Frequent visitors at the psychiatric emergency room – A literature review

Frequent visitors at the psychiatric emergency room (PER) constitute a small subgroup of patients, yet they are responsible for a disproportionate number of visits and thus claim considerable resources. Their needs are often left unmet and their repetitive visits reflect their dissatisfaction as well as that of PERs' staff. Motivated by these dilemmas, this study systematically reviews the literature about frequent visitors at PER and seeks to answer two questions: What characterizes frequent visitors at PER in the literature? and What characterizes PER in the literature? Based on 29 studies, this paper offers answers to the two questions based on a strength weakness opportunities and threats (SWOT) analysis. The results of the review and subsequent analysis of the literature revealed the multiplicity and complexity of frequent visitors' characteristics and how they appear to converge. Commonalities were more difficult to identify in PER characteristics. In some cases, this happened because the characteristics were poorly described or were context specific. As a result, it was not easy to compare the studies on PER. Based on SWOT and the findings of the analysis, the paper proposes new venues of research and suggests how the field of mental health might develop by taking into account its opportunities and threats.

Introduction have shown that PER are increasingly used by visitors with non-urgent needs [22][23][24], implying that the quality and availability of care for visitors suffering an acute crisis is being compromised [24]. These findings are in line with other studies that have pointed out that a number of visits of persons to PER have increased among those who do not represent Btruep sychiatric emergencies but who use PER as a source of support [22,23,25]. Consequently, those non-urgent and frequent visits claim a disproportionate amount of PER's resources [24] and put a substantial financial strain on PER [5]. Although PER represent a medical specialty, it can be concluded that they function as a Brevolving door^for particularly persons who frequently visit PER [11,26,27] and who often have non-urgent mental health problems and needs [22].
Research question 1: What characterizes frequent visitors at PER in the literature?

Psychiatric emergency rooms
The field of health care and mental health care in particular have undergone significant changes over the past decades. As a consequence the same can be said about the development of PER and staff working in a psychiatric care setting that had to be adjusted in order to meet the new challenges.
PER is both time and staff-intensive [25]. Since the workload in PER is increasing [27] and utilization rates are soaring [28] it becomes more difficult for staff at PER to provide quality services. Central to the concept of psychiatric emergency is the subjective quality, the unpredictable nature of the emergencies, the wide range of diagnoses and symptoms of people visiting PER, lack of prior assessment or adequate planning, result uncertainty, severity, urgency etc. which puts many demands on staff at PER and makes PER a challenging workplace [29]. One consequence are inaccurate diagnoses that have shown to be associated with poor treatment outcomes and overuse of the most expensive types of services e.g. hospitalization [30].
Since the domain of PER is so far-reaching in terms of services that they provide and the variety of patients they serve, it is difficult to make a comprehensive evaluation of PER [1]. Since PER are the context where frequent visitors are comprehensively evaluated, their needs are assessed, and they are cared for, the characteristics and specific dimensions of PER are important to explore.
Research question 2: What characterizes PER in the literature?
The aim of this paper is to explore what characterizes frequent visitors to PER and PER themselves by critically evaluating previous studies and providing a systematic review by addressing the two main research questions. Given the changing nature of PER's frequent visitors and PER, this study provides an insight into the development of both over time. For each question, and inspired by Jackson et al. [31], a SWOT 1 analysis will be performed to 1 SWOT is one of the most known approaches used for analysis of a company's strategic position being the acronym for Strengths, Weaknesses, Opportunities and Threats. The idea of ensuring a fit between the external situation (threats and opportunities) and own internal qualities (strengths and weaknesses) has shown to be very popular ( [32].Hill, T. and R. Westbrook, SWOT analysis: it's time for a product recall. Long range planning, 1997. 30(1): p. 46-52.) not only in the field of business administration but also is widely being used for any kind of strategic planning or examination of projects, organizations, companies or other ventures. consider the strengths and weaknesses of the previous research as well as opportunities and threats in it. The latter two aspects will lay the ground for suggestions for future research.

Literature review
The papers relevant for this review were identified by a computer-based search done in the Web of Science and PubMed databases. A key word search was conducted by combining the search term Bpsychiatric emergency^with terms associated with a repetitive behavior or high frequency (Bfrequent^, Breturn^, Bmultiple^, Brepeat^,^recurrent^,^high^or^increase^). The key word search was limited to the titles of articles. No time restrictions were imposed in order to cover as broad a field of research as possible.

Inclusion criteria and selection process
The citations retrieved were scrutinized by reading the titles/abstracts/articles. The following inclusion criteria were used to identify studies to be included in the review; they had to be (1) published in a peer-reviewed journal in English; (2) based on original empirical data analysis; and (3) focused on frequent visitors in psychiatric emergency care settings.

