Behavioral emergencies are a common and serious problem for consumers, their families and communities, and the healthcare providers on whom they rely for help. In recent years, serious concerns about the management of behavioral and psychiatric emergencies—in particular, the misapplication and overuse of restraint and seclusion—have become a focus of attention for mental health professionals and policy makers as well as for the lay public, the media, and patient advocacy organizations. 1–7 Policy leaders and clinicians are searching for ways to balance the rights of consumers with considerations of safety and good care in an area in which it is difficult to conduct research. As a consequence, regulations concerning psychiatric emergency care have been promulgated by the Health Care Financing Administration (HCFA), now the Centers for Medicare and Medicaid Services (CMS), 8 and the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) 9 to try to address these issues.
The Expert Consensus Guidelines on the Treatment of Behavioral Emergencies, published in May 2001, 10 were based on a statistical analysis of the opinions of 50 experts on emergency psychiatry and psychopharmacology. The professional survey was undertaken to fill a gap in the research literature concerning best practices for the treatment of behavioral and psychiatric emergencies. The group of professionals who were surveyed for the Expert Consensus Guidelines is referred to here as the professional panel in order to distinguish them from the consumer panel, who are also qualified as experts based on their experiences. [Editor’s note: see the article by Allen et al. on p. 16 for a summary of the results of the professional survey.]
In fact, an important finding from the survey of mental health professionals was the emphasis placed on consumer involvement in treatment decisions. When asked which factors it was most important to consider in selecting an acute intervention in a behavioral emergency, the experts in emergency psychiatry considered collaboration between patient and clinician whenever possible extremely important in achieving the best short-term and particularly the best long-term outcomes. The editors of the guidelines therefore recommended undertaking a survey of a representative “expert panel” of consumers who had personal experience with restraint, seclusion, or emergency medications. The expert panel noted that, given the relatively modest differences in efficacy between available medications, consumer preferences, which may be more likely to reflect differences in side effects, could play a deciding role in determining the best general approach to a behavioral emergency during the usually brief period when patients are unable to participate in decisions and physicians are serving as proxy decision-makers. Individual advance directives or credible consumer surveys are two methods by which consumer preferences might be ascertained. The present survey of consumers concerning their emergency services experiences and recommendations was undertaken in response to the need for such input from consumers.
In this article, we describe four emergency services forums conducted in 2002, each of which involved a written consumer survey and a workshop to develop and prioritize recommendations for improving psychiatric emergency care. In the first section of the paper, we discuss the methodology used in the survey and workshops. Next, we present the results of the consumer survey and summarize the top recommendations from the consumer workshops. We then consider the implications of these findings for improving psychiatric emergency care. Finally, we discuss the limitations of the current study and make recommendations for future research in this area.
Emergency services forums were held between May and August 2002 in four locations: Poughkeepsie, New York, Syracuse, New York, Marshalltown, Iowa, and Los Angeles, California. The purpose of the forums was to obtain the recommendations of consumers for improving the provision of psychiatric and behavioral emergency services and to use these recommendations to develop educational materials that could be used by a wide range of stakeholders in mental healthcare delivery.
The forums were planned and implemented by representatives of Comprehensive NeuroScience Inc., (CNS) a medical research organization, by John L. Sheets, MSW, MPH, a consultant in psychiatric rehabilitation services design from Syracuse, NY, and by representatives of consumer advocacy and support organizations. All the forums were moderated jointly by John Sheets and by Steve Miccio, Director of People, Inc., a mental health advocacy and service agency in Poughkeepsie, NY.
Selection of Participants
By necessity, the panel recruited for this project was a convenience sample. A key element in the willingness of consumers to participate and contribute to the forums was the involvement of consumer advocacy and support organizations.
Based on personal accounts given at a public hearing on emergency services held in New York State and on advice given by advance directive educators working for the New York Association of Psychiatric Rehabilitation Services (NYAPRS), the authors perceived a significant mistrust of formal representatives of hospital emergency services. The authors therefore decided that the most successful recruitment method would be to have participants be invited by persons they knew and trusted to have their best interests and the best interests of consumers in general at heart. Consequently, it was decided to organize all forums in partnership with locally known and respected consumer advocacy organizations and other trusted persons, who would be asked to extend the invitations to participate. To test the concept, one of the authors (SM), who is director of a consumer advocacy and services agency in Poughkeepsie NY, hosted the first forum and invited consumers to participate.
Partnerships were subsequently made with two advocacy and support organizations and a case management service in Syracuse, with the consumer advocate for the managed care organization for mental health services in the state of Iowa, and with a patients’ rights advocate with the Department of Mental Health for the city of Los Angeles. The organizers of the forums contacted the partners by phone and, when possible, in person, and provided them with background information on the study and with a model forum announcement describing the purpose and process of the forum that could be used in recruiting efforts.
Persons who had had at least one emergency services experience involving medication, seclusion, or restraint were invited to contact the forum partners to express interest in participating. The forum partners in each area were then asked to select 12–15 participants who met the requirements for study participation and whom they thought could tolerate the day-long experience and would be willing to address their emergency services experiences both verbally and in writing.
In each forum, when participants were asked “why they chose to participate,” the major reason given for accepting the invitation was that the participants appreciated that someone was willing to listen to and validate their experiences and that they wanted an opportunity to make emergency services better for others.
Each 1-day emergency services forum involved the following activities:
- An overview of the project presented by the moderators, followed by an opportunity for participants to share their reasons for accepting the invitation to participate in the forum
- Review and signing of an institutional review board (IRB) approved informed consent form for participation in the study
- A written survey in which participants were asked to recall their experience with emergency services and make recommendations for improvement
- A group discussion of the impact of the survey guided by the question “What did filling out the survey mean to you?”
