = 9.9 years) and a majority were female (n = 8 female, 67%; n = 4 male; 33%).

Procedure

All study protocol was IRB approved. Recruitment of AI homeless adults was possible with the assistance of multiple local service agencies using convenience and purposive sampling methods. Each agency agreed to recruit three to five interview participants using approved informational brochures that contained eligibility criteria and study contact information for interested participants. Recruitment of healthcare/social service providers occurred via snowball sampling. Three initial providers were invited to participate. Target service providers for this project were those who specialized in working with the homeless population as counselors, case managers and those assisting with housing programs. They were then asked to recommend/nominate two or three additional colleagues to be invited for study inclusion. These new participants likewise recommended colleagues for participation. We recruited additional participants until saturation.

Interviews were conducted by University project staff between 2013 and 2014. All participants provided informed consent. Interviews were conducted individually in private settings. Most often, AI interviews took place at service agencies and service provider interviews took place at the University or in providers’ offices. Interviews lasted from 12 to 66 minutes with an average length of 37 minutes. Individuals who participated were given a $20 gift card.

Data Transcription and Analysis

Research staff electronically transcribed all audio-recorded interviews verbatim. Data analysis was performed using qualitative data software NVivo10 (QSR International, 2012) in a series of 3 distinct steps. Three coders were assigned transcripts so that each interview was independently read and analyzed by two members of the team. We used a general inductive approach to analysis such that themes were derived directly from the data (Thomas, 2006). The first analysis step involved a process of open coding during which researchers read each transcript carefully, marking potential trends in the data and developing a series of memos or notes to organize and flag emergent themes (Creswell, 1994). Team members then re-read the transcript to assign initial codes (“nodes” in NVivo) identifying general themes. Second, the two coders assigned to each transcript met to compare thematic results; overall, identified themes were highly consistent between coders. In the few cases where inter-rater agreement was not achieved, a third researcher familiar with the data provided input on final theme identification. These inter-coder meetings resulted in a single list of concrete themes for each transcript encompassing a range of topics such as conceptions of wellbeing, perspectives on integrated care, and health risk behaviors. For the current manuscript, the research team moved to a third step to carefully identity any themes in the data indicative of barriers to healthcare. Through this focused coding procedure, we deliberately scanned our open code lists to identity recurring themes, combine or refine open codes where relevant, and select transcript excerpts related to each major theme (Charmaz, 2006).

Results

Healthcare Barriers Reported by Homeless Participants

Transportation

Multiple AI homeless participants reported lack of transportation to healthcare facilities as a frequent barrier to care. For many, walking was their predominant form of transportation given the financial burden associated with other modes of transport. One participant shared: ‘If I have enough funds to get on the bus or anything’ (Male Homeless 8). Other participants were reliant on others for rides. A male participant stated that his group living situation was remotely located, and he was only able to carpool into town a few times a week due to time and gas expenses: ‘If there’s only one person or two people going into town then it’s not worth the ride’ (Male Homeless 6). The uncertainty associated with these types of transportation are compounded by the second barrier described below, phone accessibility. If transportation falls through and individuals do not have phones to cancel their appointments, they often experience even less access to care as a result of these missed appointments. Further, mental health issues sometimes intersected transportation and communication barriers, as evidenced in this excerpt:

I have social issues, it’s hard for me to get on the bus and I didn’t have a ride that day, my worker didn’t, I wasn’t gonna meet with her that day, so, I was not able to get down there, to the appointment, so that was the last I ever seen of them. And they were quick to send me a letter to seek services elsewhere (Female Homeless 5).

This quote illustrates the overlapping nature of barriers to care: anxiety, transportation, and restrictive policy barriers combine here to prevent service utilization.