Results
In total, 67 items were identified in the Web of Science database and 23 in the by PubMed database. After duplicates were excluded, the total number of studies was 65. An additional 36 were excluded because they did not meet the inclusion criteria. Table 1 lists the author(s), year, data's origin, method, sample size and data of the remaining 29 articles.

Description of findings
Of the articles included in the review, 13 were published before 2000 (1970 to 2000) and 16 after 2000 (2001 to 2016). The articles were mainly published in journals within the fields of health care sciences, public, environmental and occupational health and psychiatry with very few exceptions in nursing, social work or emergency medicine. The majority of articles (14) was published in Psychiatric Services, which before 1995 was called Hospital and Community Psychiatry. All the journals are ranked and have impact factors varying between 0.769 and 5.605, as reported in Web of Science for 2015. Data for 16 of the articles was collected in USA, 9 in Europe and 4 in Canada, showing the clear dominance of the American research community in this field.
Twenty of the studies included had a quantitative approach focusing on register data of patients. Three additional quantitative studies concentrated on psychiatric emergency services, one on staff members, and another one on an implementation of an intervention. Out of 29 studies, one had a qualitative approach, interviewing staff about their categorization process of frequent visitors [2]. Five studies used a mixed method approach, combining interviews (or observations) of patients/nurses with register data. Out of 29 studies, one focused on a patient group below 18 years [33], whereas the remainder of 28 focused on the adult population, though this was not explicitly stated in many of the articles. None of the studies included explicit organizational/economic perspectives.
The quantitative data was analyzed using statistical tests, mainly regression analysis. In the qualitative study, discourse analysis was used with an underlying social constructionist framework. Furthermore, one article was a case control study [34], and another was a community based participatory research study [35].
Research question 1: What characterizes frequent visitors at PER in the literature? Table 2 is a summary of the definitions used in the literature reviewed, either as determined in the introduction or method section or as part of the results if some categorization was made there.

Strengths
One strength of the literature is that most of the articles are based on large samples, which allows for generalization to the population. Several papers went beyond descriptive analysis and applied more sophisticated statistical techniques, such as logistic regression or linear regression analysis or both [e.g. 3,5,16,27]. This leads to a more thorough understanding of the causal inference. Most of the studies used a longitudinal design, which showed how frequent visitors as a group change over time and allowed for better understanding of behavioral (proxied) patterns of that patient group. A few studies used new methodological approaches, e.g., mixed method designs [27,[34][35][36]40]. This provides a better understanding for the complexity of frequent visitors as a group, no less from the perspective of significant others and nurses.
The descriptions of frequent visitors in the literature are another strength. On an operational level, frequent visitors were often described as being single [e.g. 39 15,16,26,33,34,39,41,44]. This substantial variation of characteristics indicates that this group has a rather heterogenic need profile, implying a complexity in supplying care for those persons. In summary, many variables are repetitive (e.g., diagnosis variables, which are found in almost all articles), whereas other variables are context-specific characteristics, like race or socioeconomic factors.

Weaknesses
One weakness of the literature is the lack of current data. Though six studies were published after 2010, the data used in them was relatively dated; the latest study was  [52]. This has left a gap in the years covered and a need for studies using data more collected recently so as to understand current developments. In addition, majority of the studies used a quantitative approach based on register data, suggesting a stronger focus is needed on more qualitative or mixed method designs that provide an opportunity to gain an indepth understanding about frequent visitors. Basing studies on register data also meant that they lacked the perspectives on non-patients, such as organizational and staff perspectives or from significant others. Especially the perspective of the frequent visitors of PER might be important to consider, yet it is rarely taken into account when quantitative methods are used. Majority of the studies did not report practical implications and thus lacked relevance for praxis. The studies used many different definitions and terms for frequent visitors, which unnecessarily complicates comparison of the studies. Though the diagnoses of frequent visitors was mentioned in most articles, they were described and analyzed in different ways (e.g., three diagnoses are given per visit or the most common diagnosis is picked or articles mention the principle diagnosis or severe primary diagnosis). This creates diagnostic confusion particularly in the context of longitudinal studies, which most of the studies are, and thus they lack transparency.
Additionally, only in few quantitative studies a control group was used [e.g. 5, 16, 34], which could be seen as a methodological weakness.

Opportunities
One way of moving forward in the research area would be to develop the concept of frequent visitors further by conducting qualitative studies from which qualitative definitions of frequent visitors could be derived. This could allow for an in-depth understanding of frequent visitors and their needs, which would prepare the way for the development of effective interventions.
Another way forward would be to explore the different dimensions of the definition of frequent visitors, by exploring the views of different external actors. This might contribute to a more holistic definition because different perspectives would be taken into account.
Finally, one could study frequent visitors' specific needs, life styles, behavior, and social networks in order to develop the interventions that are necessary and that could be implemented with a long-term perspective.

