Background
Severe mental illnesses (SMI) are associated with a complex mixture of clinical and social needs [
1]. SMI-patients are often defined in relation to their long-lasting treatment situation, including criteria concerning the persistent nature of the disease (operationalized as >2 years of care) and an indicator of dysfunctioning (e.g. GAF score of less than 50) [
2,
3]. SMI constitute only a minority of all mental illnesses; in many cases of mental illness dysfunctioning is mild and not chronic. Due to the chronic nature of SMI, continuing support from Mental Health Care (MHC) is crucial. The study on long-term care dependent patients (LZA-study) in Amsterdam found that the estimated amount of SMI-patients in care increased from 3000 in the year 2000 to 4576 patients in 2005 [
4]. This increase was not explained as an increase in the total number of patients, but as in increase in the percentage of patients that was in treatment by the MHC; i.e. a better coverage by the MHC institutions among SMI-patients. In the past decades, the care for SMI has changed from long-term institutionalization to ambulatory care in the community and in sheltered housing. The current treatment is aimed at treating SMI-patients in an ambulatory setting instead of a clinical setting. Outreaching treatment teams, such as ACT (assertive community treatment) have become more and more common and since this increases the ambulatory treatment options, (compulsory) admissions are less often necessary and can often stay short.
Although treatment opportunities for SMI-patients have improved over the years, some patients still experience periods of social decline, and their behavior can cause nuisance to others. Of all long-term mental health patients in Amsterdam, 4–6 % experienced a crisis MHC contact outside office hours annually i.e. five crisis MHC contacts per 1000 inhabitants from 2000 to 2004 [
5]. Although part of the SMI-patients function well with adequate treatment, the group remains vulnerable. For a small group of patients crisis interventions and compulsory admissions are common, the quality of life is below average, unmet needs remain present and suicide rates are up to nine times that of the general population [
6,
7]. It seems that their social vulnerability, their complex needs of care, poor treatment adherence, comorbidities, and an unstructured existence, impede the way for providing the right care and even influence it negatively [
8,
9].
The expected effects of dehospitalization in the ′80s and ′90s were: less restriction resulting in less aggression, more commitment from the social network, and a better (re-)integration into society. However, there were also fears, like the risk of overburdening the social network, increased use of medication and an increase in crisis MHC contacts [
10]. Several publications report the incidence and risk factors for crisis MHC contacts within the mental health care system [
5,
11‐
13]. Crisis MHC contacts are generally divided between those within office hours and those outside of office hours.
In Amsterdam, in the ′90s, similar to the situation in London [
14], the fears seemed to have become reality. The public complained about overt psychiatric behavior and the associated nuisance. Dutch media reported an increase of problems in the streets that were clearly caused by people suffering from psychiatric disorders, such as paranoid or uninhibited sexual expressions, but also signs of deterioration [
4,
10]. In addition psychiatric emergency units were overburdened. The psychiatric crisis chain in Amsterdam was reorganized in response, and the public health service got an active role in the coordination and screening of psychiatric patients [
5,
15], [Raat, Municipality of Amsterdam, 2001].
The police gets involved in most crisis situations that develop outside of the scope of MHC and result in a public expression, mainly notified by third parties like other citizens, neighbors or housing corporations. In the past years, the Safetynet department of the public mental health service (PMHC) pre-screened all persons that experience a public crisis event and for whom the police suspects psychiatric problems. The number of these pre-screens has increased over the years, from 5 to 9 per 1000 adult inhabitants per year from 2003 to 2013 [Research and statistics department of the city of Amsterdam, Safetynet]. This might be due to the fact that a request for these consultations became more and more part of the regular process among the police. About 30–40 % of these contacts result in a referral to a crisis MHC contact within the unit for emergency psychiatry (SPOR), this percentage has remained fairly stable over the years. Those not referred to SPOR result in a variety of other services including social care or are followed by judicial trajectories.
These crises in the public domain have a large impact on the client, the public and the police. Little attention has been given to these public expressions and interventions in the literature. A study in Groningen about the role of the police in mental health crisis situations [
16] showed that half of the individuals in crisis were disengaged from MHC in the year prior to the crisis. A body of literature is focused on factors that are related to (the re-occurrence of) psychiatric emergencies within MHC. Some studies shed light on socio-demographics and show an increased risk for migrants [
17,
18] and those with a small social network [
11]. Other studies focus on the influence of continuity of services and transitions, and show that a change of service provider, discharge from a clinic, are associated with an increased risk [
8,
19‐
21]. However, most of these studies only included persons that were compulsory admitted after a crisis.
