Patient-centered values and experiences with emergency department and mental health crisis care

Abstract

Little is known about what patients value in psychiatric crisis services or how they compare community-based services with those received in the emergency department. Three focus groups (n=27) were held of participants who had received psychiatric crisis services in emergency departments or a community mental health center. Participants described care experiences and preferences. Focus groups were audio recorded, transcribed, and coded using a value-based lens. Themes included appreciation for feeling respected, basic comforts, and shared decision-making as foundations of quality care. Participants preferred the community mental health center. Research should address long-term outcomes to motivate change in psychiatric crisis care.

Introduction

Rising rates of use and workforce shortages challenge emergency departments (EDs) as they work to maintain their mission of serving all patients regardless of their acute medical issue or payer status (deBruyn 2017; Solberg et al. 2016; Sharieff et al. 2013; O'Malley et al. 2005). Emergency department use by people with psychiatric needs, especially those with serious mental illness, contribute to these challenges. People with psychiatric needs have higher rates of ED use and longer ED stays than others (Nolan et al. 2015; Hendrie et al. 2013; Centers for Disease Control 2013; Slade et al. 2010; Owens et al. 2010). Further, their rates of ED use have been increasing over time, as hospitals close and community care options leave needs unmet (Swartz 2016; Appelbaum 2015; Halmer et al. 2015). They are more likely to use the ED when they have multiple or more severe psychiatric symptoms (Lunsky et al. 2015; Nossel et al. 2010), assault-related injuries (Cunningham et al. 2015) and co-occurring chronic conditions such as diabetes, chronic kidney disease, or sleep problems (Sajatovic et al. 2016; McPherson et al 2015; Kaufman et al. 2011; Hackman et al. 2006). Common factors associated with extremely long ED stays, or psychiatric boarding, are suicidality, being uninsured or having public insurance (compared to commercial) and being admitted or transferred (rather than discharged) (Pearlmutter et al. 2017; Smith et al. 2016; Misek et al. 2015; Stephens et al. 2014; Weis et al. 2012; Wharff et al. 2011; Park et al. 2009). Combined, these factors set apart people with psychiatric needs an important segment of the ED consumer population.

Available information on satisfaction with ED services among people with psychiatric needs focuses on areas requiring improvement. Patients and staff desire improvements in care settings and systems, including quiet and safe spaces, shorter waiting times, and better patient-provider communication (Olive 2016; Broadbent et al. 2015; O'Regan et al. 2009; Griswold et al. 2008). People with psychiatric needs report feeling stigmatized and discharged without anticipated treatment (Wise-Harris et al. 2016). Nursing staff report awareness that manual restraint can lead to psychological, as well as physical, injury (Chapman et al. 2016), but feel that workload demands and limited training prevent more personalized care (Wand et al. 2016; Wand et al. 2015; Saurman et al. 2015; Montreuil et al. 2015; Jelinek et al. 2013; Dunnion and Griffin 2010). Evidence on the positive impacts of suggested improvements in EDs is lacking.

Dedicated regional psychiatric crisis care may shorten ED length of stay and allow stabilization that obviates the need for inpatient psychiatric care (Zeller et al. 2014). And, there is emerging evidence that residential crisis care is less expensive than ED use, achieves similar community outcomes for mental health patients, is valued by users and, when combined with self-referral, enhances users' confidence and ability to cope (Olso et al. 2016; Hamilton et al. 2016; McKenna et al. 2015a & b; Fenton et al. 2002).