- A workshop exercise to generate recommendations for improving emergency services care.
In recent years, expert consensus has played an important role in the development of practice guidelines alongside the “gold standard” of meta-analysis of clinical trials and other experimental data. This is because the sheer number of possible combinations and sequences of treatments for many diseases makes it difficult to provide comparative recommendations based entirely on clinical trial data. 11,12 An expert consensus method has been developed for sampling and analyzing expert opinion in a reliable, quantitative manner to help fill gaps in the evidence-based literature. 13 This method has been used to develop practice guidelines for a number of psychiatric disorders 14–23 and to query consumer advocates regarding policy recommendations for the treatment of schizophrenia. 18 This article describes the first use of this method to sample the opinions of consumers concerning their experiences with and preferences for emergency services.
The 43-question survey used in this study was created with the consultation of staff who have been both consumers and providers of mental health services and who serve as Advance Directive Trainers for NYAPRS. The survey asked respondents to provide information about the care they received when they came in or were brought in to an emergency room or crisis unit because they were having a crisis or psychiatric emergency. Consumers were asked to answer questions about what was helpful and what was not helpful in responding to their particular crisis situation and to give recommendations for how psychiatric emergency care could be improved. The organization of the survey is presented in Table 1.
The Rating Scale for the Survey
For many of the 43 questions in the survey, respondents were asked to rate options on a 9-point scale adapted from a format developed by the RAND Corporation. 24 Other questions asked respondents to select an answer and check a box, answer yes or no, or to write in an answer. A number of anchoring systems were used for the 9-point scale. These included level of agreement (1 or 2 = strongly disagree, 3 or 4 = disagree, 5 = neither agree nor disagree, 6 or 7 = agree and 8 or 9 = strongly agree), importance (1–3 = not important, 4–6 = somewhat important, and 7–9 = most important), preference (with 1–3 = least preferred, 4–6 no strong feeling, and 7–9 = most preferred), and helpfulness (with 1–3 = not helpful, 4–6 = somewhat helpful, and 7–9 = very helpful). Figure 1 presents examples of questions from the survey.
Data Analysis for Options Scored on the 9-Point Rating Scale
For each option rated on the 9-point scale that asked about importance, preference, or helpfulness, we first defined the presence or absence of consensus as a distribution unlikely to occur by chance by performing a chi-square “goodness-of-fit” test 25 (p < 0.05) of the distribution of scores across the 3 ranges of the rating scale (1–3, 4–6, 7–9). Next we calculated the mean and 95% confidence interval (CI). The rating for each option was designated based on the lowest category in which the CI fell. If the lower boundary of the CI was at 6.5 or greater it was considered to fall in the top range (e.g., most important, most helpful, most preferred); if the lower boundary was at 3.5 or greater the rating was considered to be in the middle range (e.g., somewhat important, somewhat helpful, no strong preference); if the lower boundary of the CI fell below 3.5, the rating was considered to be in the lowest range (e.g., not helpful, not important, least preferred). Within the top range, we designated an item as “option of choice” if at least 50% of the respondents rated it as 9. For items that asked about level of agreement, we looked at the mean for the panel, with means of 3 or lower considered to indicate strong disagreement, means between 3 and 4 to indicate disagreement, means between 4 and 6 to indicate uncertainty, means between 6 and 7 to indicate agreement, and means above 7 to indicate strong agreement. In this way, we were able to determine the consensus of the panel on different options (e.g., the level of their agreement with different statements, their most and least preferred types of treatment, the importance they attached to various kinds of staff behavior). For some options, there was no consensus as determined by chi square test, indicating that the pattern of responses was not significantly different from a random distribution across the ranges.
Note that the full consumer panel did not respond to every question. For example, respondents who had not experienced seclusion or restraint were instructed to leave those questions blank. Respondents were also told that they could leave questions blank if they did not feel comfortable answering them. For this reason, the n and percentage of respondents varies from one question to another, as is indicated in the presentation of the results.
Development of Workshop Recommendations
At each of the four forums, the survey served as a lead-in to a workshop in which the participants developed and prioritized recommendations for improving psychiatric emergency care in five phases of the emergency services process. Each participant was asked to write and post his or her recommendations for improving emergency services. All the recommendations were then reviewed by the moderators and, with input from participants, duplications were eliminated and similar items were clustered into more comprehensive recommendations. After a period of individual and small group review of all recommendations, each participant was asked to rate (vote for) the top three recommendations in each phase. The votes were counted and the top-rated recommendations for each phase were then organized into that forum’s rank-ordered recommendations for improving emergency services care. Items receiving the same number of votes were included as tied recommendations.
Fifty-nine individuals participated in the four forums: 14 in Poughkeepsie, New York, 12 in Syracuse, New York, 18 in Marshalltown, Iowa, and 15 in Los Angeles, California. Of those 59 participants, 57 completed written surveys (2 individuals in the Marshalltown forum did not complete surveys). All participants had had at least one and most had had multiple experiences with an emergency room or inpatient emergency service, often in multiple cities and states.
Table 2 presents information on the demographics and emergency services experiences of the participants. The average age of the participants in all the forums was early 40s. Overall, a higher percentage of women (60%) than men (40%) participated in the forums, although the numbers of men and women were equal in the forum in Iowa. The majority of the participants were Caucasian in three of the forums; however, the California forum was much more ethnically diverse, with 27% African-Americans, 33% Hispanics, and 20% Asian-Americans. Only two participants had a native language other than English (Spanish). The educational level of the participants was relatively high. The participants had experienced emergency services in a total of 17 states (New, York, Vermont, Pennsylvania, Ohio, Maryland, Georgia, Florida, Iowa, Illinois, South Dakota, Minnesota, Missouri, Colorado, Arizona, Nevada, Oregon, and California) and three different countries: the United States, the Philippines, and Mexico. The most common diagnoses the participants had been given at some point in their lives were major depression (70%), bipolar disorder (51%), posttraumatic stress disorder (46%), schizophrenia (40%), substance use disorder (30%), and personality disorder (30%). Other diagnoses reported by the group included schizoaffective disorder, anxiety disorders (e.g., panic disorder, obsessive-compulsive disorder, generalized anxiety disorder), dissociative identity disorder, eating disorders, trichotillomania, and dysthymia.