Threats
There is a danger of misinterpreting the results of the studies if the settings, definitions, and terms are so varied. Because the majority of studies were conducted in US, the findings lack a broader perspective. Further, most of the studies missed taking the findings to a higher theoretical level. They did not explain their results by using the underlying theoretical reasons, but instead were driven mainly by empirical aims. Thus, although the studies succeeded in identifying patterns and characteristics of frequent visitors, they did not provide reasons for the results. As a consequence, the situation of frequent visitors would not be improved based on these studies because knowing only about their patients' patterns is not enough. More focus should be paid to practical implications in order to make a difference in frequent visitors' situations.  Categories Borganizational context,^Bgeographical context,^Btypes of services and facilities, Bprocesses and procedures,^Bservice usage,^Bhealth care system,^and Bstaff^are shown below in Table 3. The characteristics of the PER were derived from the literature included in the review that were mainly described in the introduction and method sections of the articles.

Strengths
The studies' strengths lay in their transparency concerning the location of the PER, its ownership, and the PER's capacity. In the studies reviewed, the PER was often located at a university hospital or a public hospital; in one case the PER was community-based [40]. Some studies did not mention the ownership of the hospital, but almost all stated the name of the hospital or city. Most of the articles provided some sort of a description of the PER, such as having open access, 24 h 7 days a week, being available for everyone, or being combined with inpatient or outpatient services. In this way, some sort of context description was provided, which is important in order to correctly interpret and understand the results of the study. Another strength is that several studies mentioned the total number of visits per year. Particular when provided in combination with number of inhabitants of the city or the catchment area of a PER, it gives a good insight about the size of the hospital and its capacity. In many studies, the number of visits per year was described in the results section, which could be one reason why it is not included in Table 3. One could also draw conclusions about the patients visiting a PER by knowing about the PER's location, e.g., rural, suburban, or whether the PER is located in an economically disadvantaged area characterized by high unemployment rates or is in an area dominated by a certain ethnographic subgroup, which were mentioned a number of times.

Weaknesses
The primary weakness of the studies lays in the limited descriptions of organizational structure and processes within PER as well as incomplete descriptions of the local and national contexts in which PER were embedded. Only a limited number of studies mentioned the health care system and its specifics. None of the studies discussed organizational structures such as power and hierarchy or explored the relationship between those structures and the use of resources, be they tangible or intangible. Further, all but one study [33] focused on adult patients and thus excluded adolescents as a group. This is especially a weakness, given that a number of psychotic disorders can be detected at the early age [e.g. 53]. By diagnosing e.g. depression in adolescents, interventions could be implemented earlier and with a preventive purpose [54], which could be beneficial for the patient and increase the efficiency of the service provision.
A further weak point of the studies is that they seldom described the catchment areas, making it impossible to know the total number of visits in relation to population served. Few studies [e.g. 15,46,48] have addressed the dynamics within the population in terms of its growth, composition, density, or involvement with other special treatment units. Further, the studies have not addressed the infrastructural aspects such as transportation and accessibility of PER. Finally, the studies have limited discussions on the types of services PER offer, which makes it hard to compare PER across different studies.

Opportunities
One aspect that could enrich studies of PER would be exploring the team aspect of staff work. Studies in other health-related contexts with a focus on acute and intensive care settings have indicated that teamwork and the dynamics in health care teams are important, given that teams handle complex work assignments better than individuals [55][56][57]. The complexity of PER and its assignment might thus serve as a golden opportunity in exploring the potential importance of team work and its outcomes. These types of study would be of particular interest in contexts like Swedish PER units, where the staff works in teams with the help of the triage method. Another opportunity for learning more about PER would be to approach them from an organizational perspective, exploring their organizational structure, economic resources, hierarchical relationships between different categories of staff, and how the division of labor between these groups is organized. Exploring these aspects might allow for a more nuanced view of PER and their structuresas well as a better understanding of how those are related to PER's organizational efficiency and performance. This could be of particular relevance since several studies mention how scarce resources at PER are [e.g. 5,15]. Exploring staff members' experiences of their working settings, lived world, and well-being could provide yet another opportunity to better understand how to motivate PER staff and consequently to improve the quality of their work and their level of work satisfaction. Finally, understanding the role of geography (i.e., location in urban or rural areas and communicational and transportation conditions) could shed light on PER's functioning and the differences that appear between the studies.

