Acute psychiatric care: approaches to increasing the range of services and improving access and quality of care

Abstract

Acute services for mental health crises are very important to service users and their supporters, and consume a substantial share of mental health resources in many countries. However, acute care is often unpopular and sometimes coercive, and the evidence on which models are best for patient experience and outcomes remains surprisingly limited, in part reflecting challenges in conducting studies with people in crisis. Evidence on best ap­proaches to initial assessment and immediate management is particularly lacking, but some innovative models involving extended assessment, brief interventions, and diversifying settings and strategies for providing support are potentially helpful. Acute wards continue to be central in the intensive treatment phase following a crisis, but new approaches need to be developed, evaluated and implemented to reducing coercion, addressing trauma, diversifying treatments and the inpatient workforce, and making decision‐making and care collaborative. Intensive home treatment services, acute day units, and community crisis services have supporting evidence in diverting some service users from hospital admission: a greater understanding of how best to implement them in a wide range of contexts and what works best for which service users would be valuable. Approaches to crisis management in the voluntary sector are more flexible and informal: such services have potential to complement and provide valuable learning for statutory sector services, especially for groups who tend to be underserved or disengaged. Such approaches often involve staff with personal experience of mental health crises, who have important potential roles in improving quality of acute care across sectors. Large gaps exist in many low‐ and middle‐income countries, fuelled by poor access to quality mental health care. Responses need to build on a foundation of existing community responses and contextually relevant evidence. The necessity of moving outside formal systems in low‐resource settings may lead to wider learning from locally embedded strategies.

Acute mental health care, including a­cute inpatient wards and services that man­age mental health crises in emergency de­partments and in the community, consumes a large proportion of the resources dedicated to mental health in many countries ¹ . However, it continues to be often unpopular, is sometimes experienced as traumatic or coercive, and shows little evidence of re­sulting in sustained improvements in outcomes.

Nonetheless, ready access to crisis response remains of high importance in the eyes of many service users, carers, clinicians and referrers to mental health services. Thus, innovations that result in better experiences and outcomes and more efficient use of resources have high potential for overall impact. In this paper, we take stock of current service models and their evidence base and identify innovations with promise for the future.

We begin by considering initial response to the acute crisis, including assessment, triage and initial care planning. We then dis­cuss the settings in which intensive inter­vention to resolve the crisis is delivered. Fi­nally, we offer some cross‐cutting perspectives on­ crisis care delivery, focusing on con­tributions from the voluntary sector; the role of service users and peer workers in designing, leading and delivering crisis services; remote delivery of crisis care; and crisis prevention.

Regarding geographical scope, it is not feasible to take a truly worldwide perspective on acute mental health care. However, while the majority of the authors of this pa­per are based in the UK, and thus tend to draw especially on examples from the Na­tional Health Service (NHS) of that country, we also include authors from several other countries, and conclude with a section that focuses on low‐ and middle‐income countries (LMICs) where specialized forms of crisis service are not present.

We focus primarily on services for adults of working age rather than on specialized models for children and adolescents or old­er adults. Distinct crisis services for these latter groups are relatively uncommon in most countries, and the extent to which the services we discuss in this paper also serve them varies greatly.

Service design and development should be rooted in evidence, and we would have preferred to focus primarily on interventions and service models for which evidence is robust. However, practical and ethical challenges in recruiting participants who are experiencing a mental health cri­sis have hampered research in this field ² , so that the evidence base is far from proportionate to the importance of acute men­tal health care. We therefore include not only approaches and models that are root­ed in evidence of reasonable quality, but also others that appear of sufficient poten­tial value for robust evaluation to be needed.

ASSESSMENT AND IMMEDIATE MANAGEMENT OF THE CRISIS

Mental health presentations in the emergency department of the general hospital

For many people experiencing an acute mental health problem, attending the emer­gency department (ED) of a local general hospital is the default option in a crisis ³ , and in some mental health systems primary care referrals may be directed to this setting. Despite efforts to develop alternatives, mental health presentations to the ED have been reported to be on the rise across the US ⁴ , Australia ⁵ and England ⁶ . Attendances are reported to have risen again following a dip during the early phases of the COVID‐19 pandemic ⁷ , ⁸ .

A review of evidence from seven countries ⁹ found that the most common mental health presentations to EDs are self‐harm, suicide attempt, suicide ideation, depression and schizophrenia, with mental health crises making up around 4% of all ED presentations.

Despite these high levels of use, EDs are often reported to be poor environments for mental health care. They tend to be hectic and may expose service users to long waiting times and distressing sights and sounds. Assessments take place in a very different and more institutional environment from service users’ usual social context, and ED assessment has been reported to be more likely to result in hospital admission than when similar crises are assessed elsewhere ¹⁰ .

ED staff may not have the training required for working effectively and empathically with people in mental health crisis ¹¹ . Negative attitudes towards people with mental health presentations have fre­quently been reported ¹² , especially towards those who present on multiple occasions following self‐harm and who may have a “personality disorder” diagnosis ¹³ .

The quality and volume of research investigating the effectiveness of different approaches to improving mental health assessment and treatment at EDs does not match what is needed. Challenges include the highly diverse nature of tasks undertaken in EDs, and more widely in general hospitals where a liaison psychiatry model is employed; lack of high‐quality routine data; difficulties with linking general hospital and mental health provider data sources; and difficulty selecting appropriate outcome measures to reflect brief contacts ¹⁴ .

An international systematic review of models for mental health care in EDs found just 17 relevant studies, relating only to Australia, Canada, UK and US ¹⁵ . Mental health staff may be integrated into the ED team, supporting it with patient assessment and triage. A psychiatric liaison service may work across the ED and the general hospital as a whole. Agreements of various forms may be established between the ED and a psychiatric service within the same hospital, so that the latter can provide input to ED patients on referral. Finally, as discussed further below, mental health EDs may be located away from the general hospital. A variety of benefits have been reported for these models, mostly related to service use measures such as waiting times, restraints, or unplanned departures from the ED department. Most studies do not include clinical or patient‐reported out­comes.

Whichever model is employed, a challenge in the ED is ensuring that, within the brief period of a crisis assessment, a warm and supportive therapeutic relationship is rapidly established, to avoid traumatic and coercive experiences of care and create a context for collaborative decision‐making about next steps ¹⁶ , ¹⁷ . More research focused on clinical communication, therapeutic relationship, and approaches to assessment in mental health crises in the ED would be valuable.