Phone Accessibility

The homeless participants frequently discussed a lack of consistent telephone access and the negative impact this had on making and keeping physical and mental health appointments. Barriers as a result of limited phone access ranged from difficulty scheduling appointments to not receiving reminder calls to being unable to cancel appointments when needed. As described above, missing appointments can lead to service refusals from facilities. Being able to contact a clinic when an issue arises is critical to continuing to receive services. The following quote demonstrates these challenges:

When you’re homeless on the streets it’s kinda hard for them to call you and remind you of your appointments and stuff ‘cause you know you never know where you’re at and a lot of the time you don’t have a phone or access to a phone (Male Homeless 3).

To take it a step further, another participant described phone access as a necessity to accessing services:

Well… you got to have a phone. I don’t have a phone, I can’t call them, I can’t have them call me….they [local services] help you out a lot, it’s just that you have a phone….that’s the only way you can seek contact, so there’s no other way (Male Homeless 1).

This participant is able to clearly identify phone access as a barrier to care. Yet, the systems in place to serve him have not been able to see or work around this issue to provide patient-centered care that would meet his needs.

Further reflecting the cycles that lead to poor care for homeless AIs, the following demonstrates not only the challenges presented by a lack of phone accessibility, but how not being able to cancel an appointment can lead to discrimination and judgment:

There was a different doctor who I went to besides the (Local clinic), I was trying to seek help with other things, doing med management with a doctor, and she kinda prejudged me right off that if I missed any appointments, they were just going to drop me off their schedule, kick me out of the clinic, which they did. A lot of people in my situation with mental health problems it’s not always easy getting to your appointments (Female Homeless 5).

In this case, the participant wanted to take care of her health issues and attend appointments, however, access barriers that prevented her from doing so emerged as a major theme in the homeless data.

Discrimination

Homeless participants identified discrimination during visits to healthcare stemming from several factors including ethnicity, gender, socioeconomic status (including homelessness), (perceived) substance use, and appearance. Most homeless participants reported experiencing discrimination on some level, with a majority recounting discriminatory experiences because they were AI: ‘Just the way you are treated when you go in there you get treated if you’re different color. If you’re not the so called supreme race, the white race, you ain’t nothing to them’ (Female Homeless 4). This was supported by another participant:

You know, I hate to say it, because I don’t like to… think that that exists, but it does…I think yes, it has affected actions with doctors and hospitals… I haven’t lied to them about anything, and I just feel like sometimes that…they choose not to hear me. And do I want to call it a race thing? I don’t want to call it that, but I do believe that’s what it is, you know? (Female Homeless 2).

These experiences clearly affect the patient-provider relationship, the quality of care provided, and the willingness of patients to engage with the healthcare system.

Providers’ stereotypes regarding “drug-seeking” were commonly identified by AI homeless participants as another major source of discrimination: ‘For some reason they (medical providers) seemed to be like, ‘we’re on a hunt for drugs’, you know’ (Female Homeless 7). Participants felt that healthcare providers came to these conclusions entirely based on AI appearance or tribal membership. This resulted in participants feeling uncomfortable, upset and unwilling to return to healthcare facilities: ‘And I’ve had a lady comment to me about that they didn’t serve alcohol there, and so yeah, I got pretty aggressive on that, on that note’ (Female Homeless 7).

To summarize, homeless individuals have limited options for service facilities. Participants were further limited by access, such as a means of transportation and communication. Even if those barriers were overcome, participants frequently experienced discrimination on the basis of personal characteristics like race, socioeconomic status or both.

Cold and Bureaucratic Culture

This theme emerged from stories about discourteous and undignified treatment from healthcare providers. This theme encompassed discrimination and yet was broader in nature. Participants in this study felt misunderstood by healthcare providers and a lack of connection when asked why they were seeking care. One such negative experience can have a major impact on patient perceptions of care. For example:

I actually had an experience where she never remembered who you are…you don’t want to go into a place where you’re just another number to them. You kinda want to go in, you want to feel comfortable with these people enough to talk about your problems and what’s going on with you and it’s really hard to do that a lot of people, a lot of Native peoples really that’s kind of offensive to them, I know I was personally offended by that whole situation, you know…she didn’t even remember my face (Female Homeless 5).