In this article we describe how public crises develop, what events precede them, and what characteristics of care surround them, in order to identify leads for prevention of crises and possibly better cooperation between the involved parties. The aim of our retrospective case-file study is to determine profiles of SMI-patients and their care among those experiencing multiple public crisis events in the period from 2004 to 2012. The collaboration between the MHC and PMHC for this group has not been described previously. Due to the chronicity of the illness, long-term treatment is needed and therefore patterns of care including the duration and intensity should be studied over a longer period of time [
22]. A chronological overview of the characteristics of public crisis events and the offered help for these individuals provides insight into the frequency of crisis interventions in relation to characteristics of pre- and post-events and treatment. Altogether these findings make it possible to identify high risk groups with multiple crisis events and predict crisis situations and may contribute to the prevention of crisis situations.
Description of the mental health care system in Amsterdam for psychiatric patients in crisis
Mental Health Care (MHC) and addiction care for SMI patients in Amsterdam are financed by the health insurance. In addition to regular services, MHC has several outreaching systems. It started with the rehab team that looked for vulnerable people anywhere in the city and if necessary, tried to persuade them to accept care. This later changed into teams working according to the assertive community treatment methodology [
23]. Within office hours each institution operates their own crisis unit, outside office hours a central crisis unit is operational called SPOR [
24]. The addiction care has a separate crisis unit, the clinical detoxification unit (CODA). At SPOR a (resident) psychiatrist decides whether or not a client needs to be admitted (compulsory or voluntary) [
11,
15].
An acute compulsory admission in the Netherlands can be ordered in an acute crisis situation to avoid danger resulting from a psychiatric disorder. The procedure requires a psychiatrist to assess the patient. Based on the psychiatric condition and the danger to be averted, the mayor decides whether to assign the ACA. Within 3 days of admission to a mental hospital, a judge decides on the prolongation of the measure, by maximally 3 weeks. For a long term compulsory admission a decision of a judge is required before the start of the admission. These admissions have a duration of 6 months, after which a prolongation can be requested through court. Another possibility is discharge under court-ordered terms, such as the obligation to undergo treatment.
Clients could be referred to a crisis unit by medical doctors and by social psychiatric nurses (SPN’s) from the Safetynet department of PMHC.
The PMHC is financed by the city and intervenes when vulnerable people with mental and social problems are not able to provide for basic needs themselves and are not able or willing to organize the care they need to function in society in the opinion of professional care givers. Since 2006, integrated care is offered by cooperations between PMHC, shelters, MHC, addiction care, social services and the welfare agency. For each client an individual treatment plan is made, and a casemanager responsible for the coordination of care is assigned. The PMHC Safetynet department responds to reported crisis situations of vulnerable people. The SPN’s make an inventory of the situation and organize care, or coordinate existing care. Annually, about 2000 of these situations are reported. In addition, more than 5000 crisis signals came from the police who either responded to a signal from citizens, or responded directly to a situation they encountered in the public domain involving a vulnerable person. They could request an immediate consultation from a SPN 24 h a day, to decide which trajectory should be followed. In about 40 % of these crisis interventions, the client was sent to the SPOR. Some were sent to CODA. The rest was either sent home, with a referral to regular care or stayed to undergo the judicial consequences of their actions.
The situation concerning crisis signals from the police as described above was the situation up to December 2014. Since 2015 the police can directly refer patients to the SPOR and CODA. SPN’s from the Safetynet can be consulted in case of doubt. The situation in this article therefor describes the null-situation, to which future data can be compared [
25].
Discussion
Of the total study group of 323 SMI-patients, 92 had one or more crisis interventions by the PMHC in a period of 9 years, and 47 had three of more PMHC crisis interventions in this period (15 %). Three distinct profiles could be distinguished within this group.
The first profile existed of SMI-patients with ongoing MHC contacts, the second profile experienced CI’s mainly after discharge from the psychiatric clinic and the third profile lacked continuous MHC contacts.