Models offering extended assessment and diversion following ED attendance

An international data synthesis found that studies varied greatly regarding proportion of ED attenders admitted to hospital ⁹ . Efforts to reduce this and to improve the quality of initial assessment following an ED attendance have resulted in service models that extend the period of mental health assessment in an environment intended to be more calming and conducive to good quality mental health care than the ED.

A range of such approaches has been de­veloped and described internationally. Psychiatric emergency services (PES; for which other names include comprehensive psychiatric emergency program, CPEP; and emergency psychiatric assessment, treatment and healing, EmPATH) are wide­spread in the US, where emergency psychiatry is a distinct subspecialty, and in Canada. They are linked to one or more EDs ¹⁸ and staffed by multidisciplinary psychiatric teams, including mental health nurses and psychiatrists (available on‐call if not on‐site), usually providing 24‐hour ac­cess.

Unlike the standard ED approach of triage and transfer, PES have extra capability to observe and provide intensive treatment, typically for a period of up to 24 hours, aiming to stabilize the crisis within this time and reduce the need for admission. Routine data on the impact of a PES serving a large area of California and linked to several EDs indicated that it substantially reduced both ED waiting times and admission rates ¹⁹ .

Similar models are reported in other countries. For example, in Australia, a behavioural assessment unit with six beds within an ED in Melbourne was designed to provide a calming environment, mental health assessment and observation, aiming to discharge home within 24 hours. A before‐after comparison indicated reductions in ED delays and restrictive interventions ²⁰ .

Psychiatric decision units have been es­tablished in a small number of centres in the UK ²¹ and are accessed via psychiatric liaison teams in the ED. They offer a stay of between 12 and 72 hours, providing re­clin­er chairs rather than beds (subject to some criticism ²² ) and aiming to ensure a calming environment, psychosocial assessment, brief interventions, and onward referrals. In general, although there are promising reports of impacts on service use, substantial evaluations of extended assessment and triage services ­following ED attendance are so far lacking, and impacts on patient experience need to be better understood.

A further model that may be linked to the ED is the brief admission ward where, rather than a full‐scale hospital admission, initial admission is to a ward in which intensive assessment and treatment planning takes place within a strict time limit, characteristically a few days. Several early trials of this model suggested rather modest benefits ²³ , although they were conducted in contexts where intensive crisis alternatives were generally unavailable. A more recent UK version of this model did not find an impact on length of stay ²⁴ , and we are not aware of substantial recent evaluations of triage or short stay wards linked to EDs or of a recent comprehensive literature synthesis.

Assessment centres outside the general hospital

Crisis assessment services may also be situated away from the general hospital in freestanding centres, within community mental health service premises, or co‐located with specialist psychiatric hospitals. Evidence is lacking regarding which locations are best and for whom. Notwithstanding the ED disadvantages discussed above, links between acute mental and physical health care are important (for example, following self‐harm, and for people with both physical and mental health problems, or who present with functional somatic and neurological symptoms).

Thus, even in mental health systems where referrals from primary care and self‐presentations are directed elsewhere, as in many European countries, mental health care is still needed in EDs. Integrating this with general hospital and mental health care systems effectively, and achieving continuity of care between acute and continuing care services, is a complex task presenting different challenges in each national system ²⁵ .

In the 1960s and 1970s, community mental health assessment centres, often called emergency clinics, were an important innovation in some countries, including the US and UK. These services provided walk‐in assessment, triage and sometimes brief treatment, often informed by the crisis intervention theory ²⁶ , which regards a crisis not as a manifestation of mental health prob­lems but as a general human response to se­vere psychosocial stressors, presenting challenges but also opportunities for growth. Similar models later emerged especially in the Netherlands, Italy ²⁷ , and Ger­man‐speaking countries, although investigation of their activities suggested that they tended not to focus on people with severe mental health problems ²⁸ , ²⁹ .

Today, there are numerous internation­al examples of mental health crisis assess­ment centres, some of which employ conventional models of clinical assessment and intervention not dissimilar to ED services, while others are more innovative in offering alternative models. The PES discussed above may be located away from general hospital premises, even though they retain close links with EDs. Such services may also be established to prevent people in crisis being referred directly for assessment to psychiatric wards, which has been observed to be associated with high rates of admission. In Switzerland, for example, establishing a unit for clinical decision‐making to assess referrals rather than referring directly to wards was reported to have reduced unnecessary admissions and costs ³⁰ .

Overcrowding in EDs and infection con­trol considerations during the COVID‐19 pandemic have resulted in some countries in further development of crisis assessment centres outside hospital. For example, a sur­vey in England found that men­tal health providers in 80% of areas had established an alternative to their local EDs for mental health assessments, most often on a site where other mental health services were delivered ³¹ . Psychiatrists reported that these often provided a better environment than EDs for mental health care, but had very limited capacity for providing physical health interventions. Concerns were raised that removing mental health professionals from EDs may increase stigma among acute hospital staff and negatively affect care for the many people with both physical and mental health problems. An Italian service system has been described ³² in which the community mental health centre, already used as a setting for some crisis assessment, shifted its focus towards greater crisis care provision during the pan­demic.

Crisis centres in the community may al­so aim to provide a more clearly distinct alternative to standard clinical approaches. For example, a model that has emerged in England over the past decade is the “crisis café”, sometimes referred to as “safe havens” or “sanctuaries” ³³ . These services provide walk‐in assessment, support and triage for people experiencing a mental health crisis. They are designed to provide a less formal and clinical environment, and are usually delivered by the voluntary sector with staff who do not have formal mental health professional qualifications, although they often have considerable rel­evant experience. Some are also staffed by peer support workers and a few are led by people with lived experience of mental health problems (e.g., the Well‐bean Crisis Café in Leeds, England). They are usually open outside typical working hours (evenings and weekends), when other forms of support may not be available, and are located separately from any other health service.

Crisis cafés provide a source of immediate support. People in crisis can usually access them without a referral, which may prevent a crisis escalating to a point where ED attendance or admission results. The potential of these services to improve access and choice is clear, but research evaluating their effectiveness and safety is still lacking.

Community crisis assessment

High anxiety, enervating depression or cognitive disorganization may all prevent some people in mental health crisis from actively seeking and accessing help. Perceived stigma of mental health services, or previous experience of unsatisfactory treatment following help‐seeking or of an unsympathetic response at hospital EDs ³⁴ , may also create barriers.