Wake County, a large urban county in North Carolina, has developed an emergency medical services alternative-destination program to transport 9-1-1- patients who need psychiatric crisis services, but not acute medical, care to WakeBrook, a new community mental health center that provides psychiatric crisis and residential care. WakeBrook is situated on a medical campus shared with WakeMed Hospital which has a level I trauma center ED with a dedicated children's emergency department. WakeBrook provides 24-hour crisis and assessment services which serve as the entry point to its longer term psychiatric services. While average ED wait times across the state are less than 40 minutes, patients needing long-term psychiatric hospital beds typically wait just under four days (Hoban 2016). A primary goal of WakeBrook crisis and assessment services is to alleviate this backlog by avoiding unnecessary transport to EDs, psychiatric hospitals, and criminal justice system facilities; stabilizing acute crises of patients and assigning appropriate psychiatric dispositions. In the face of statewide concerns over inadequate resources for mental health services and a constantly changing practice landscape, WakeBrook represents a promising model for crisis care that can overcome, for Wake County, some of these issues (Creed, et al., 2016; La et al. 2016; Luthra 2016; Swartz 2016; Thomas et al. 2012; Thomas et al. 2009; Swartz, 2003). While WakeBrook has successfully served a significant proportion of patients who would otherwise have been taken to an ED, little is known about what psychiatric patients value in psychiatric crisis services or how they compare dedicated psychiatric crisis services such as WakeBrook's with typical ED services for psychiatric patients.

The present study addresses this gap through qualitative analysis of three focus groups. Participants had received psychiatric crisis care in area EDs or at WakeBrook in the past year. We asked participants to describe their care experiences and explain their preferences for each setting. The goal of this study was to develop a better understanding of what patients with mental health and substance-related disorders value in order to inform county and state policy on psychiatric crisis services. This paper contributes to the literature by capturing how individuals who use psychiatric crisis care describe their ED and community mental health center stays including detailed assessment of accommodations, meeting of basic needs, infrastructure, staff interactions, and their assessment of the appropriateness of care. It concludes by drawing out their preferences and reasons for them.

Methods

Study design

A qualitative focus group design was used to explore participants' experiences and values about psychiatric crisis care. Focus groups were held in order to develop group perspective: 1) across emergency service settings to compare those settings, 2) across time to facilitate making judgments in the present about the quality of care during a time of crisis in the past, and 3) across participants to develop a group understanding of common and unique experiences. Focus group questions were developed through an iterative approach to solicit participants' understanding of and values regarding care. The groups explored their experiences with accommodations, interactions with providers and staff, medication and treatment plans, role of family and advocates, length of stay, and follow-up care (Supplement 1. Focus Group Guide). The groups also discussed differences in valued attributes of psychiatric crisis care between emergency departments and WakeBrook and their preferences for one setting over the other. Groups were scheduled in succession until saturation of themes was achieved: when we began to hear similar answers to our questions and no new information was volunteered.

Recruitment

Participants were screened and recruited from the WakeBrook Primary Care Clinic, local National Alliance on Mental Illness (NAMI) support groups, and other local psycho-social support groups. Inclusion criteria included being age 21 years or older, English speaking, able and willing to provide written consent, able to attend a focus group, recommended for study participation by clinical or social work staff familiar with their diagnosis, and receiving psychiatric crisis care at an ED or WakeBrook in the past year. When participants arrived for the focus group, a member of the research team privately confirmed each participant's capacity to consent by administering a three-item questionnaire. Each patient was asked to describe in his or her own words: 1) the purpose of the study; 2) the benefits of the study; and 3) the risks of the study (Palmer et al. 2005; National Institutes of Health 2009). Informed consent was obtained from all individual participants included in this study. The study design was approved by the Office of Human Research Ethics of the University of North Carolina.

Setting

Three focus groups (n=6 to 11) were conducted with a total of 27 participants from the Wake County, North Carolina area. Wake County is a large fast-growing urban county with just over a million residents. The focus groups were held in a community room at WakeBrook. WakeBrook was selected as the site for the focus groups because it is located on a bus line, offered a comfortable environment, and was familiar to many of the participants. No providers were present during the focus groups to allow for open exchange, but medical and psychiatric providers were on hand in the event a participant required immediate care.

Discussion was guided by a study team member who is an experienced mental health services researcher and who has conducted research interviews and focus groups with patients with mental illness. A research assistant was also in the room to audio record the discussion and take notes on non-verbal cues. All participants were provided with lunch and received a small cash token of appreciation to compensate for travel cost and time.