Some individuals feel unheard and diminished as a result of negative interactions. For example: ‘The things that I’ve been through I’ve always had people say well, do you have someone that can say that they saw this happen’ and I’m like “no” and then they didn’t believe me after that’ (Female Homeless 6). Several participants in this study were generally apprehensive and had difficulty trusting others. This permeated all of their interactions: ‘I ain’t to trusting not any more anyway. I use to trust a lot of people and don’t no more’ (Male Homeless 7). Trust in providers is a key prerequisite for seeking care and following through with provider recommendations.

Barriers Reported by Service Providers

Access to Care

Service providers reported many of the same barriers identified by homeless participants, with some differences in framing and depth of understanding. For the homeless AIs, transportation and phone access were the most frequently discussed barriers and the most commonly occurring themes. Service providers tended to talk about this in terms of broader “access” problems as opposed to identifying the specific details or struggles related to transportation or phone difficulties. For example, ‘…just trying to get access to healthcare, is difficult at times’ (Male Service Provider 4) and ‘…to get in and get meds taken…or set up a therapist and appointments, it’s pretty difficult’ (Female Service Provider 6). These examples highlight issues that anyone in the general population may be presented with, not the extent to which this affects the homeless AI population.

Providers also discussed how access issues led to urgent or emergency care visits: ‘A large number of them use the emergency rooms at any hospital’ (Female Service Provider 8). In essence, the providers saw how the emergency room became the primary care office for many homeless clients, in part due to basic access barriers. What was less frequently discussed by providers was the connections and compounding factors that caused barriers and healthcare problems to proliferate. Many did not acknowledge how facing these barriers often left utilizing urgent care or emergencies rooms as the only option.

Discrimination and Mistrust

Responses from service providers, like AI homeless participants, also revealed discrimination and mistrust as a problem in the healthcare system:

I really believe and have seen that some of the medical providers in town, treat Native American people different than non-Native American and I’ve seen it when some of our Native American young women have gone to have kids, and social services is right there, or the hospital treats the Native American people different than they treat other people, and I think that sometimes they get a bad rap that they’re trying to get medication because they’re addicted when actually they aren’t and so, yeah we’ve seen a lot of that going on, yeah (Female Service Provider 7).

Providers very frequently cited discrimination in the local healthcare system, but almost exclusively outside of their own practices or agencies. Another participant noted a cultural difference in the trust building process:

I would say that there is definitely with a lot of the Native American [clients] that I have worked with, the trust building process is um, very different. There’s a lot of reluctance about coming into an agency like [facility], especially the fact that all of our staff are primarily Caucasian, I think that that can be a little intimidating of an environment but even, even with that once they get to know the staff the trust building timeline kind of is much longer (Female Service Provider 1).

Thus, this provider described AI skepticism for formal systems of Western care, which has been previously discussed in relation to historical trauma, medical racism, and medical mistreatment (Marrone, 2007; Walls, Johnson, Whitbeck, & Hoyt, 2006; Burgess, Ding, Hargreaves, van Ryn, & Phelan, 2008). Still, developments in cultural safety remind providers to allow each individual patient to express his/her viewpoints and concerns without presumptive conclusions based on cultural or ethnic background.

Restrictive Policies

Service providers in this study expressed frustration with rules and regulations that made it challenging for them to meet the needs of their clients. AI individuals also discussed some of these frustrations, but largely in terms of access instead of systematic policy problems. Many facilities have missed appointment policies that can prevent individuals from seeking care at that specific clinic.

Missed appointments a lot of times, there’s like a rule, three strikes you’re out type of thing, so if somebody either pulls a no-call or no-show, or tries to show up intoxicated sometimes they’re asked not to come back, so we’re left searching for a different provider for them (Female Service Provider 5).