Assessment at home may be more feasible and less frightening or distressing for many. It enables evaluation of someone’s living situation, current coping, and potential risks in the home. It can help clinicians to consider social precipitants of a crisis, which may otherwise be overlooked ³⁵ . Home‐based assessment may engage the family from an early stage, helping clinicians to understand and manage a crisis ³⁶ . For these reasons, home‐based crisis assessment services have been developed as part of the community psychiatry movement, with “psychiatric first aid” multi‐disciplinary teams in the Netherlands in the 1930s ³⁷ , ³⁸ being an early example.

Community teams providing longer‐term care may be well placed to respond to crises for people on their caseload, allowing assessment by clinicians who already know the person in crisis. Indeed, providing a 24‐hour crisis response is a fidelity criterion for high‐intensity assertive community treatment (ACT) teams ³⁹ . Flexible, stepped care models have been developed internationally and can offer a prompt crisis response to new referrals, as well as longer‐term care of varying intensity, to meet people’s current needs. Two examples (for both of which a robust evidence base has yet to be established) are the German RECOVER programme ⁴⁰ and the FACT (flexible ACT) model developed in the Netherlands ⁴¹ . However, most community mental health services are not 24‐hour, or resourced or organized to respond rapidly to needs for crisis assessment across a whole community, including people not previously known to services.

Dedicated crisis resolution and home treatment teams (CRHTTs) have therefore been developed, with the sole function of providing assessment and short‐term, multi‐disciplinary home treatment for people during a mental health crisis. Pioneered in the US ⁴² and Australia ⁴³ , CRHTTs are now provided nationally in England and Norway, and in many areas across Europe, North America and Australasia ⁴⁴ . Established fidelity criteria for CRHTTs include standards for ease of referral, rapid response time, a 24/7 service, assertive engagement and comprehensive initial assessment ⁴⁵ .

Two key challenges for community crisis assessment relate to providing a rapid response, and managing safety and risks.

Regarding rapid response, in‐person assessment within four hours from referral has been adopted as a nationally audited performance indicator in England. Yet, a 2016 survey of CRHTTs in England found that target response times varied from one hour to one week, with less than half of teams routinely providing a response within four hours. Less than a third of Norwegian CRHTTs achieve good fidelity for the rapid response criterion ⁴⁶ . CRHTT staff highlight the competing pressures of responding rapidly to new referrals while reliably maintaining frequent, scheduled home treatment appointments with people being offered crisis support ⁴⁷ .

To address this issue, a recent trend in England has been to split crisis assessment and brief crisis home treatment functions into two different teams. This split model is now provided in over a third of English health care regions ³³ . Crisis assessment teams, sometimes called “first response” teams, have achieved marked improvements in service accessibility and response times in local evaluations ⁴⁸ , and offer a “no wrong door” point of access for people in mental health crisis of any severity. However, they risk introducing new discontinuities between assessment and treatment, with opportunities for information to be lost or people in crisis being required to tell their story multiple times to different professionals. As yet, no robust evidence compares effectiveness or users’ experience of integrated CRHTTs versus split assessment and treatment teams.

Regarding safety and risk, crisis assessment at home is not suitable when someone requires urgent medical tests or treatment (for example, following an overdose or other self‐harm). Escalating risks to the person in crisis or others may be harder to manage by lone clinicians in an unfamiliar home environment than in a clinical setting. A Cochrane review cautions that people with the highest risks or using drugs and alcohol were typically excluded from studies that have provided positive evaluations of CRHTTs ⁴⁹ .

Thorough information gathering and careful triage are therefore essential before home‐based assessment is offered. 24‐hour crisis phone lines staffed by trained clinicians, with links to other local or national health service helplines, may help to achieve this, and improve the accessibility of crisis support ³³ . Effective system integration with police and ambulance services is required for circumstances where the need for immediate access to hospital or clinic‐based care becomes apparent during a home assessment, and help from emergency services is necessary to ensure safe conveyance of the person. This is further discussed in the next section of this paper.

Practical measures to help ensure the safety of staff, such as a lone working policy with check‐in and follow‐up processes, alarms for staff, and team capacity to visit in pairs when indicated, are also recommended ⁴⁴ . Challenges are compounded in remote areas, and the role of telepsychiatry in crises is discussed further below.

Initiatives to facilitate prompt assessment following police contact

A 2016 literature review estimated that, for around one in ten individuals, the police were involved in their pathway to mental health care ⁵⁰ , although, while the author searched for all English language studies, only studies from North America were found. In a Canadian city, around half of mental health‐related police contacts resulted in apprehension using mental health legislation, and half of these led to a hospital admission ⁵¹ . Concerns have been reported around the world that police officers, without adequate training or support, are often acting as frontline mental health workers, potentially resulting in worse outcomes for people in mental health crisis, increased trauma and coercion, and higher numbers of unnecessary arrests ⁵² and escorts to hospital ⁵³ .

Various service models have been developed to improve outcomes for people in mental health crisis following contact with the police. They usually consist of police and mental health staff responding to mental health‐related emergency calls together. Some successes have been reported in reducing unnecessary use of mental health legislation. For example, in Toronto, Canada, a model involving additional training and a joint response by mental health nurses and police officers was found to result in lower rates of involuntary escorts to hospital and of arrest and injury, although total numbers of escorts to hospital increased ⁵⁴ .

In the UK, around 70% of NHS providers now have a street triage service involving various models of joint response by police and mental health professionals, ranging from telephone liaison to (in a few cases) 24‐hour joint response ⁴⁷ , ⁵⁵ . A systematic review of co‐response models found studies carried out in Australia, Canada, UK, and US ⁵⁶ . There were indications that these services reduced the use of police powers to detain people under mental health legislation, and of police cus­tody.

Feedback from both police officers and health staff working in street triage teams or similar models is generally positive ⁵⁵ , ⁵⁷ , but there has been a lack of research investigating service user experiences and outcomes ⁵⁶ . The research that does exist suggests that service users value responders with expertise in mental health and skills in de‐escalation ⁵⁴ .

There are many challenges in delivering joined‐up responses across different organizations with very different roles, and models which may lead to greater police involvement in management of mental health crises may prove unacceptable or have unintended negative consequences. For example, the Serenity Integrated Mentoring model (SIM), deployed in England by around half of NHS Trusts, is designed to be a concerted approach by mental health care services and the police to better supporting people who frequently use emergency services. Reports that it resulted in inappropriate diversion from health services and in approaches mainly based on enforcing boundaries have led to the #StopSIM coalition of service users campaigning against the model’s deployment, supported by allies across the mental health sector ⁵⁸ , ⁵⁹ , ⁶⁰ , following which policy makers have required Trusts urgently to review its further use. Much of the debate has focused on the ethics of police involvement and on its lack of underpinning evidence base, exemplifying the risks of rolling out mod­els that are not supported by robust evidence.