Profile 1: Crisis during continuous MHC
Results show that even stable contacts with MHC cannot always prevent the occurrence of decompensation (mainly due to medication nonadherence) and social problems. Although some CI’s were preceded by an increased frequency of contacts and signals from family, friends and neighbors, the actual occurrence of CI’s was not prevented in this profile. Often because voluntary interventions did not result in improvement, and because compulsory interventions are only possible when the threat of danger is clear. In these cases, the current system does not seem to provide for more preventive action and since some periods of decompensation seems to be inevitable, compulsory admission seems to be an effective way to restore a stable situation. A more preventive approach might be feasible for crisis situations in the social domain. In some cases, the home situation had deteriorated severely in spite of regular MHC contacts. From this study, we cannot conclude whether more preventive action was feasible, nor whether the lack of motivation of the client hampered all possible interventions, but the fact that signals of an upcoming crises are picked up is a first prerequisite for prevention. Especially since these CI’s often occur at home, and the Dutch care system is switching more towards integrated care in the neighborhoods. In different neighborhoods, collaborations between different partners are being created, depending on the local problems, and care infrastructure. These collaborations support vulnerable client in society, by focusing not only on their mental health, but also on all other domains of life, including housing, finances, and participation in society. Possibly these collaborations might prevent escalation by early signaling in the home situation or staying in contact with family and friends and the availability of multiple leads for intervention.
Profile 2: Crisis after discharge clinic
This profile experienced severe crisis events and interference of PMHC was frequently necessary. CI’s were often preceded by increased crisis MHC contacts, official signals from the MHC and discharge from a psychiatric clinic. It seems that the subjects in this profile are not able to cope outside the clinic. Discharge is often due to withdrawn court authorizations. CI’s seem to be a tool for police, MHC and PMHC to obtain a court order but legal regulations often hampered this. The transition to compulsory admission is based on the degree of danger resulting from psychiatric symptoms. Often this was not an option and a proper alternative was lacking, resulting in new CI’s within days and creating a huge burden for the police and other parties involved, leaving clients very vulnerable. For some severely mentally ill persons, compulsory admission and long term clinical care seem the only solution within the treatment options at the time of the study. They are not able to cope outside a psychiatric clinic, with or without ambulatory care. A high number of previous admissions are a predictor for readmission within 6 months [
27]. In the light of the current trend to decrease the number of beds in the clinics and switch to more ambulatory care, the burden on and of patients in this profile might increase. New alternatives should be sought to provide sufficient support for patients in this profile outside the clinical setting, but with intensive outreaching guidance, strong motivational expertise, and possibilities for preventive coercion.
Profile 3: Crisis during unstable MHC
CI’s were often related to the marginal living situation and circumstances of these clients, due to drugs and/or alcohol dependence and/or homelessness. The need for care often arose at the (crisis) moment. Due to these circumstances, the care during a CI was not always focused on psychiatric problems, these problems were not directly visible. No appropriate care seems to exist for the clients in this profile who, in general, feel no need for care or even actively avoid care. Their sudden upcoming demand for care, that diminishes within days, renders mental health care without sustainable options, and the motivation of MHC and addiction care to offer short term solutions whenever the clients demands these, decrease each time a care trajectory or clinical episode is prematurely ended by the client. It is clearly difficult to achieve and maintain frequent contact within this profile of whom many are alcohol and/or drug dependent. Based on the available data, it is impossible to tell how much effort was put in to maintaining contacts. This study showed however that CI’s provide an opportunity to resume or start MHC care and contacts. For more than half of the individuals within this profile, increased ACT (after admission) resulted in improvement of the situation. ACT also reduces the likelihood of disengagement from services [
28]. Therefore it seems important to keep putting effort in reaching out to this profile of possible care avoiders, even though they constitute only a small proportion of all SMI-clients.
Strengths
By combining data from three large institutions (two MHC and PMHC) that provide care for SMI-patients in Amsterdam, we were able to provide longitudinal insight in patterns of care. Reviewing every patient journal in detail made it possible to extract much information about crisis interventions. This made it possible to determine profiles of a group of SMI-patients experiencing multiple CI’s. With these results, the PMHC and MHC can possibly identify cases and predict crisis situations.