Experiences with psychiatric emergency care.
The questions about the respondents’ emergency experiences were presented to them with the following instructions:
In answering the next set of questions, please think about a particular time that you came in or were brought in for emergency psychiatric care. It can be the last time you received this kind of care. Or you can tell us about a time that you remember particularly well. We realize that some of these questions may bring back painful memories, but other consumers have told us that it is important to ask about these kinds of experiences if we are going to be able to develop recommendations for improvement. If you are not comfortable answering any question, please feel free to leave it blank. We also realize you may not remember everything in detail and that’s OK—just answer the questions the best you can.
Of the 57 respondents, 21 (37%) focused on the their first experience with psychiatric emergency care, while 36 (63%) focused on a later experience. Fifty-three respondents said that the incident they were describing had occurred at least 1 year earlier, with the average being 8 years earlier (median = 6 years, high = 28 years, low = 1 year); four respondents said they were recalling an incident that had occurred more recently. Approximately half the panel said they came in for care involuntarily and half voluntarily; 13 respondents (23%) said they came in by themselves, 17 (30%) that they were brought in by a family member or friend, 14 (25%) that they were brought in by the police (in 5 of these cases, a family member and/or mental health professional also came in with them), 6 (11%) by a mobile crisis unit, 3 (5%) by ambulance, and 3 (5%) by a mental health professional (56 respondents total).
Of the 53 respondents who provided information about where they were cared for, 29 (55%) were treated in a general hospital emergency room, 22 (42%) in a psychiatric hospital emergency room, 14 (26%) in an inpatient psychiatric unit, and 5 (9%) in a crisis center. Of course there was sometimes an overlap in these reports, with 12 respondents receiving care in more than one setting (e.g., 6 patients were cared for either in a general or psychiatric emergency room and also on an inpatient unit).
The types of problems that caused the respondents to come in or be brought in for care are shown in Table 3. The most common were “feeling out of control” (68%), “relationship problems with family, friend, or spouse” (58%), and “being very afraid” (54%). Most of the respondents (88%) reported that more than one type of problem was responsible for the emergency, and 72% reported 3 or more different types of problems. Five respondents reported coming in because of problems both with alcohol and street drugs. Whenever respondents reported coming in because of problems with alcohol and street drugs, they also reported other types of problems.
Experiences and relationship with staff.
Respondents were asked how strongly they agreed or disagreed with a number of statements about how they were treated by staff. As a whole, the consumer panel reported poor satisfaction with the treatment they received. Looking at all the respondents who either strongly disagreed or disagreed with the statements (ratings of 1–4), we found that the majority of the respondents did not feel that the staff had treated them with respect (63%), seen them in a timely manner (65%), listened to them, their story, and their version of events (68%), spent enough time with them (77%), adequately addressed their problems (80%), or described the nature of the proposed treatment, its risks, benefits, and alternative options before asking them to consent to treatment (82%). The mean score for these items ranged from 2.61–3.61 (scores of 3 or 4 = disagree). There was no consensus about whether the staff was equipped to deal with specific needs and issues related to the respondents’ ethnic, cultural, racial, or religious backgrounds, although 53% of the 51 respondents to this item felt they were not.
When asked if they had experienced any particular act of kindness or caring from staff that supported them and helped them through the crisis and, if so, to describe it, 32 respondents said they had experienced such an act, while 22 said they had not (3 respondents did not answer this question). Four important themes for improving emergency services emerged from the respondents’ answers to this question. These themes, which are listed below with representative examples of the write-in responses on which they are based, could be used to help guide the creation of practice guidelines, operational policies, and pre-service and inservice training curricula.
Importance of being treated as a human being and allowed to retain one’s dignity
Quite a few respondents reported positive experiences in this area.
They were attentive to comfort needs, bringing food and drink, blankets and pillows. I was not made to disrobe and don a gown, but allowed to remain in street clothes.
During the ER visit, the kindest thing was that my psychiatrist happened to be at the hospital and met me. She stayed with me and made it much easier because she always respected me and always did whatever she could to maintain my dignity.
When I was in restraints and needed to go to the restroom, they let me out. This trust was important to me, as many people confined to restraints end up going to the bathroom in their pants.
Some respondents, however, had felt stripped of their dignity and autonomy.
I was immediately strapped down, given two injections, and my clothes were taken. I was given a hospital gown. This last hospitalization I did not see a doctor. I was admitted and all the paper work was done after the fact and all the incidents are conflicting. I am trying to find out what actually happened.
Because I tried to kill myself by overdosing with alcohol and medications, I was treated meanly and the staff was cold. The police were laughing and joking with each other about what a loser I was.
Importance of staff listening to what the person has to say, respecting his or her wishes as much as possible, and answering questions and providing information about what is going on
Again, quite a few respondents felt that certain staff members had done a good job in this area.
The intake staff took the time to talk to me and get my history and were supportive of my decision to stay in the hospital for help.
A charge nurse took time out of her busy schedule to answer some questions I had and alleviated some of my fears.
One head nurse took the time to really listen and made changes as necessary. She talked to other staff about my complaints and pursued the matter for several days until it was resolved. I would go back to this hospital just because of this one person.