INTENSIVE TREATMENT FOLLOWING CRISIS

Management of crises in hospital

Despite their ubiquity in mental health care systems, there has been surprisingly little definition or discussion of the role, function and design of acute inpatient mental health wards. Bowers et al ⁶¹ provide a conceptual model of inpatient treatment. The primary admission tasks for inpatient care may include any or all of: assessment, treatment of acute illness, providing safe and highly tolerant accommodation, rehabilitation, and the resolution of personal stress.

Inpatient wards are uniquely able to enforce treatment, provide constant observation to contain risks, and tolerate behaviour which would be unmanageable or unacceptable in the community. Inpatient admission also offers respite from and space to address stressors in the person’s home environment, and the potential, through 24‐hour care, for providing high levels of interpersonal contact and therapeutic engagement ⁶¹ .

Thus, there is clearly a role for inpatient wards in managing and supporting those who are most acutely unwell at times when community services are unable to offer a safe alternative. Nonetheless, in the context of the narrative of deinstitutionalization, acute inpatient wards tend to be seen as an expensive legacy of a past institutionalized system of care, with admission reflecting a failure of care, rather than as unique and specialist clinical services playing an important role within a balanced mental health system ⁶² .

Internationally, bed provision is inevitably influenced by the national and regional configuration of mental health care systems ⁶³ . In general, across Europe, there are mental health care systems with predominantly community‐oriented approaches, such as those in the UK, Italy and Spain; areas with a high availability of community, residential and hospital services (mainly in Scandinavian countries); and areas where the deinstitutionalization process is still incomplete and inpatient services are the main source of care, such as in rural France, or where it is still in its very early stages, as in several Eastern European countries ⁶⁴ .

A recent study involving 22 high‐income countries in Europe, North America and Australasia found wide variation in the extent of inpatient provision: the mean number of beds per 100,000 population was 64, with an interquartile range of 46‐93 ⁶⁵ . Throughout Europe and elsewhere, psychiatric inpatient bed numbers have tended to decrease in recent decades, and this trend has been marked in some countries: for instance, bed numbers fell by 62% in England between 1988 and 2008 ⁶⁶ .

Much literature on inpatient care focuses on negative patient experiences and risks. Potential iatrogenic harms include institutionalization, exacerbation of psychotic symptoms from intense social contact with others, injury or victimization from other patients, loneliness due to separation from their home environment and social network, despair and depression arising from the environment and seeing other very unwell patients, and stigmatization ⁶¹ . Women are vulnerable to sexual harassment or assault, especially in mixed‐gender inpatient wards ⁶⁷ .

Evidence suggests that acute inpatient mental health wards are often unsafe, with high levels of intra‐ and international vari­ation in levels of conflict and containment ⁶⁸ , ⁶⁹ . During inpatient care, patients may experience high levels of restrictive practices (physical and mechanical restraint, forced medication); discrimination based on ethnicity, gender or diagnosis; crime (physical or sexual assault, criminal activity, drug taking); and blanket restrictions and rules. In England, the most frequently occurring incidents in this setting involve aggression and self‐harm ⁷⁰ .

Safety incidents are often associated with high physical, emotional and financial costs. The physical and psychological harm to the patient, which may increase length of stay as well as having a negative impact on health‐related quality of life ⁷¹ , is often underestimated even in those ser­vices which aspire to operate trauma‐in­formed models, in which an aim is to avoid retraumatizing the many patients who have previously experienced significant trauma ⁷² . In some cases, injuries to staff may also occur, leading to costs of replacement and impacts on burnout, stress and morale ⁷³ . The financial cost of restraint, seclusion, rapid tranquilization, and one‐to‐one nursing have not been examined in any depth. One incident on a ward may increase the likelihood of further incidents via a disturbed ward milieu and social contagion ⁷⁴ .

Negative service user and carer experiences of involuntary detention are frequently reported and are of particular concern, given the contrast between such detentions and the principles of collaboration and consent usually advocated as central underpinning values for mental health treatment ⁷⁵ , ⁷⁶ .

Rates of involuntary detention in psychiatric hospitals under mental health legislation have risen in some high‐income countries and fallen in others in recent decades ⁶⁵ . Explanations of why this is oc­curring remain confused. A complex com­bination of societal, service‐related and legal factors is probably implicated ⁶⁵ . Evidence regarding the relationship of bed numbers and availability to detention rates is mixed and inconclusive ⁷⁷ ; however, in countries where the drive to cut inpatient beds has been strong, there are widespread concerns and perceptions that lack of bed availability has resulted in higher thresholds for admission to hospital, a greater likelihood that those who are admitted will be involuntarily detained, a higher concentration on wards of people who are very acutely unwell and whose needs are complex, and a disturbed ward milieu. These factors combine to create high risks of iatrogenic harm. Detention also tends to establish a pattern of increased risk of future detentions ⁷⁸ .

Inpatient admission offers rapid access to needed medication, intensive monitoring and assessment to inform medication review, and enforcement of treatment if required – all of which may be problematic in community care ⁶¹ . However, prescribing practices are reported in many settings as relying too heavily on high‐dose medications, polypharmacy and supplementary as‐required doses ⁷⁹ , and there is a dearth of evidence on effective non‐pharmacological approaches to managing acute illness and violent behaviour ⁸⁰ . A literature on cognitive‐behavioural interventions for psychosis adapted to inpatient settings is beginning to develop and provides exam­ples of feasible approaches for people with complex needs, but does not yet offer con­clusive evidence to underpin a large scale transformation ⁸¹ . Moreover, there is a strik­ing lack of good quality evidence to underpin inpatient care for people with a “personality disorder” diagnosis.

Recent years have seen the development of interventions designed specifically to reduce conflict and use of restrictive practices in inpatient wards. A recent systematic review ⁸² identified two programmes with trial evidence of effectiveness, Safewards ⁸³ and Six Core Strategies ⁸⁴ , both of which now commonly inform practice ⁸⁵ . These are multi‐component team‐level interventions, which target avoiding or mitigating potential flashpoint situations resulting from interactions between patients, staff‐patient interactions, or the ward regulatory or physical environment. The need to improve therapeutic engagement and the culture of care on wards more generally has also been emphasized ⁸⁶ .