Data analysis

Each focus group was audio recorded and transcribed; demographics (i.e. gender, race, ethnicity, age group) were also noted. Our coding scheme incorporated a values coding framework, in which participants assigned value to their crisis care experiences (Saldana, 2009). The coding process was informed with a value-based lens in order to explore intrapersonal and interpersonal participant experiences. Values codes were determined both a-priori, to inform research questions of a larger study, and post-hoc, to capture themes important to participants. Following initial coding, the team also analyzed cross-cutting themes present in all focus groups in order to capture valued attributes of crisis care that emerged in group discussions. Coding and thematic categorization was conducted manually by three members of the research team each blinded to the coding decisions of the others; discrepancies were reconciled through discussion (Saldana 2009). In order to further validate features and attributes of valued crisis care, the research team confirmed final themes with one stakeholder involved in mental health care advocacy and one stakeholder provider of psychiatric crisis care (Saldana, 2009).

Results

Participant Characteristics

Participant characteristics (n=27) reflected a range of demographics: half (52%) were men; one-third were minority race or ethnicity; ages ranged from 20s to 60s.

Themes

In the following sections, we present themes related to participants' values and preferences in crisis care services. Table 1 summarizes the various features of care and their attributes commonly cited by participants across focus groups. Patient identified attributes of each feature highlight the unique aspects of care that are valued in the psychiatric crisis experience, and range from structural issues related to crisis care facilities to interpersonal preferences when coordinating with clinicians and staff.

Table 1. Features of psychiatric crisis care and their valued attributes.

Feature	Valued characteristics
Accommodations	Placement in a private area/room as opposed to a public space Hot blankets, pillows, windows
Basic needs and infrastructure	Access to food, Comfortable and clean garments, Adequate sleeping arrangements, Hygienic bathroom spaces, and Materials to support meaningful use of free time
Freedoms	Involuntary commitment can be appropriate for safety and to meet referral criteria Transparent and open communication with providers regarding their care Outside recreation
Communication	Clarity and timeliness of communication, so that there was a mutual understanding of the care process Conversation is especially important around difficult topics, e.g. health risks, restraint, involuntary commitment Common courtesies and respect are a critical foundation of fruitful conversation Kindness is greatly appreciated Shared decision-making and self-advocacy can be planned for ahead of time, even if it is not always possible during psychiatric crisis
Length of stay	Shared decision-making in length of crisis care Understanding that extended stay was associated with improved treatment options, medication management, and follow-up referrals Enhanced rapport with crisis staff to shorten wait time
Visitation during crisis care	Presence of family and friends in crisis de-escalation Pre-defined visiting hours for guests Opportunities to approve or exclude potential guests during crisis care stay
Follow-up care	Discussion of options so that referral site is matched to preferences Referral set up prior to discharge, including calls and paperwork Attention given to how the referral site is working for that person, so that poor-fit referral is not repeated

Accommodations

The focus groups included discussions of the physical space in which participants were placed while they received care during a psychiatric crisis. Participants described their accommodations and discussed whether or not they thought they were appropriate and what they valued in these spaces. Participants valued having privacy. One participant explained: “I like my time alone.” Participants valued their privacy for a variety of reasons. One participant expressed that she “didn't want people to see [her] like that.” A few participants mentioned discomfort in being placed in a room with multiple people while waiting for a private room to open: “Here [WakeBrook] I've been…with 3 or 4 people in the TV room.” Some participants reported being able to overhear private conversations between other patients and their providers in the emergency department: “Sometimes they're asking a lot of personal questions and you can hear the guy or girl answer.” Interestingly, while participants expressed discomfort from hearing other patient's personal health information, they did not mention other patients overhearing their private conversations. Additional items were mentioned by participants as providing comfort during their stay. For example, one participant mentioned: “they give you a hot blanket and pillow in the emergency room…I've always been put immediately into [a room], maybe it's curtained off. With a nice bed and pillow.”