Some respondents, however, felt ignored and unheard.
I felt that my treatment was bad because I felt nobody was listening to me.
They just seemed to ignore me. I was in a locked room and could see the staff through a window. They were not kind.
Importance of being asked what you need or want
Again, some respondents felt that staff members had done a good job in this area.
She checked on me to see how I was doing and if there was anything I needed.
The male nurse who first took my vital signs was very nice, talked to me, acted very compassionate, asked me if I wanted something to eat.
I asked for a vegetarian meal—it was 8 or 9 at night—and I got a hot cheese sandwich.
However, others felt that at least some members of the staff had ignored them and their needs.
In the ER, I was left alone in a locked fairly empty unit for many hours.
I felt totally ignored as staff never came out of their office. The only person that showed care, concern, and time was the housekeeper who was very supportive.
Nurses acted nice but when I was overly medicated, they did not help me. My bed stayed wet and I was drooling.
The techs help you more than the people who are more qualified. I got the sense that the techs care more than the doctors and nurses.
The male nurse was the only one I saw who cared at all.
Importance of being soothed and helped to calm down and not be afraid
The nurse practitioner was very kind, soft-spoken, and receptive to my mood and feelings of being overwhelmed. She soothed my senses.
An ER nurse was there with me the whole time I was in the ER.
They were kind to me and they tried to calm me down. I was scared and they tried to get me to relax.
One nurse stayed full time and engaged me in conversation that had nothing to do with my crisis. This was a good “time-out” for me and helped me relax.
The importance of staff having an positive outlook and reassuring the person that he or she can get better
I had tried to kill myself. The staff was very helpful at that time, they gave me hope to keep on living.
Although most of the staff was detached, there was one overnight nurse who took the time to talk to me and reassure me that everything would be okay.
Experience with medications.
The respondents were asked if they had received medication during the crisis and, if so, what types of medications they had received. Their responses are summarized in Table 4.
Respondents were asked how strongly they agreed or disagreed with a number of statements about their willingness to take medication during the crisis and whether it had been helpful. The panel tended to be divided on these questions, as shown in Table 5. For example, 50% of the respondents said they had not been asked about medications that had been helpful in the past, whereas 44% reported that they had; 35% felt that medication had been forced on them against their will, whereas 58% did not; 37% said they had not wanted medication and had not taken it willingly, whereas 46% said they had; 36% did not feel that the medication had helped them, 42% felt that it had, and 22% were in between (neither agree nor disagree). However, a majority (61%) of the respondents did report that the medication they took during the crisis caused troublesome side effects.
Experience with restraint.
Thirty-six (63%) of those who were surveyed had been placed in restraints at some time during a psychiatric crisis. Most of the episodes of restraint the respondents were describing had occurred at least a year earlier and on average 8.8 years earlier (median = 7 years, high = 33 years, low = 2 months). Two respondents reported more recent episodes of restraint occurring 6 and 2 months earlier. Three-quarters of the respondents reported that they had seen other people in restraints while they were receiving care for a psychiatric emergency, an experience nearly all the respondents reported finding upsetting (42%) or terrifying (51%). When asked how well they remembered what had happened while in restraints, 11 respondents (31% of the 35 individuals responding to this question) said they barely remembered anything, that it was a blur; 7 (20%) said they remembered things pretty well but some things were a bit hazy, and 17 (49%) said they remembered it vividly and in detail.
Twenty-four of the 36 individuals who had experienced restraint (67%) reported that staff had not tried anything else before putting them in restraints. The respondents who said staff had first tried other interventions mentioned strategies such as talking to them, leaving them alone, giving them an injection, or putting them in locked seclusion.
Respondents reported that they were kept in restraints for an average of 6.5 hours (median = 3 hours, high = 24 hours, low = 1 hour). (The mean duration of restraint reported in a survey of approximately 50 psychiatric emergency services was 3.3 ± 2.9 hrs. 26) Twenty-one (62% of the 34 respondents answering this question) said they had been given medications either just before or while they were in restraints.
Respondents were asked how strongly they agreed or disagreed with a number of statements about their experiences while in restraints. A majority of those responding to these items felt that staff had not made a responsible decision in putting them in restraints (69%), that they had been kept in restraints too long (68%), that no one had listened to them or responded to their requests (77%). Approximately half of the respondents said that they were not checked on very often while in restraints. Finally, 54% said that being in restraints had made them unwilling to seek out psychiatric care after this experience.
Respondents were also asked to list things that helped them and things that made it harder for them while they were in restraints.
Things that helped while in restraints.
Overall, respondents stressed the importance of having someone (a staff person, a family member, a friend, or a peer advocate) stay with them or check on them frequently, talk to them, and explain what was happening and why they were being restrained. For example, they mentioned the importance of “having a family member there, someone to explain why I was in restraints, someone to offer alternatives,” “of talking out my fears and frustrations with staff who would really listen to me,” and of being talked to “as a person, not a victim.”
Things that made being in restraints harder.
The respondents expressed concern about the physical discomfort associated with being in restraints (e.g., being tied too tightly, muscle cramps, cold, worry about vomiting and choking) as well as frustration at not being released to go to the bathroom. One respondent recalled, “The staff was intolerant to my frequent urination problem and made me wait so long for a bed pan that I had to urinate on the table; then they became angry with me and left me wet.” Some respondents also reported dissatisfaction with the physical setting (e.g., darkness, plastic mattress, lack of stimulation of any kind). Members of the panel reported that staff either did not check on them enough or displayed a negative and unsympathetic attitude towards them (e.g., “the staff looked at me as if I was a criminal.”
Experience with seclusion.