An umbrella review of interventions to reduce coercion in mental health services concluded that there is supporting evidence for staff training interventions ⁸⁷ . However, evidence for initiatives which have tried to improve the therapeutic quality of wards, such as scheduling protected time for ward staff to engage with patients, has tended to be inconclusive. Boredom is identified as a common problem for patients on inpatient wards, but further empirical evidence is needed about its impacts and the best ways to address it ⁸⁸ .

Another area where practice varies internationally and where evidence to support best solutions is lacking is the location of wards. In some countries, embedding acute wards in general hospitals is seen as advantageous, offering close links with physical health care services, normalization of mental health and accessibility to local communities ⁸⁹ . However, potential drawbacks include wards that have not been specifically designed for mental health patients, and lack of access to safe open space.

There is a need for better understanding of how to design healing environments that offer private space, light, access to fresh air, and attention to details relevant to recovery (e.g., making the environments autism‐friendly) ⁹⁰ . The identification and international dissemination of examples of good practice would be very valuable, as the nature and probably the quality of ward environments varies greatly between countries. Other questions that have yet to be fully addressed include the value of specialized wards based on diagnosis or other indicators of need, and separation by gender ⁹¹ .

Staffing is a further area in which there is scope for innovation to improve care. The staffing of wards remains a nurse’s domain, largely providing the 24/7 care for inpatients. The approach to staffing is often constrained by budgets and custom rather than evidence, and we lack high quality research regarding safe staffing levels or optimal skill mix on inpatient wards. Clinical decision‐making still tends to be dominated in most settings by psychiatrists, often via a traditional ward round model. More extensive involvement of other multidisciplinary team members such as psychologists and occupational therapists has great potential to enrich both decision‐making and therapeutic environments and activities, though limited size of the specialist health professional workforce may constrain this ⁹² . The opportunity to further enrich the skill mix by enabling the roles of peer support workers, mental health advocates, housing officers and social workers could help heal disconnections from the community and address those key issues which precipitate and prolong admissions, such as social isolation, poverty and poor housing.

The future of acute inpatient provision requires serious attention. Services can improve, and listening to the patient voice is key to this ⁸⁶ , ⁹³ . There is a broader need to listen to those voices marginalized as a result of gender, ethnicity or diagnosis, including those labelled with “borderline personality disorder”, who may be at most risk of receiving a poor service ⁹⁴ . Achieving high quality community care and supporting people outside hospital is rightly a policy priority internationally, but it is vital that this is accompanied by sustained efforts to re‐design and improve the provision of care in acute inpatient settings, rebalancing multidisciplinary teams, listening to service user voices and investing in interventions that demonstrate improvements in patient outcomes.

Home treatment

Early crisis home treatment programmes formed part of a broader deinstitutionalization movement, seeking to minimize stigma and normalize mental health crises. In this section we discuss intensive treatment at home. We note that in many systems the same teams are offering both crisis home assessment (discussed above) and intensive home treatment.

Treatment at home from CRHTTs may reduce the perceived stigma and coercion associated with hospitalization. Because it requires negotiation and takes place on the territory of the person in crisis, it potentially reduces power imbalances and respects people’s autonomy ⁹⁵ . It may encourage a greater focus on interpersonal issues and involvement of the family and wider support system ³⁴ , ⁹⁶ . It may also avoid difficulties of transferring coping strategies and skills learnt in a hospital setting to a home environment ⁴¹ .

A Cochrane Collaboration review of com­munity crisis intervention for people with severe mental illness ⁴⁹ included six trials of CRHTT‐style services (and two residential community crisis services). It found evidence that CRHTTs can reduce inpatient service use, improve clinical outcomes and patients’ experience of care, and reduce costs. Observational studies similarly suggest that the introduction of CRHTTs in a local area can help reduce overall men­tal health inpatient admissions when well‐implemented ⁹⁷ . A qualification to this prom­ising evidence base is that crisis home treatment will not be suitable for people with the highest risks to self or others, and CRHTTs have not demonstrated effectiveness in averting involuntary hospital admissions ⁹⁸ .

CRHTTs do not originate from a highly specified theoretical model. Key characteristics of model services have included: a multi‐disciplinary team; 24/7 availability and a rapid response to crises; intensive short‐term home‐based treatment (typically of less than six‐week duration and with visits more than once a day); collaboration with families and other involved services; working with people in crisis who would otherwise be admitted to hospital, and facilitating early discharge from hospital for those who are admitted ⁴³ . There is some empirical evidence that having a psychiatrist in the team and extended opening hours are related to CRHTT effectiveness ⁹⁹ . A more highly specified CRHTT model and an accompanying fidelity scale have been developed ⁴⁴ , with fidelity scores shown to relate to inpatient admission rates and satisfaction with care ¹⁰⁰ , but the relative importance of individual fidelity criteria and the critical ingredients of CRHTTs have yet to be established.

Implementation of the CRHTT model has proved challenging. Model fidelity is typically low or moderate in CRHTTs in England and Norway – the two countries where it has been scaled up nationally ⁴⁵ , ¹⁰¹ . Criticisms from service users and families have included poor continuity of care within CRHTT team‐working, a narrow therapeutic focus on risk and medication (with a corresponding lack of other meaningful therapeutic interventions), and lack of support for or involvement of families ³³ , ⁹⁹ , ¹⁰² , ¹⁰³ . CRHTT staff have highlighted difficulties in establishing role clarity for CRHTTs across the mental health system, and in joint working with inpatient services and longer‐term community care teams ⁴⁶ .

Three initiatives may offer helpful ways to address some of these difficulties and improve the effectiveness of CRHTTs. First, a UK trial ¹⁰⁴ showed that a service improvement programme for CRHTTs over one year, involving coaching from a senior clinician, regular fidelity assessment, and access to an online bank of practice resources, increased model fidelity and led to reductions in inpatient admissions and bed use. Second, a recent Swiss trial ¹⁰⁵ reported that a CRHTT was able to reduce inpatient bed use, despite focusing almost exclusively on facilitating prompt hospital discharge rather than preventing admissions, which shows the importance of working closely with inpatient wards to end inpatient stays as soon as home treatment becomes a viable alternative. Third, a number of models for enhancing the involvement of families in acute mental health care have been developed, which typically include a focus on communication, language use and joint decision making ¹⁰⁶ .