Basic Needs and Infrastructure

Participants described the dignity of having their basic needs met during their care. Elements mentioned in this category included access to food, ability to wear comfortable garments, sleeping arrangements, hygienic bathroom spaces, and meaningful use of free time. Windows and natural light supported participants' wellbeing during their crisis care stay. One participant said, “If you weren't depressed before you went in there [ED], you're gonna get depressed while you're in there, so I think they need to put in windows.” Another said, “the sunshine helps.” A pantry space in the emergency room area with snack options was reported as a comforting freedom. Meal time restrictions were not favored among the group; in fact, a participant said snack options outside of the routine meal time were important when, “I wake up in the middle of the night ‘cause I've got the munchies because of the medicine I take.” Participants discussed access to clean bathroom spaces as a highly-valued feature, preferring bathrooms that offered privacy and ability to bathe. One participant added that, “I came in with nothing and I wanted to take a shower so bad. They [WakeBrook] gave me brand new under clothes.” Wearing clean clothes or ‘scrubs’ was favored as preserving some physical comfort and modesty as compared to open-back hospital gowns. Other participants echoed the importance of comfortable clothing, saying, “In the emergency room, I'd be given like a paper gown to wear. At least over here [WakeBrook], I'm wearing clothes.” Participants understood that facilities lack a sufficient number of beds in the crisis care environment and sometimes require temporary sleeping arrangements that included chairs, reclining chairs, and roll-out mattresses. One participant said, “But I've seen people [at WakeBrook] that slept …with no pillow. They give you a little blanket.” Participants expressed great appreciation for times when they were able to sleep in beds or at minimum provided with a blanket and pillow during their stay. Participants also mentioned that access to television, reading material, and fresh air were desired to help pass time.

Freedoms

Participants discussed both lack of and expression of freedom: involuntary commitment, restraint, and recreation. Participants discussed their experiences and the appropriateness of being involuntarily committed (IVCed) during psychiatric crisis care at an ED and WakeBrook. Generally, participants recounted negative experiences with placement on involuntary commitment. However, several participants believed that at the time of their crisis, being on involuntary commitment was appropriate for them. A participant expressed: “every time I've been IVCed, I've needed to be IVCed.” A patient who was involuntarily committed after voluntarily coming to an ED believed it was an appropriate course of action since he “felt like I still needed to be under lock and key.”

Participants also explained why they thought they were on involuntary commitment in the ED. A patient was told by a nurse that she could get a bed quicker if she were on involuntary commitment. One participant recounted his experience of being on involuntary commitment after coming to the ED for a physical medical complaint. “Prior to going there for the [other] issue, I went there for psychiatric treatment, and since it was on my record, they IVCed me just for being there.” This participant believed he was treated differently because of his psychiatric diagnosis, and as a result experienced a longer length of stay.

Through these conversations, common themes emerged regarding what participants value if they have to be involuntarily committed. They valued learning the rationale for involuntary commitment and being informed of all medical decisions being made on their behalf by providers:

“I came in [to WakeBrook] voluntarily with someone and they didn't tell my sister…So, I was wondering how can they keep you here if you haven't seen the doctor or anything, because normally the doctors can tell if you need to be here or not, but nobody here talked to me. And it was kind of upsetting because all they said was ‘you're just going to stay here,’ and I wasn't counting on that.”

Participant discussion on restraints covered their experiences while they were restrained, and the appropriateness of restraints at the time of their crisis. Most participants associated negative experiences with being restrained. One participant came into care at an ED feeling distressed, and being restrained only exacerbated his discomfort. He “started freaking out worse than what it was before they did this.” Participants said physical restraints did little to calm them down, but being able to talk to someone gave them some comfort. One participant mentioned that she “didn't get restrained [at the ED] so that was good. All they did was talk to me and that helped a little bit.”

Participants valued having free time for recreational activities outside, “[at WakeBrook] I still had my freedom too because they took us outside to play basketball or whatever we wanted to do outside.”

Communication

Participants valued clarity and timeliness of communication. They described how they appreciated explanation for how the care process would work, whom they would see and when, and details about the admission criteria and treatment options.

“With me, I have been on so many medications that, when [ED] physicians speak medications to me, I want to know what they're talking about, ‘We are going to try you on this.’ ‘You're gonna try me on that? What does that do, what are the reactions to that?’ And they don't have any problems with sharing with me what the side effects are. And then if I say I don't want that because of those side effects, then in my own experience, they've been like ‘Okay, well would you mind trying this?’ So, in my own experience, medication has been explained to me. That's because I ask.”