Additional themes unique to service providers included the discussion of collaboration across diverse agencies. While collaboration was predominately positively viewed in, issues with not understanding the struggles of homelessness arose internally and from within other services and management:

I think there’s sometimes a disconnect within the organization between management and what people are really seeing in the field that can kinda get, would be hard to get some things done. But I mean people are always trying to provide the best care out there for their clients (Male Service Provider 4).

I think the number one problem I have is internally, and not in the immediate office, but again in the larger entity, and with other service providers there’s a great misunderstanding as to how homelessness impacts, it’s like a spider web, it’s in everything. So if they’re upset this person isn’t working on their sobriety all I can say is, ‘How well would you do that yourself?’ and I try to make it very personal so that they can understand you cannot have a lot of normalcy if you don’t have a place to live. (Female Service Provider 8).

Also highlighted by service providers was the frustration and difficulty complying with systemic policies and procedures that do not work well for homeless patients:

A classic thing would be some places are open from 8 to 4. We know that homeless things happen, some of our shelters- homeless shelters- expect people to be in there at 6. If a person ends up becoming homeless, through domestic violence or anything from 4-6, they don’t [receive] services until the next day. That means that we have to meet their needs between 4 at night until 8 o’clock the night morning… homeless or issues come up at odd hours… (Female Service Provider 4).

Discussion

This study is important for several reasons. First, we examined barriers to healthcare as documented by both service providers and homeless individuals. Past studies have explored barriers to care for those experiencing homelessness predominantly through either the perspective of the service provider or the perspective of the homeless individual (Call et al., 2006; Clucas & StClaire 2010; Garroutte, Sarkisian, Goldberg, Buchwald, & Beals, 2008; Gelberg et al., 1997; Wen et al., 2007). Second, this study focused on barriers to healthcare for AIs who are experiencing homelessness. Many studies have looked at barriers for homeless individuals, but few of these focused on AI homeless participants (Wen, Hudak, & Hwang, 2007; Ensign & Panke, 2002). Given the disproportionate representation of AIs among the homeless (Whitbeck, 2012; US Census Bureau, 2010), it is important to extend the homeless literature to include the AI population. Themes that emerged were not sole issues, but rather each in some way overlapped with at least one other barrier creating a larger, compounded challenge. Barriers that were reported were: transportation, phone access, discrimination (based on race/ethnicity or socioeconomic status), cold and bureaucratic culture, access to care, discrimination and mistrust, and restrictive policies.

Comparisons between Homeless and Service Provider Responses

A major challenge faced by most, if not all, AI homeless in this study was dealing with discrimination. Service providers agreed and stressed how often they heard from AI individuals about instances where discrimination played a central role in discontinuation of service usage with specific providers or facilities. While both service providers and homeless AI identified discrimination as a barrier, the descriptions provided by the homeless participants more vividly highlighted the discrimination they faced.

For homeless AIs, seemingly basic needs such as transportation and phone access specifically were cited as major challenges. They considered these challenges to be of lesser immediate importance survival and perhaps also unattainable on a consistent basis. Service providers also identified this barrier and took it a step further by inferring that it directly connected to participants’ spotty use of preventative care.

While service providers acknowledged the complex individual-level and systemic barriers faced by their homeless clientele, many still expected homeless people themselves to adapt to fit into the current healthcare system. When providers discussed prohibitive institutional policies, many expressed passive acceptance to the systemic flaws rather than seeing the disconnect as a call to action. This may be due to provider burnout or even perceived helplessness. For homeless participants, the frustrations of a rigid system of care frequently meant that they stopped seeking out or receiving services beyond emergency and urgent care.