Most attention internationally has been given to the open dialogue approach (ODA). ODA is a model of crisis and continuing care characterized by a rapid response to a crisis presentation, care centred around regular meetings of the whole support network of the person in crisis; and a psychologically informed approach to care facilitated by clinicians trained in family therapy. Three evaluations of ODA in Finland have reported promising findings ¹⁰⁷ , although robust trial evidence for effectiveness and transferability to other health care contexts has yet to be provided. A randomized controlled trial of an adapted ODA approach within a contemporary CRHTT context is currently in progress in England ¹⁰⁸ .

Both crisis assessment and intensive home treatment are in some service systems undertaken as functions within community mental health teams that also provide longer‐term care ¹⁰⁹ , ¹¹⁰ . This has advantages for continuity of care and therapeutic relationships. However, community teams also providing a range of other functions may struggle to deliver sufficiently intensive support and may not be well‐placed to work with people not already on their caseloads.

Treatment at home may not be helpful for people who are extremely socially isolated, or for whom tensions or abusive relationships with others in the household are contributing to the crisis, or when other household members require respite from their caring roles. “Family sponsor homes” – short‐term crisis placements with host families, who are trained and supported by mental health teams – have been established in the US and England ³³ , although practical and legal challenges have limited the implementation of this model internationally.

Acute day units

Acute day units (ADUs) typically offer programmes combining therapies, activities and social contact to people experiencing mental health crises who are close to the threshold for admission and attend several times a week for a number of weeks. Traditional names include day hospital or partial hospitalization service, but the more recent use of terms such as ADU or recovery centre reflects a concern that the term “day hospital” may have unduly institutional connotations ¹¹¹ .

The history of ADUs extends over most of the last century, with Moscow in the early 1930s sometimes identified as their birthplace, prominent models established around Europe and the US before and after the Second World War, and provision expanding rapidly in many countries between the 1950s and the 1980s ¹¹² .

The evidence base for ADUs is arguably the most robust for any admission alternative. The authors of a Cochrane review concluded that around one in five of those otherwise admitted to an acute psychiatric ward could successfully be treated in an ADU setting, with similar clinical and social outcomes ¹¹³ . The most recent UK trial showed greater service satisfaction and symptom improvement for ADU service users ¹¹⁴ , but new trial evidence has been lacking worldwide over the past 15 years, so that it cannot be assumed that such findings would be replicated in contemporary service systems which tend to have high thresholds for hospital admission and other approaches, such as CRHTTs, providing alternatives to admission. However, a recent naturalistic study compared outcomes for ADU and CRHTT care, finding greater service satisfaction and better outcomes for depression and well‐being for the ADUs ¹¹⁵ .

Despite the robust underpinning evidence, a decline of ADU provision has been documented in the UK ¹¹⁶ , and may have accelerated during the COVID‐19 pandemic, while little new evidence has been published elsewhere in the world. Reasons for this may include a perception that the model is unduly institutional, the substantial premises required to support a comparatively small number of service users, and the rise of other admission alternatives.

Care of an ADU form may also be inte­grated into community mental health cen­tres, where these are central to service provi­sion. However, qualitative work as well as trial evidence suggests some specific advan­tages which may not be shared by other admission alternatives: ADUs have important potential to address loneliness, social isolation, and lack of purposeful activity, and are also a potential environment for fostering both formal and informal peer support ¹¹⁷ . Evidence for the importance of social connection, sense of belonging and peer support in mental health recovery is growing, and purposeful activity also has established significance for recovery. A resurgence of the ADU as the principal acute service in which these elements are a central focus would thus be timely.

Residential community crisis services

Like ADUs, crisis houses and other community residential alternatives to hospital admission have a history spanning many decades. They are characteristically services allowing a short stay of a few days to a few weeks, with 24‐hour staffing and therapeutic programmes that range from relatively clinical services aiming to replicate the interventions delivered in hospital in a less coercive and institutional setting, to more radical alternatives aiming to support different ways of resolving crises and to enhance service user choice ¹¹⁸ .

An early US example was Soteria House in California, which from 1971 to 1983 aimed to manage first and second episodes of psy­chosis with minimal medication in a community setting, with some reported evidence of success ¹¹⁹ . Subsequently, crisis houses have been described around the world in a variety of formats. In the UK, provision has been growing in recent years, with just over half of catchment areas having some access to crisis house provision in 2019 ³³ .

The evidence underpinning the crisis house model is substantial, though not conclusive. Relatively few randomized controlled trials have been reported, reflecting the challenges of conducting such trials with people in crisis ² . A systematic review ²³ included five randomized trials and 11 non‐randomized studies of community residential alternatives to admission. Services were diverse in theoretical model, content and workforce, and included 11 US, two UK and two Swiss studies. Summary conclusions were that, according to the limited available evidence, community residential alternatives show similar, or in a few cases better, clinical outcomes to hospitals, with similar or lower costs and greater service user satisfaction.

A subsequent US review ¹²⁰ included “subacute” services, not necessarily 24‐hour staffed but available for urgent admis­sion with the aim of averting crisis. Equivalent or better clinical outcomes and great­er user satisfaction were reported compar­ed to acute wards, with lower costs also found in some studies.

Throughout this literature, the authors note that community acute residential services support a population overlapping with, but not the same as, acute wards, often excluding people who are assessed as posing a substantial risk of violence or who have been compulsorily detained ¹¹⁸ . We are aware of no randomized controlled trial of community residential alternatives to hospital in the past 10 years.

Positive reports regarding service user experiences, therapeutic relationships, and the availability of non‐standard therapeutic models are prominent in the literature on crisis houses ¹²¹ , ¹²² , ¹²³ , ¹²⁴ . This, together with evidence of satisfactory outcomes and simi­lar or lower costs compared to inpatient care, provides a justification for including community residential alternatives to inpatient acute care as a standard part of the range of services in any mental health system where choice, flexibility and cost‐effectiveness are prioritized. Despite this, we are not aware of any countries where inclusion of crisis houses is a standard element in acute care, although the model is found in many countries.

The literature on residential community crisis services suggests that the models implemented are diverse ¹¹⁸ . While this is an impediment to drawing generalizable conclusions about their outcomes, it is a potential strength in developing a flexible crisis care system in which a range of needs are met. Needs vary greatly at the time of a mental health crisis: for example, a service user beginning to take medication following a relapse of psychosis or bipolar disorder may benefit from a crisis house that incorporates some clinical professionals and approaches, while someone experiencing escalating distress and risk of self‐harm in the context of complex trauma and/or a “personality disorder” diagnosis may benefit more from a less clinical approach, in which relational care, psychotherapeutic approaches to trauma and complex emotional needs, and the support of peers might be the main elements. An optimized crisis care system might thus include multiple residential alternatives offering a choice of approaches to service users and referring clinicians.