“The doctors [at Wakebrook] listened…they tested out different drugs on me and they came up with a good idea.”

Importantly, they acknowledged the value of conversations even when they were difficult. “[A WakeBrook physician] still put a little fear in my heart, but he kept talking to me and everything, which he was telling me right and everything.” Immediate support in the face of distress was appreciated, “I was crying and the [WakeBrook] technician came in and saw me crying and she knew me, so she immediately got somebody.” In contrast, they expressed dissatisfaction when treatment options were raised without adequate detail or not discussed at all and in the face of long waiting times.

Basic respect was a common theme of participants. They described how extending common courtesies helped both patient and staff come to a joint understanding which facilitated the care process. “I feel like [ED staff] listened to me. I really do. Because it got me into a different place where I belonged. It helped me move to the next step where the next person understood, and then from there the next person understood.” “I have a mental health advance directive. And it's been notarized and everything. And I find that over at [the ED], they don't accept it as readily as they do here [WakeBrook].” Counter examples reflect a lack of understanding of psychiatric crisis by ED staff and likely also limited resources: “And the nurse said, ‘Well you know, I took an oath to save people and help people. And I can't in good conscience help you if you want to hurt yourself.’ And she said, ‘I can't work with you.” “I said, ‘Please give me something for the voices.’ She said, ‘Well I don't hear the voices, so there's nothing I can do for you.’” “A nurse aide …kept saying to me…’You know you're just doing this on purpose to get attention.’” “[Sitters] have conversations in their language. I don't understand and I think that's rude. They're talking about you and you know it.” Participants generally felt that mental health needs were not taken as seriously as physical needs at the ED. Participant statements indicated they appreciated the kindness shown to them in both settings. They said over and over how having company, someone to talk to, was very comforting. “I've had [ED] nurses that kind of pat me, saying you know, hey it's going to be alright. Sit and talk to me.” “[WakeBrook staff] used to sit up with me at night and talk to me and be very encouraging.” In contrast to the common narrative regarding police officers, participants noted these individuals as influential to their valued care experience.

“The lieutenant that works at [ED]…came and got me and took me into this other room. He was like, ‘How about you sit in here until they can get you a room?’…And then he came and sat in there with me, he said, ‘I don't want you by yourself while you're going through your emotions.’ He said, ‘I won't say anything unless you say something to me.’ So, I said okay that's fine. And he sat there the whole time with me and even after they actually put me into a room, he was right out there. He went and grabbed a chair and sat out there.”

Participants' assessment of the appropriateness of care was nuanced and insightful. They were clear that shared decision-making and self-advocacy, i.e. working with providers to identify the treatment goal, consider treatment options and make a choice (as described above regarding medication choice), were not always possible when someone was in the midst of psychiatric crisis (Mental Health America 2016). In situations like these, participants could still advocate for themselves by planning ahead. They said their medication records were on file, “They have my medication on file. So the first thing they do is come ask me… looking on the computer, and say you're still taking so and so and so. And how you're feeling” and sometimes they had a psychiatric advance directive. “[My psychiatric advance directive] says where I want to go if I'm in a crisis, and places that I don't want to go, and medications that have worked on me, and gave me bad effects.”

They felt that there were clearly times when involuntary commitment was appropriate. There were also times when they wanted to be admitted but were not. Participants returned repeatedly to the notion that, when possible, clear communication was ideal.

Length of stay and visitation during crisis care

Participants' experience with length of stay in crisis care varied widely. While most participants stayed in care one to three days, others noted extended stays of up to three weeks. Discussion primarily centered on issues of involuntary commitment, medication management, and bed availability. As characterized by one participant, the most valued experience came as a result of improved medication management in order to find “the correct medication that works with me and for me.” Participants who came to an ED complaining of physical comorbidities cited longer stays in ED facilities. Some participants noted that their shortened length of stay at WakeBrook was influenced by their frequent use of crisis services. Another participant was accepting of her length of stay at WakeBrook because, “It was a choice for me because I wanted to make sure they executed a good follow-up plan for me.” Participants' awareness of delays within the crisis care system was due to both frequent use of services and an understanding of system-level limitations that influence crisis care. Taken together, most participants valued shared decision-making, clear communication, and autonomy, particularly related to medication management and follow-up care, as supporting quality length of stay experiences.