Limitations
The data from the Safetynet system come from an administrative database. All details of psychiatric emergencies were made in free text by a large panel of professionals from the department and may not be as accurate and complete as those made with diagnostic research instruments and structured interviews. Second, although our data included a period of almost 9 years, this period did not coincide with the start of a care trajectory. This may have limited the insight in pathways of care. Also, the long follow-up period of the study inevitably also included changes in the organization of care, and the introduction of new interventions, such as Assertive Community Treatment (ACT). In addition, we may not have covered all MHC contacts. For example, the data did not include contacts with MHC outside the region, contact with private mental health practices, or contacts with prison mental health services. Finally, the study population focused on the subgroup of patients with multiple CI’s. A quantitative evaluation of the whole group is in progress and will provide additional insight into factors in which this problematic groups differs from a group in which less public interventions were necessary. Since it is an observational study, we provide insight in the current process in which MHC, PMHC and the police work closely together. However, we could not compare this to different situations in which one of the parties was absent or other options in interventions were available. It might therefore be very insightful to conduct similar analyses in other cities or countries with different systems as a comparison.
Leads for prevention
The study had two aims; to identify high risk groups and predict crisis situation(1) and contribute to the prevention of these situations (2). High risk groups seem most readily definable by developments in a patient’s situation over time instead of personal characteristics. These developments include changes in the availability of important persons in the social network, like the hospitalization or death of a parent; a change in housing situation, which includes discharge from a clinic; and personal deterioration or deterioration of the living environment. In order to use these signals for prevention, they need to be picked up and considered to be important by the involved professionals, who then need to have options to act. Especially this last step seems difficult, since it often involves people not motivated for voluntary options and the situation is not yet severe enough for involuntary options. Motivational techniques, involving the social network when available, therefore seem to be a key element in prevention of crises. In addition, patients might feel a need for support in other domains than mental health, such as housing or shelter or hygiene. When this support is provided by professionals with an integral approach, this might contribute to prevention as well. Outreaching care with unplanned house visits forms an important element in timely signaling and also helps in maintaining the relationship between patient and care professional in times of low motivation of the patient.
A few more specific leads for preventing crises became apparent. These are not new, but are confirmed by these results: intensive and immediate follow-up after discharge from a psychiatric or addiction clinic [
29], the value of signals of concerned family members, control of medication adherence. It also needs to be said that some crises cannot be predicted and prevented, underlining that a public response remains necessary.
Conclusions
The general conclusion is that the interference of parties in the public domain seems necessary for a small group of SMI-patients that are already known and in treatment by the MHC. The results showed that the police and PMHC play an important role in signaling crisis situations and referring individuals to the MHC, but also in referring clients to other types of care and support. Where public mental health care has an important role in picking up signals and referring to different organizations, the police has an important role in signaling problems in the public and domestic domain and in recognizing that these problems might be related to underlying mental or social problems. The base of this collaboration between MHC, PMHC and police could be further developed in order to tackle the complexity of the SMI-patients. Since many cases do not require acute psychiatric care, it seems that quick and effective triage is essential to decide which care system should be involved. Several problems emerge from this study, differing by subgroup. Medication non-adherence is a problem in all profiles, especially in the first profile. Continuity of care, including continuation of care after discharge from the clinic (aftercare) but also prolongation of clinical care are important issues in the second profile. Motivation, outreaching contact, and prevention of deterioration emerge as the most important issues from the third profile. Creative interventions and cooperations seem necessary to provide effective care for all three of these profiles of complex chronic patients, especially in the light of the current new wave of de-institutionalization in the Netherlands.
Competing interests
The authors work at the institutions under study. All institutions are represented among the authors and have contributed. The authors declare that they have no competing interests.
Authors’ contributions
MH has gathered and analyzed the data, and has written the manuscript. LdM has contributed to the analyses of the data and the design of the profiles, and has been involved in drafting the manuscript. JT, JD, MW and JZ have been involved in the conception of the study, and in revising the manuscript critically. JP has contributed to the collecting and analyses of the data, and in revising the manuscript critically. MdW has designed the study, contributed to the collection of the data and the design of the profiles and has been involved in drafting the manuscript. All authors have given final approval of this manuscript to be published.