FURTHER PERSPECTIVES ON CRISIS CARE

Crisis prevention

Our primary focus in this paper is on the management of mental health crises. However, the best option is clearly to prevent such crises if at all possible, investing instead on maintaining good mental health and supporting recovery in the community ¹²⁵ . A rapid evidence synthesis found that several interventions recommended by the UK National Institute for Health and Care Excellence (NICE) guidelines have some supporting evidence regarding prevention of crises and/or relapses of illness ¹²⁶ . These include early intervention services for psychosis, intensive case management models, and a range of pharmacological and psychological interventions for psychosis and bipolar disorder. Investing in full implementation of such models has potential to reduce crisis care use. Beyond such clinical models, social stressors and adverse social circumstances are contributors to crises, and a comprehensive programme to reduce adversity and inequality, as well as to implement interventions for severe mental illness that are clearly evidence‐based, is arguably the optimal approach to crisis prevention ¹²⁵ .

A wide range of approaches focus directly on preventing crises, including early warning signs monitoring and relapse prevention programmes, some in digital form, collaborative crisis plans, and advance statements or directives. Supported self‐management, often incorporating relapse prevention, is a straightforward intervention that shows evidence of effects on a range of clinical and social outcomes ¹²⁷ , so that wide implementation appears desirable in an optimized mental health system. The time following a crisis is an obvious target for delivery of interventions to prevent further crises: a large trial of a supported self‐management intervention delivered by peer support workers in sites around England found that it reduced repeat use of acute services ¹²⁸ .

Collaborative planning for what should happen at the time of a crisis is currently the intervention that appears most effective in preventing compulsory hospital admission, the form of acute care that it is most desirable to avoid ⁹⁸ . Ideally, as advocated in the Independent Review of the Mental Health Act in England, this should include advance statements that have legal force regarding what should happen when compulsory admission is contemplated ¹²⁹ .

The role of the voluntary sector

In many high‐income countries, the voluntary sector (including charities and community and service user groups) is increasingly playing a role in the provision of mental health support, valued for the distinctive approaches it offers and its greater focus on equalities.

Factors accelerating the contribution of the voluntary sector to crisis support include: a) recognition that the restricted focus of statutory acute mental health care results in people falling through the gaps in provision ¹³⁰ ; b) service user dissatisfaction with crisis support provided by secondary mental health services ¹³¹ , ¹³² ; and c) disproportionately high rates of involuntary detention for people from some minority communities, and concern that their needs are not well addressed by statutory services ¹³³ .

The distinctive contribution of voluntary sector services results from the way they work, whom they work with, and their roles within local communities ¹³⁴ , ¹³⁵ , ¹³⁶ . Their foundations are often in grassroots organ­izations and activism, and they tend to be “underpinned by an ethos of informal­ity, promoting accessibility, using relational‐based approaches, and valuing self‐organization and service‐user‐defined outcomes” ¹³⁰ . Hierarchies are often flat, and service user, volunteer and staff roles may overlap. They are thus potentially better placed to meet the needs of marginalized groups, and of those who are either unable to access or mistrust mainstream health services, such as people from racialized communities ¹³⁷ , homeless people, or those excluded because of complexity of difficulties or diagnoses such as “borderline personality disorder” (although coverage of marginalized communities may be uneven).

For example, Hutchinson et al ¹³⁸ found that men using not‐for‐profit mental health services in London were more often unemployed and had more unmet needs than local users of public mental health services. Those using the voluntary sector service cited wanting to escape “the system”, with levels of dissatisfaction with public sector mental health services reported to be particularly high among Black Caribbean participants.

Among the models discussed above, cri­sis cafés/safe havens and crisis houses have developed predominantly in the not‐for‐profit sector. Distinctive characteristics of their intended approaches ¹³⁰ , ¹³⁹ can include: a positive stance on mental health; a holistic understanding of crises that locates them in the biographical, social and relational context of people’s lives; space and time for people to speak about their distress; a safe, calm and welcoming environment and relational safety; informality and a light touch in terms of assessment and note‐keeping; greater autonomy, choice and responsibility for clients; strong therapeutic and peer relationships; enabling people to maintain their connections to “normal life” and the community; and a less stigmatizing and less clinical approach, with providers of care including peer support workers and volunteers embedded in local communities.

Types of help offered by such crisis services include emotional support and individual and group therapy; peer support and mentoring; social and therapeutic activities; programmes to better manage mental health; advocacy; and liaison with and signposting to both public sector and other not‐for‐profit organizations. Thus, mental health crisis management often sits alongside services that can support recovery and enable people to deal with financial, housing and social issues.

As well as these specific crisis support services, many other not‐for‐profit organizations play a role in crisis support, crisis prevention, recovery, and addressing inequalities in access and support. These include those supporting particular groups at risk of poor mental health – for example, members of the lesbian, gay, bisexual, transgender and queer (LGBTQ) community, those who are deaf, communities from specific ethnic or refugee backgrounds – and those responding to life crises such as bereavement, rape or homelessness ¹³⁰ .

The research literature on the contribu­tion of not‐for‐profit and community or­ganizations remains relatively scant internationally, and stronger evidence regarding their roles in local systems, experiences and outcomes would be very valuable. Reported advantages suggest that approaches developed in some not‐for‐profit crisis services have potential to address the problems with accessibility, acceptability, equality, and appropriateness to specific communities often reported in public mental health services ³³ , ¹²⁵ . A case can thus be made both that this sector should be recognized and incorporated within a comprehensive crisis system, and that it provides a model for rethinking dominant models of crisis care to ensure a response that is accessible, acceptable and appropriate for all members of the local population ¹³⁰ .

The contribution of service user‐led and co‐produced initiatives, and of peer support

Change to crisis and acute services has been a consistent focus for action in the mental health service user (or consumer) movement for many decades ¹⁴⁰ . In the 1970s, activists in the UK demanded rights‐based reform of the conditions and treatment in psychiatric hospitals ¹⁴¹ . Later, in the context of “community care”, user‐led organizations established themselves as sources of mutual support, patient advocacy and forums for campaigning and involvement work ¹⁴² . Informal peer support naturally occurred when people with mental health problems came together, and mental health service user groups went on to develop more organized forms of peer support, including for people experiencing mental health crises and acute distress ¹⁴³ .