Thirty-nine (68%) of those surveyed reported that they had been placed in seclusion at some time during a psychiatric crisis. This had occurred an average of 6.4 years earlier (median = 3 years, high = 24 years, low = 7 months).
When asked how well they remembered what had happened while in seclusion, 9 respondents (23%) said they barely remembered anything, that it was a blur; 6 (15%) said they remembered things pretty well but some things were a bit hazy, and 24 (62%) said they remembered it vividly and in detail. Respondents reported that they were kept in seclusion for an average of 12.2 hours (median = 8 hours, high = 48 hours, low = 0.5 hour). Of the individuals who had been placed in seclusion, 26 (67%) reported that staff had not tried anything else before putting them in seclusion. The respondents who said staff had first tried other interventions mentioned strategies such as talking to them, warning them about what would happen if they didn’t calm down, putting them in the TV room, and giving them an injection of haloperidol. Twenty-two (56%) said they had been given medications either just before or while they were in seclusion. When asked if they had been given things they needed (e.g., water, medicine, food) if they stayed in seclusion for over an hour, only 15 (38%) said that they had.
Respondents were asked how strongly they agreed or disagreed with a number of statements about their experiences while in seclusion. A majority of those responding to these items felt that that staff had not made a responsible decision in putting them in seclusion (69%), they had been kept in seclusion too long (78%),and that they had not been checked on very often while in seclusion (55%). Finally, 60% said that being in seclusion had made them unwilling to seek out psychiatric care after this experience.
Respondents were asked to list things that helped them and things that made it harder while they were in seclusion. The concerns respondents expressed about their experiences in seclusion were similar to those they had about their experience with restraint. Again, they stressed the importance of having staff check on them, talk to them, ask how they were doing, and treat them with respect.
Things that helped while in seclusion.
However, certain incidents of kindness and being treated with respect stood out in the respondents’ memories. Six respondents reported that having staff check on them, talk to them, ask them what they needed, and treat them with respect had helped them get through being in seclusion. One respondent recalled “One time there was this female security guard who stayed with me until I was picked up and she hugged me just at the right time and talked to me with dignity and respect. I’ll never forget her.” Six respondents noted that the presence of some sort of external stimulus, such as music, a window, reading materials, television, was helpful.
Things that made being in seclusion harder.
Seventeen of the respondents (44% of those who had experienced seclusion) reported that one of the hardest things about being in seclusion was the isolation (“lack of human contact, being alone, no one to talk to, loneliness”), especially at a time when they were frightened and anxious. One respondent described “being totally alone, not having anyone to talk to when I was afraid and anxious.” Another said, “At that time in my life I could not stand to be alone. The psych doctor told them, but they didn’t listen. I totally freaked out.”
Six respondents had found it difficult that no one explained why they were being placed in seclusion or how long they could expect to be there. Just as with restraint, a number of respondents expressed dissatisfaction with the physical environment in which they were secluded (dark, cold, sterile, bare mattress). Several respondents also noted that staff had treated them with a lack of respect.
Abusive staff behavior while in restraints or seclusion.
We asked the consumer panel if they had ever experienced certain types of abusive treatment during an episode of restraint or seclusion. Of the 43 participants who responded to any items in this question, 8 (19%) did not report ever experiencing any abusive behaviors, while 6 (14%) said they felt that they had been placed in restraints or seclusion at some point as a punishment rather than to help them calm down but did not report any other abusive experiences. Twenty respondents (50% of the 40 respondents who answered this item) said that staff had refused to let them out of restraints or seclusion to go to the bathroom and 19 (46% of 41 respondents) said that a staff member had cursed at them or been verbally abusive. A total of 14 respondents reported experiencing one or more of the more severely abusive behaviors we asked about: being kicked, punched or slapped by a staff member, (12 [29%] of 42 respondents); having been injured while being placed in restraints or seclusion or while there (9 [22%] of 41 respondents); being placed in seclusion without any clothes (8 [20%] of 40); being sexually harassed or assaulted (5 [13%] of 39); and two reported all four behaviors.
Positive staff behavior while in restraints or seclusion.
We also asked the consumer panel to tell us about anything that staff or others did that stood out as particularly comforting and reassuring or that helped them feel better while they were in restraints or seclusion.
Four respondents replied negatively to this item—saying that nothing stood out as having been helpful and that they had felt that no one cared for them. Just as in the earlier questions about what was helpful and not helpful during restraint or seclusion, the majority of those who responded to this question stressed the importance of having staff talk to them and listen to them, treat them with respect, reassure them that they were safe, and see that their physical needs were met (e.g., for food, drink, being able to go to the bathroom). One respondent recalled the following behaviors as reassuring and comforting:“Coming into the room and talking, reassuring me I was not a bad person and not judging me. Making sure needs are met (asking frequently). Getting me the least restrictive care possible as soon as possible. Not using it as a punishment or a way of ‘discouraging negative symptoms,’ especially when it’s not something I can control in crisis.”
Recommendations for Future Care
When asked about staff behaviors that would be most important to them if they were to experience a psychiatric emergency or crisis again (with a rating of 1–2 = not important, 4–6 = somewhat important, and 7–9 = most important), the consumer panel rated all of the behaviors we asked about as most important:
- Having the staff listen to me, my story, and my version of events (mean = 7.9)
- Being asked about what treatment I want (mean = 7.9)
- Trying to help me calm down before resorting to forced treatment (mean = 7.9)
- Being asked about what treatments were helpful and not helpful to me in the past (mean = 7.6)
Figure 2 presents the panel’s medication preferences if they should need to take psychiatric medications while being treated for a psychiatric emergency in the future (with a rating of 1–3 = least preferred, 4–6 = no strong feeling, and 7–9 = most preferred).