Although many participants did not have close involvement with family during crisis care, those who prioritized this relationship cited it as a key contributor to the value of their experience. Participants described family and friends as fundamental players in the de-escalation process once entering crisis care. As one participant noted regarding the ED, “I started getting hysterical and stuff. And then [my sister] says, ‘I think she would feel more comfortable if, if I was with her,’ meaning me. So, I felt a little bit better.” Others described the role of family and friends as a comfort and sometimes used these individuals to facilitate the care process by encouraging disclosure of symptoms with crisis care staff. Valued features of this experience included pre-defined visiting hours for family and friends as well as being able to “weed out” potential visitors. In some cases, participants were denied access to their family due to concerns over risk of harm. Outside of visitation by family and friends, participants also appreciated company from crisis care staff and individuals designated to stay with them as they awaited care. These individuals ranged from employee staff to police officers, and were uniquely influential in providing elements of valued care during participants' care experiences, such as food, supportive communication, and comfort. As one participant noted, “Yeah. I'm sitting [in the ED] crying, like I'm very emotional, because I was depressed. So, I was sitting there and [the lieutenant] said, ‘I'm going to get you some food.” So, he went and got me some food, got me a drink, got me some snacks, got me a blanket. So, he made me more comfortable…”

Follow-up care

Participants valued having information about potential referral sites, so they could make a judgment about where to go, to “…know something about a facility…told me about it and see if I liked it.” They also appreciated having paperwork completed to facilitate the referral. “And [the ED] set out the paperwork, and when I got [to WakeBrook] all I had to do was hand them the papers that the doctor wrote up for me. And that was it, they took me in. So, he did do a good job referring me to this program.” They wished that the discharging site would take note of the referral success. As a counter example, one participant said, “In fact, I had episodes where … my conditions became worse. And, instead of looking at what other options were there, they just kept sending me, discharging me back to [Place] and telling me to follow up with my regular therapist.” Participants' descriptions of the quality of referral ran the full gamut from no plan to always receiving an aftercare plan. Some said they were simply told to go to a particular referral site, but no calls were made on their behalf. Others were transferred to long term facilities. There was no clear distinction in opinion about referral quality between the EDs and Wakebrook. However, one person said, “Facility based crisis: The staff [at WakeBrook] is just top-notch to me. I mean I keep in touch with them. I come back here to primary care.”

Valued attributes of crisis care

Table 2 describes attributes of crisis care that were valued across features among focus group participants: respect, comfort, and shared decision-making. Participants appreciated when respect was shown to them, such as basic courtesy, attention to basic needs like food and clothing, attention to individual preferences, clear and timely communication, and kindness shown during a difficult time. Respect was an important attribute of each of the seven features of psychiatric crisis care that they described. Basic comforts were described as important for well-being, including hygienic facilities, clean clothing, access to healthy food, and natural sunlight. Comfort was an important attribute of two of the seven features of psychiatric crisis care regarding physical space and its experience: accommodations and basic needs and infrastructure. Shared decision-making was emphasized as a foundation of quality care. Participants were well aware that they might not be able, or as able as usual, to participate in shared decision-making during times of crisis. They described valued strategies to address this tension, including transparent communication, treatment planning both before and during crisis care, and attention to their preferences regarding visitors who might or might not aid in decision-making. Shared decision-making was an important attribute of the five remaining features of psychiatric crisis care regarding interactions and processes: freedoms, communication, length of stay, visitation and follow-up care.

Table 2. Valued attributes of crisis care.

Attribute	Examples
Respect	Basic needs Clear, timely, open communication even about difficult topics Kindness, courtesy Attention to preferences
Comfort	Food, bedding, clothing, natural light, outside time, cleanliness
Shared decision-making	Transparent communication Treatment planning Visitors both allowed, and able to screen

Setting preference

Although participants saw pros and cons in both ED and WakeBrook settings, participants preferred WakeBrook. They emphasized caring, knowledgeable staff and referral to WakeBrook inpatient, residential, or detoxification care as the most important factors driving their preferences.