Overlapping Themes: The Interconnection of Barriers

Our findings reinforce and add to conclusions from past studies in that many of the barriers to care discussed by participants such as transportation (Gelberg et al., 2004), discrimination (Johansson et al., 2005), and access (Call et al., 2006), have been previously identified for homeless individuals or AIs specifically, but rarely in the same study. Reflections from our participants suggest that these barriers to care can interact and proliferate to create deeper obstacles. Each of these can be individually daunting, but adding in mental and physical health needs creates compounding barriers and complications. Perhaps being faced with so many barriers to care is one of the factors leading to the disproportionate health rates seen in the AI population. Regardless, effort should be made to minimize the barriers AI’s face when seeking healthcare services. A possible first step for change is acknowledging the interconnectivity of barriers.

Another example of interacting themes was the connection between feeling disrespected and the level of trust between patient and provider. Feelings of disrespect lead to greater mistrust of providers and dissatisfaction with the services (Clucas & St Clair, 2010). Those unwelcoming experiences and past encounters with discrimination increase the chances that the individual will feel dehumanized resulting in a decrease in likelihood of continuing to seek services (Wen, Hudak, & Hwang, 2007) as well as impacting how they respond to future interactions at the same or similar institutions.

This lack of trust has historical and contemporary underpinnings. AIs are underrepresented among physicians and other health professionals (Cooper & Powe, 2004). When patient and provider are of different backgrounds, there is increased potential for cultural differences to act as a barrier. A history of medical racism, historical trauma, and medical procedures performed without consent intersect such that AI community members may justifiably lack trust in governmental, medical, and institutional settings and providers (Carrese & Rhodes, 2000; Zerger, 2004). As a result, AIs may be more likely to delay entry into the healthcare system until later stages of disease or illness and may have higher rates of non-adherence to advice given by providers (Zerger, 2004).

The Many Faces of Discrimination

As previously mentioned, discrimination was a major barrier to healthcare reported by both AI and service provider participants. In this study, individuals who experienced discrimination first attributed it to their ethnicity and second to their homeless status, with these intersections creating a compounding discrimination effect. These experiences of discrimination were not exclusively at the hands of the provider, but from non-clinical staff as well. This speaks to the need for revising the system and the communication that occurs at every level of the healthcare institution. Experiences of discrimination decrease willingness to access care or to adhere to what is being recommended (Pascoe & Smart Richman, 2009). The system needs to do more to build trust and engender a feeling of “cultural safety” among patients, which could start with providers examining their own biases and reactions to patients (Epstein & Street, 2011). Providers can also work with the patient to create health plans that fit their life styles and beliefs which will overall lead to higher levels of understanding and adherence.

Limitations

One limitation of this study was the range of service provider specialties represented in the provider interviews. None of the participating service providers were medical professionals (e.g., physicians or nurses). Rather, our participants were employed in social, human, and mental health service arenas. Other limitations include a geographically limited sample that is not representative of all AI tribes. Another weakness was only sampling homeless participants from within the healthcare system, thus potentially missing the most marginalized individuals who are not receiving services.

Implications

The results from this study reveal modifiable barriers to care that can be addressed through policy and system changes. Basic efforts could be made to increase availability of transportation, phone access, and creating flexible schedule structures for services. Improvements in cultural safety training (Epstein & Street, 2011) for healthcare providers may increase trust, understanding, and comfortability for help-seeking AI adults. Many providers dedicate their lives to service professions in underfunded and understaffed agencies. A very high proportion of AIs live in urban areas (Health Disparities, 2015), and yet urban service options are largely underfunded (IHS, 2016), often forcing AI individuals to go miles to reach the nearest IHS facility. To meet homeless AI needs where they are at, travel distances need to be limited by dedicating more funding to meet urban care needs. Within a majority based system, there are many cultural divergences that can lead to disparity-creating policies (Gone, 2007). However, there are clear steps that can be taken based on the experiences of these individuals to improve systems of care. In addition to basic changes described above, systems would do well to engage AI clients in decision-making collaboration, training providers, staff, and system leaders in basic principles of culturally safe care, and enacting basic policy changes that prevent the practice of systematically cutting off those most in need of receiving services.