The panel overwhelmingly preferred pills or capsules as the formulation of choice; this option was given the highest rating of 9 by over 63% of the respondents (N = 54) and was rated a 7 or 8 by 30% of the respondents. There was no strong feeling about liquid medicine or an injection that is agreed to voluntarily (means = 4.9 and 4.7, respectively), with no consensus among the panel on these options, while the panel overwhelming rejected an injection forced on the person involuntarily, which was rated least preferred (mean = 1.5).
Class of medication:
When asked which specific types of medications they would prefer to take if they needed a medication to help them calm down so they would not hurt themselves or others, the respondents preferred anti-anxiety medication (mean = 7.3, rated most preferred by 73% of the respondents); there was no consensus about a combination of an anti-anxiety agent and an antipsychotic, which received a mean rating of 5.7, but was rated 7, 8, or 9 by 47% of the respondents. An antipsychotic alone was least preferred (mean = 4.4).
When asked about specific medications, the number of respondents varied depending on the agent being asked about (n = 18–37 respondents), with the number of respondents generally higher for medications that received higher mean scores. Although no medications were rated as most preferred by the panel as a whole, the respondents preferred the benzodiazepines lorazepam (mean = 6.5, N = 36), diazepam (5.4, N = 37) and clonazepam (5.3, N = 35) and the antipsychotics quetiapine (4.6, N = 30), risperidone (4.1, N = 35), and olanzapine (4.0, N = 28). All the other medications we asked about received mean ratings below 4.0.
Frequency of monitoring while in restraints or seclusion.
When asked how often they felt someone should check on them if they needed to be put in restraints or seclusion, 27 of the 56 respondents to this question (48%) felt that someone should be there all the time, 20 (36%) felt that someone should check pretty often (at least every 15 minutes), and 4 checked both options (18%). Only 3 (5%) respondents felt that checks every 30 minutes would be acceptable, and 2 respondents checked all three options.
The survey respondents were asked to rate a number of treatment interventions and other strategies to help them calm down and avoid being placed in seclusion or restraints or being given forced medication if they should experience a psychiatric emergency in the future. The top five strategies, which were all rated as interventions of choice (i.e., they received a rating of 9 from over 50% of the respondents) were:
- Being listened to about the kind of treatment I want
- Having someone contact my own doctor or therapist
- Being given a medication I request because I know it has helped me in the past
- Having crisis counselors for trauma available
- Having a comfortable bed.
Other interventions that were rated as very helpful were a calm soothing environment, being given something to eat and drink, being able to make a phone call to a friend, being given the opportunity to walk, a clean comfortable waiting room, and the availability of an art room to express feelings. When asked to write in other things that might help them calm down and avoid being placed in seclusion, restraints, or given forced medications, respondents also mentioned being treated with respect and as an adult, contact with and companionship of family members or friends, access to their own doctors and therapists, and access to clergy and religious or spiritual counseling and observances, and availability of external stimulation and distraction.
A major theme that runs through all the responses is the importance of having staff members spend time with them, listen to them, and treat them with respect. As one respondent said “One-to-one interactions help the most. I go to the hospital because I am depressed and suicidal—feel sad, alone, and worthless. The last thing I need is to be left all alone—especially in an ER room with all kinds of ways to do self-harm. I think ‘Obviously they want me to kill myself.’”
Recommendations for Preventing Future Crises
Strategies of choice.
When asked what staff could do after a crisis but before they left the ER or were discharged from the hospital to help them stay well and avoid future psychiatric emergencies or crises, the following were rated as strategies of choice:
- Having someone identified who will call a few days later to see how they are doing
- Being helped to identify and avoid triggers of future episodes
- Being encouraged about their chances of recovery
- Having the ER or hospital set up a follow-up visit with a new doctor if they don’t have one
Other helpful strategies.
The panel also considered the following strategies very helpful: referrals to peer support groups, having someone help them figure out how to pay for medications, and being given self-help materials (workbooks, books about their illness). Write-in recommendations included referral to peer support group/hotline (with phone number of specific contact person), ensuring that the person has a support system in place before discharge, being sure the person has food/drink at home or arranging for help with shopping, helping the person explain the situation to family members, contacting the person’s doctor and letting him or her know what is happening, arranging for assistance with taking medication, arranging for transportation to follow-up appointment, follow-up to see that the person gets to follow-up appointment, reassurance that it was a good decision to come in for help and that one can return when needed and that staff will help, encouragement about the potential for recovery, availability of consumer advocates in the ER as counselors, availability of legal advice if needed, and availability of semi-independent subsidized housing at an affordable cost.
Important contact numbers.
When asked if they knew their doctor’s number, 17 (32%) of the 53 respondents to this question said they knew it by heart, 26 (49%) said they carry it with them all the time, and 19 (36%) that they have it written down at home in a place they can find it right away (there was some overlap in responses). Four respondents noted that they didn’t have a regular doctor, although 3 of these respondents did report having a medical contact number. The overlap in the answers may in part reflect the problem one respondent expressed in a write-in comment “I know it by heart, except in a crisis situation everything goes out the window. I will now carry it with me.” One respondent reported that, because he goes to a clinic, the doctor is not on call after hours and the phone message tells him to dial 911.