Discussion

Focus group participants were compelling and articulate in highlighting respect, basic comforts and shared decision-making as valued attributes of psychiatric crisis care. These findings are consistent with emerging literature on patient preferences in psychiatric crisis. For example, people with psychiatric needs and co-occurring conditions are more likely to go to the ED than others, (Sajatovic et al. 2016; McPherson et al 2015b; Nolan et al. 2015; Hackman et al. 2006). Psychiatric ED patients have identified the importance of listening, sensitivity, and respect as potential strategies to improve the care experience (Griswold et al. 2008), and women with acute stress reaction who had experienced domestic violence have expressed the need for a safe quiet setting for care (Olive 2016). There has been some attention devoted to the architectural environment of the ED and its impacts on psychiatric patients (Broadbent et al. 2015). A mental health liaison nurse and ED care plans have been valued by patients as a means to support better communication between patients and emergency staff (Wand et al. 2016; Dunnion and Griffin 2010; Griswold et al. 2008). The literature also articulates the importance of nursing to build relationships with both psychiatric patients and caregivers and parents (Montreuil et al. 2015). And, there is evidence that having the option for shared decision-making in the form of self-referral to follow-up care was perceived as enhancing quality of care, improving confidence in treatment choice, and ability to cope (Oslo et al. 2016). Our participants' insights into 1) their typical functioning and patterns of service use, 2) their efforts to prepare for times of crisis, and 3) their temporary impairment during crisis make important contributions to a literature that frequently focuses on their disability (Oslo et al. 2016; Kessler et al. 2001). These findings underscore the importance of psychiatric advance directives as well (Swanson et al. 2006). Future work should address strategies to improve respectful communication protocols, and shared decision-making for individuals receiving psychiatric crisis care in both emergency department and dedicated psychiatric crisis care facilities.

Focus group participants provided important insights into what a therapeutic landscape looks like, detailing the elements that foster well-being. Themes of basic comfort are straightforward if not always easy to fund. Natural resources and life-style time management strategies may be the resources that are most affordable for the variety of psychiatric crisis provider organizations. In community populations, exercise, outdoor activity, and sunlight exposure are all associated with reduced depression and anxiety scores (An et al 2016; Xu et al. 2016; Beute and de Kort 2014; Hahn et al. 2011). In individuals with mental health conditions, there is also evidence that diet, exercise, and sun exposure are associated with improved mental and physical health (Garcia-Toro et al. 2014). Tellingly, a survey of youth and young adults in the general population highlighted physical activity and sunlight as the two most frequently used strategies to deal with mental disorders and to support mental health and well-being (Reavly et al. 2011). This evidence supports the preferences for exercise and sunlight expressed by our participants. Psychiatric crisis care facilities should work to ensure that patients enjoy exercise and sunlight to the extent that they are physically able. Other infrastructure and protocol changes may require more long-term planning and funding. For example, if people need to be in restrictive environments, allowing in more natural light may require space or building modifications. Overall, there is evidence that reducing toxic stress by addressing basic needs, such as natural light, quality food, clean and comfortable residences, and access to outdoor recreation can support mental health and well-being (Shern et al. 2016). Future work should examine short-term impacts of these basic features on escalation of behavioral problems, patient and family or caregiver satisfaction, and psychiatric health outcomes for people using psychiatric crisis care. This evidence is particularly critical in psychiatric crisis care because this is the time and place where assessment and referral happen, with long-term implications for patients. The factors that make up a salubrious setting that supports therapeutic alliance should be the cornerstone of patient-centered care and recovery services.