Threats
Again, a threat can be seen in the dominance of the US-based studies, which results in a limited view on the context of PER and their patients. Further, the studies lack information on the methods used within psychiatric emergency care, e.g., in developing countries. Studies tend to focus on a medical/clinical perspective, which leads to a lack of multidisiplinarity. Putting more stress on the nursing/caring perspective within PER could be of importance because the nursing staff is highly involved and are the first to encounter the patient. Another threat is the lack of exploration of geographical aspects of patients vis à vis PER (e.g., rural area vs metropolitan area). The organizational and management structures within PER are also underexplored; knowing more about them could further the understanding of the role and functioning of PER.

Discussion and conclusion
This literature review posed two research questions: 1) What characterizes frequent visitors at PER in the literature? and 2) What characterizes PER in the literature? Both were explored by the means of a systematic review combined with a SWOT analysis. One aspect that emerged from the literature review is the inconsistent use of the terms BPER^and Bfrequent visitors.T he broad spectrum of differences in terms can be partly explained by the variation of definitions, variation in different health care settings and welfare systems, geographical and climate differences, and, not least, different populations served. The diversity of conceptualization of PER and frequent visitors represents a challenge. According to McArthur, even a common definition concerning how emergency psychiatric care defines itself needs to be found [8]. In addition, an accepted operational definition of frequent visitors has not yet been proposed [18]. This literature review attempted to provide a first step in developing accepted operational definitions for PER and frequent visitors by surveying the literature on both concepts: PER and frequent visitors. The two are interdependent and thus should both be acknowledged in future definitions of either concept. Though there has been some discussion about what a Btrue^psychiatric emergency constitute, the basis for all definitions should be found in the urgency of the visitors' need for care, either as experienced by the person or by others.
In most of the papers, frequent visitors are quantified and objectified, with the studies dealing with frequent visitors at PER based on register data and archival data. There is a lack of qualitative studies investigating the perspective of persons in care in terms of their needs, their satisfaction with PER services, and their life style and living situation. Such studies are needed in order to provide appropriate support and help to frequent visitors and might identify whether the type of help needed could be offered outside of PER. The studies are empirically driven and do not seek to establish models or theories. One way forward would be to look at the findings of the studies and to try to further develop existing models and eventually create new ones. This would also allow for learning more about, e.g., frequent visitors' attitudes and behaviors by applying existing models. One such model could be, e.g., the Tidal model [19,58], which was developed for psychiatric care settings. Another possibility for seeking explanation for the results would be the application of existing theories, e.g., Giddens' structuration theory that poses that society is based on social actions and should be understood in terms of agency (relationship with other people based on interactions) and structure (rules and resources). Agency and structure exist in duality (i.e., they involve reciprocity between actors and collectives) [59]. This review revealed that the studies focus mainly on the characteristics of the agent (the frequent visitors) and do not do enough to take into account the implications of PER and the challenges PER faces (structural aspects) when caring for frequent visitors. In order to understand the agency and the structure, one needs to study their interactive nature. Applied here, it means that future studies should focus on understanding the role of PER for frequent visitors and the role of frequent visitors for PER, which so far has not found its way into the literature [cf . 18].
The opportunity and threat discussions further revealed the need for studies that address person-centeredness. Person-centered frameworks have recognized the important role of the care environment with its hinders and facilitating roles [60], thus context needs to be considered more when investigating frequent visitors at PER. Such investigations should include physical setting, organizational systems, professional competencies, human relationships, and hierarchies [61], as well as the interpersonal context [20]. PER provides a unique context for caring processes to occur and for interactions between staff and patient. The first encounters between frequent visitors and staff and how staff interacts with and cares for them is part of the therapeutic relationship [e.g. 62, [63][64][65]. Concepts of transference and countertransference might need to be taken into account, in particular in this setting.
Further the literature has shown that frequent visitors psychological and service needs are complex, vary from patient to patient and between the contexts. Thus, understanding patients' needs better would allow for better fitted and tailored interventions that also strive for continuity. The latter could only be achieved by collaboration with external actors.
In reviewing the literature, it became apparent that PER continue to face challenges, given their fast-paced environments, and insufficient time for staff to provide giving diagnoses and care while serving acutely ill, vulnerable persons that need interhuman and interpersonal interaction. Instead, the literature is in agreement when addressing PERs' frequent visitors as 'hard to treat' and 'difficult patients' [20,47,[66][67][68] or those that cannot profit from psychiatric treatment [2]. Little information is revealed about how frequent visitors could profit from the newly gained insights of the studies. Future research would benefit from applying the aforementioned theories or models or other person-centered frameworks that stress acknowledging the patient as an equal partner in the health care process. Such theories or models also conceptualize the context and the person(s) of the studies, which might lead to the development of the former and improved care for the latter.