We also asked the consumers if they had worked out a plan with their doctors in case there was an emergency or crisis in the middle of the night, and if they had, what the plan was. Of the 53 respondents to this question, 27 (51%) had no plan for such a crisis. although 7 of these respondents did indicate what they would do (e.g., call 911, call friend or family to help, go to the emergency room). There was a good bit of overlap in the other responses, with 15 of the remaining 26 respondents indicating more than one plan: 17 respondents said that they would go the emergency room, 10 that they would call family or friend to help, 6 that they would call the doctor on call, 7 that they would call their doctor, and 7 that they would call 911. Only a few respondents provided details of a plan; these included calling the case manager, calling the doctor (because he said it was ok to call him “day or night”), calling the therapist at home, calling the crisis line at the mental health center (if the psychiatric emergency is not threatening to safety), calling the crisis line at the hospital ER to speak with the psychiatric evaluator, calling an emergency number where one can speak with a first contact person who then calls the doctor on call, calling someone on a support group call list, who would find someone to come over and help, and calling the ER help number to speak with a doctor.
Finally, we asked the panel about advance directives: Only 18 (32%) had an advanced directive at the time they completed the survey, although 42 respondents (74%) said they would like to do one if given the chance.
Key Workshop Recommendations
The written survey served as a lead-in to a workshop in which the participants were asked to develop and prioritize recommendations for improving psychiatric emergency care in five major phases of the emergency services process. The participants in the four Emergency Services Forums generated a total of 571 recommendations for improving psychiatric emergency care (129 in Poughkeepsie, 147 in Syracuse, 160 in Marshalltown, and 135 in Los Angeles). After all participants had posted their recommendations, duplications were eliminated, similar items were grouped into themes, and the participants voted for the top recommendations in each of the five phases. The top recommendations for each phase of the emergency services process are summarized in Table 6 (note that many of the recommendations are relevant to more than one phase of the process).
Comparison with the Professional Survey
The consumers’ preferences and recommendations, as expressed in the survey and workshops, largely agree with the ratings of the mental health professionals who completed the expert consensus survey on psychiatric and behavioral emergencies. 10
Patient autonomy and involvement in decision-making.
As noted above, and frequently echoed by the consumers, the professionals who were surveyed considered collaboration between patient and clinician whenever possible and honoring the patient’s wishes and preferences to be very important in achieving the most favorable short-term and particularly long-term outcomes for patients (Guideline 2C). Just as the consumers repeatedly stressed the importance of having staff talk to them and listen to them, the experts felt that talking to the patient was always one of the most important strategies in the initial intervention with an agitated patient in a psychiatric crisis (Guidelines 5A, 6A, and 7A).
Use of medication.
Just as the consumers stressed the importance of being asked about their medication preferences and about medications that had helped them in the past, the professional panel stressed the importance of obtaining as accurate a history as possible of the patient’s previous experience with medication (e.g., allergies, adverse reactions, response, medical contraindications) and determining the patient’s preference for treatment before intervening with medication in a psychiatric emergency (Guideline 1C). The mental health professionals’ perceptions of consumer preferences concerning types of interventions and medications largely agreed with the preferences expressed by the consumers. The professionals believed that consumers would find oral medication most acceptable; they believed that consumers would find physical restraint least acceptable and were divided as to how they thought consumers would consider injectable medication and seclusion (Guideline 2E). Interestingly, 93% of the consumers rated pills or capsules as the most preferred formulation, with no consensus on liquid medication (Figure 2), whereas the professional panel rated oral liquid concentrate, an orally dissolving formulation, or an IM formulation as the preferred routes of administrations for medication treatment of behavioral emergencies, while tablets were only a second-line option (Guideline 4F). Asked about specific classes of medications, the professionals surveyed felt that consumers would prefer benzodiazepines as treatment of choice followed by atypical antipsychotics (Guideline 2F). This agreed with the consumers’ responses when asked what types of medications they would prefer to take if they should need medication during a future crisis: the consumers gave the highest ratings to anti-anxiety medications followed by a combination of an antipsychotic and an anti-anxiety medication; among the specific agents, benzodiazepines received the highest ratings, followed by the atypical antipsychotics (Figure 2). The antipsychotic medication most frequently used for injections, haloperidol, was not recommended by consumers. Low potency antipsychotics were not recommended by either group.
Use of physical restraint and seclusion.
The consumers clearly felt that restraint and seclusion were overused, often in situations in which there was insufficient justification. When asked about appropriate interventions for an imminently violent patient, the professional panel recommended verbal interventions, voluntary medication, show of force, emergency medication, and offering food, beverage and other assistance as first choice strategies, with the use of physical restraint or seclusion considered second line options that should be tried only if other strategies had not succeeded (Guideline 2B).
In the consumer survey, 54% of those who reported being placed in restraints at some point said that this experience had made them unwilling to seek out psychiatric care after this; although there was no clear-cut consensus among the experts as to the effect of chemical and physical restraint on patients’ long-term adherence to treatment, only 38% felt that the use of restraint would be likely to have a negative impact on long-term adherence (Guideline 2G).
When asked about what level of monitoring was appropriate for someone while in restraints or seclusion, the professionals (Guideline 3F) and the consumers agreed that continuous monitoring was most acceptable and in-person evaluation at least every 15 minutes was also appropriate, but that longer intervals were not generally appropriate. The professional panel also recommended reordering of restraint at 2–4 hour intervals. These issues are now the subject of regulation.
Limitations of Current Study and Directions for Future Research
A key element in the construction of this sample was the involvement of consumer advocacy and support organizations in each area. Although not all of the participants were themselves involved in consumer advocacy groups, all were known to and invited by advocates. As such, the panel clearly represents a convenience sample. These workshops were a time-consuming and emotional activity for the participants. It is therefore reasonable to assume that those who chose to participate were motivated to communicate their experiences with psychiatric emergency care. In addition, the forum partners who were helping to organize the forums were asked to select participants whom they thought would be able to tolerate the day-long experience and would be able to address their emergency services experiences both verbally and in writing. Therefore, the panel sampled here may not be representative of all individuals who use psychiatric emergency services. The preponderance of affective illness, the higher educational attainment, and the low frequency of regular antipsychotic use relative to emergency service populations described in the literature all suggest that this sample is not representative in some ways. This is a significant limitation of this study but was perhaps unavoidable. It would be desirable to obtain data from a larger sample recruited by other means but it is difficult to imagine a sample that is not in some way self-selected and motivated by a desire to provide feedback. Finally, nearly one third of those who had experienced restraint and nearly a quarter of those who had experienced seclusion reported that they had a poor memory for the details of their experiences of restraint or seclusion.