Some issues raised by participants reflect larger system challenges. The long wait times that participants reported and encounters with staff who lacked skills to deal with psychiatric crisis are consistent with the evidence on mental health professional shortage, psychiatric bed shortage and unmet need for community-based mental health services (Thomas et al. 2009; Thomas et al. 2012; La et al. 2016; Luthra 2016; Swartz 2016). North Carolina ranks 27^th among the states in prevalence of mental illness and access to care (Nguyen et al. 2017) with widespread shortage of mental health professionals (Thomas et al. 2009). There is evidence that ED length of stay is longer for patients transferred to publicly funded psychiatric facilities compared to those transferred to private facilities (Misek et al. 2015). In fact, most recently, psychiatric boarding times have been demonstrated to be a useful metric to assess need for psychiatric beds (Swartz 2016). For example, simulations indicate that to reduce boarding times below 24 hours, North Carolina would have to add 356 state psychiatric hospital beds, an increase of 165% (La et al. 2016). In this constrained context, identifying and making appropriate referrals through shared decision-making was highly valued by participants. Focus group participants particularly valued WakeBrook crisis staff's ability to refer internally to WakeBrook's short-term inpatient services, so that there was continuity of care. Future work should examine the impact of such tight referrals on satisfaction in crisis services as well as on long-term outcomes.

Shared decision-making was a valued attribute of care that emerged throughout participants' discussion. Participants' value of shared decision-making is consistent with Institute of Medicine's recommendations regarding integrating shared decision-making into psychiatric care which has been championed by Mental Health America (2016; Alston et al. 2014; Institute of Medicine 2006). These findings reflect current evidence that psychiatric patients want more information from providers and experienced peers and to have input in treatment choices (Velligan et al. 2016; Park et al. 2014; Patel and Bakken 2010). The fact that they often feel excluded from decision-making and may lack the skills or power to handle disagreement suggests the need for interventions to teach activation skills (Dahlqvist Jonsson and Schön 2015; Woltmann and Whitley 2010). When shared decision-making is achieved, it is associated with improved treatment engagement (Dixon et al. 2016). Psychiatric advance directives, in particular, are associated with therapeutic alliance, patient satisfaction, and sense of autonomy over time (Elbogen et al. 2007; Swanson et al. 2006). Future work should develop strategies to foster and monitor shared decision-making during treatment in emergency department and dedicated psychiatric crisis care facilities. Importantly, future work should also examine if shared decision-making during crisis services can improve the quality of care over time, for example, by increasing reliance on community-based services and patient satisfaction with care patterns.

These findings should be considered within the context of several limitations. First, while some participants had received care in both EDs and at WakeBrook, others had received care in only one of these settings. This latter group could contribute to the discussion about one setting and participate in articulating preferences and reasons for them. They could not, however, compare personal experiences in both settings. Future work should compare care settings through experimental designs to assess outcomes adjusting for the complexity of patient needs. Second, participants were recruited from patients known to clinical and social work staff. First time or infrequent users of crisis care may have different experiences, perhaps more difficult, if they are unfamiliar with the setting, perhaps less difficult if they are not known as frequent users of services. Future work would benefit from random sample designs. Third, we relied on participant recall to describe and assess their care experiences. Future work would benefit from longitudinal cohort designs that rely on shorter periods of recall. Fourth, while our qualitative study design generates new themes important to typical users of psychiatric crisis care in our county, our sample was not large enough to compare perspectives by participant characteristics. Future work based on larger samples suitable for quantitative analysis should explore how patient characteristics, such as residence and travel times, socio-demographics and medical conditions are associated with care experiences.

Conclusions

Focus group participants preferred receiving crisis services at WakeBrook because it was a familiar setting where they were known by staff, had established therapeutic alliance, and where their medication preferences were known, recorded, and heeded. WakeBrook is a new facility located in two spacious buildings that provide ample natural light and views of surrounding woods. It is also on a city bus line, making it accessible for people with limited means of transportation. Altogether, WakeBrook embodies each of the valued attributes of psychiatric crisis care: respect, comfort, and shared decision-making. Investment in dedicated psychiatric care facilities such as WakeBrook that provide these valued attributes of care has the potential to increase user satisfaction at a vulnerable point and shift such care out of EDs. Future work should evaluate the impacts of dedicated psychiatric care facilities. There is a critical need for research addressing long-term outcomes to motivate these changes.