The consumers surveyed here reported a high frequency, approximately 20%–25%, of very adverse experiences. The high rate of abusive behavior reported here may surprise the provider community. While sampling bias may be a factor here, if the true frequency is even a small fraction of this rate, it represents a serious problem for which we as providers must assume responsibility. Although data on such experiences have not been reported, the finding is credible because of the wide variation in restraint practices and sporadic reports of serious injury and death. That these things occur with some frequency is undeniable and the burden is on providers to prevent abuses and deal with consumer perceptions. Regulations now mandate tracking of injuries and debriefing, which will hopefully provide larger samples and help consumers and providers process such experiences together.
Some areas of general agreement are noted between the recommendations of the consumer panel and those of the experts in emergency psychiatry surveyed for the Expert Consensus Guidelines on the Treatment of Behavioral Emergencies. These include the desirability of verbal interventions, a collaborative approach, and the use of oral medications guided by the individual consumer’s problems, medication experiences, and preferences. The majority of the consumer panel reported adverse experiences with general hospital emergency rooms and, in fact, called for the development of specialized psychiatric emergency services such as those recommended in the Expert Consensus Guidelines. It is also worth noting that one fifth of the consumer panel attributed their emergency contact to lack of access to more routine mental health care.
Although Sheline and Nelson 27 found that consumers preferred seclusion to restraint, the consumers surveyed in this study disliked being left alone in a cold sterile environment. The concept of providing distraction from internal stimulation recommended by many of the respondents is an interesting one. From the provider viewpoint, less stimulation is central to any postulated benefit of seclusion. However, there might be an advantage to some forms of stimulation for some patients if this could be provided safely.
The consumers clearly do not reject medications categorically. Almost half indicated that they wanted medications and a similar number indicated benefit from medications, although most complained of unwanted side effects. One of the moderators (JS) noted that, in nearly all cases, it was not the drug or drugs that the consumers objected to but the false promise of a simplistic cure by pharmaceuticals and the misuse of medications by practitioners.
The consumer panel clearly preferred anti-anxiety medications (e.g., benzodiazepines) over antipsychotics. When asked about specific agents, the consumers gave the highest ratings to lorazepam, although they did not endorse any medication as first line, perhaps because of their clear preference for verbal interventions. The consumers gave extremely low ratings to haloperidol (least preferred, mean = 2.5). Given the preponderance of nonpsychotic disorders in this sample of consumers, the professional panel would agree with the consumers that benzodiazepines would be the most appropriate choice for many of the conditions reported here.
This raises a philosophic question as to the basis for disregarding consumer preferences. If benzodiazepines are appropriate and most consumers prefer benzodiazepines, then when is it appropriate to administer an antipsychotic involuntarily? A single dose of parenteral medication is unlikely to alter the course of the patient’s underlying condition but may alter the individual’s attitude toward continued treatment in significant ways. It could be argued that insufficient treatment may expose the patient and others to further agitation and aggressive behavior. However, agitation often fluctuates and requires continuous assessment and repeated intervention. If a patient fails to respond to a single dose of the first medication, it is recommended that it be repeated several times before switching strategies because the dose required is unpredictable. It could then be argued, based on consumer preference, that benzodiazepines be the first line until such time as the patient improves and begins to participate more effectively, or until the failure of benzodiazepines has been demonstrated. Most patients will respond within a few doses regardless of the initial treatment. In summary, the consumers in this survey indicated that they wanted medication in many cases. Given that there is little evidence to support one medication strategy over another for managing behavioral emergencies, it seems reasonable to offer patients a choice and to consider the medications they prefer in the absence of any specific contraindication.
A second issue is diagnostic confidence. For many patients, antipsychotics, though effective for agitation, will not ultimately emerge as the treatment of choice for their condition and antipsychotic exposure will be unnecessary. For those patients for whom antipsychotics are appropriate, an initial trial of benzodiazepines for agitation might result in a delay of several hours in initiation of antipsychotic treatment. However, given the primacy of process issues for consumers, this delay during which providers attempt to improve their assessment and work with the patient toward a consensus treatment plan would be of critical importance.
When asked about discharge planning and post-discharge follow-up, the consumers stressed having someone identified who would call to follow-up a few days later. This strategy has also been recommended by the American Psychiatric Association (APA) Task Force on Psychiatric Emergency Services. 28
The two surveys that we describe here form a dialog between expert providers and experienced consumers. Though some individuals would undoubtedly see the differences as irreconcilable, the gulf between the two groups is not as wide as some might think and, in fact, there is remarkable agreement in principle. However, operationalizing best practices in emergency settings has been difficult. Consumers identify a number of staffing, training and other resource issues that the authors believe to be at least partly a result of the ad hoc organization of psychiatric emergency services. Although Gerson and Bassuk 29 called for psychiatric emergency services to be conceived as organizationally unique 20 years ago, they are still poorly defined and regulated in many areas. The recent APA Task Force report may help with this problem.
The literature also suggests that some problems experienced by consumers are related to institutional and provider attitudes. JCAHO regulations now require a consumer perspective in staff training on restraint and seclusion. In that vein, it is hoped that information from this survey will put a human face on the frightening confrontations in potential restraint and seclusion episodes and help shift the balance from reflexive control measures to more reflective care.