Since their inception in grassroots service user groups, organized versions of one‐to‐one and group peer support have become influential for crisis and acute services across the UK, US, Canada, New Zealand and Australia ¹⁴⁴ . For example, “intentional peer support” defines crisis as “emotional and psychological pain” and peer support as being with another who has experienced similar pain in a relationship of trust and “mutual empowerment” ¹⁴⁵ . This model has been introduced into acute inpatient environments in the UK ¹⁴⁶ , and small‐scale qualitative studies show that patients can find it helpful in providing person‐centred emotional and practical support and in modelling hope ¹⁴⁷ . Research into the implementation and effectiveness of peer support in crisis and acute services is ongoing globally ¹⁴⁸ and, while some study findings on discharge and readmission to acute care seem promising ¹²⁸ , a robust evidence base is still needed ¹⁴⁹ , ¹⁵⁰ .

As originally conceived, peer support is rooted in a set of values and principles ¹⁴⁴ which can sometimes conflict with clinical environments and treatments associated with acute services, such as seclusion and restraint ¹⁵¹ . Mental health service us­ers, their organizations and allies have worked to establish a set of principles and principles‐based approaches for delivering peer support services in mainstream mental health services, including inpatient and crisis care ¹⁵² . Recent research into the formalization of peer support in UK mental health services suggests that “we need to pay attention to the values underpinning peer support… [and] to resist the replication… of a para‐clinical model of peer support” ¹⁵³ , whereby peer support workers become just another kind of non‐professional staff making up numbers in clinical teams. Some are concerned that the professionalization of peer support could undermine its values and authentic practice, and might negatively affect user‐led and community groups that have established their own forms of crisis peer support outside the psychiatric system ¹⁵⁴ .

An international consortium of peer support leaders agreed that present and future peer support innovations should adhere to values and principles rooted in maintaining “role integrity”, and in civil rights, social justice, and responsiveness to local cultural world views ¹⁵⁵ . These principles should apply whether crisis services are located within public mental health systems (such as Open Dialogue ¹⁵⁶ ) or beyond them in independent user‐led projects, such as the Leeds Survivor Led Crisis Service (LSLCS).

LSLCS is notable as an independent organization offering an alternative to hospitalization and statutory crisis care underpinned by principles and values of peer support ¹⁵⁷ . A social return on investment (SROI) analysis for the service estimated that the “SROI ratio for LSLCS lies within the range of £4.00 to £6.50 of social value generated for every £1 invested” ¹⁵⁸ .

The future challenge is to sustain and develop a diversity of values‐based, innovative and responsive peer support services for people in crisis and acute states. This is likely to expand further into the digital and online space for crisis prevention and recovery support ¹⁵⁹ . Research into implementation, development and effectiveness using a range of methodologies is needed to ensure that a robust evidence base is built on current and emerging forms of peer support, both within and beyond mainstream services.

Other essential considerations for service planning in the future include the benefits of a co‐production approach and of service user leadership. Given frequently negative service user views regarding mainstream acute services, such approaches have potentially much to offer across the acute care system.

Remote acute care delivery

Most literature on telepsychiatry focuses on videoconferencing, seen as the preferred substitute for in‐person interactions, but rapid and wide accessibility suggests that there is a significant role for telephone support in crises. Voluntary sector organizations have a long history of providing such mental health support, and have been found to deal with suicidal callers as effectively as professionals ¹⁶⁰ . The use of mental health hotlines has increased greatly in the early stages of the COVID‐19 pandemic ¹⁶¹ . Telephone services may also be used in secondary mental health care as an initial contact, support and triage point: for example, all NHS Trusts in England are now required to provide a local helpline ¹⁶² .

Telepsychiatry, predominantly using videoconferencing tools, has been used for decades to overcome geographical barriers to specialized care, particularly in rural parts of Australia and Canada, and some parts of the US ¹⁶³ , ¹⁶⁴ , ¹⁶⁵ . The adoption of these services has expanded to the crisis setting to provide urgent and emergent consultation, informing care management and decisions regarding transfer to hospital ¹⁶⁶ . For example, the Mental Health Emergency Care ‐ Rural Access Program provides telephone and video triage and assessment for emergent psychiatric presentations across Western Australia ¹⁶⁷ , ¹⁶⁸ .

Urban emergency settings characterized by variations in psychiatric coverage can also be served by a telepsychiatry liaison model. Such models have shown promise in the US and Canada to increase access to consultation, reduce wait times, decrease system costs, and improve post‐ED visit outcomes ¹⁶⁹ , ¹⁷⁰ , ¹⁷¹ . Evidence indicates that a trained team following comprehensive safety protocols can reliably assess a wide range of presentations remotely ¹⁷² , ¹⁷³ . This includes the assessment of suicidal behaviour, psychosis, affective symptoms, and substance use.

Virtual care is expanding rapidly, including web‐based programmes and apps with potential usefulness in crisis settings. Patient‐directed apps designed to help individuals cope during crises can be provided at the point of care to support post‐crisis self‐management and safety planning ¹⁷⁴ . Personal videoconferencing is now emerging as a viable modality of direct care delivery, removing the need for a traditional telehealth suite and allowing assessments to take place with individuals remaining in their homes or other accessible settings. As a result, some centres are innovating and pushing the usual boundaries for crisis care delivery ¹⁷⁵ , and virtual hospital‐at‐home models may become a significant format for acute care in the future ¹⁷⁶ .

However, significant barriers to scaling up telemental health effectively include remuneration models, digital exclusion, inadequate privacy in many service users’ homes, and perceptions that quality of care and therapeutic relationships are impaired ¹⁷⁷ . Rigorous research is thus needed to inform future development of remote crisis care within specific health care systems ¹⁷⁸ , ¹⁷⁹ , ¹⁸⁰ .

Crisis care in low‐ and middle‐income countries (LMICs)

In many LMICs, as well as in under‐served areas of high‐income countries, health services are often not the first port of call for individuals in crisis and their families. This is partly due to the limited availability and poor accessibility of mental health care. The average number of psy­chiatrists per one million population ranges between 0.6 in low‐income countries to 20 in upper middle‐income countries ¹ . Even with efforts to expand access to care through integration in primary health care ¹ , service coverage in LMICs remains low, with only 14‐22% of individuals who meet the criteria for a mental disorder receiving treatment ¹⁸¹ . Past experiences of poor‐quality or coercive care that fails to meet prioritized needs may also deter help‐seek­ing ¹⁸² . Only 44‐50% of countries in Africa and Southeast Asia have legal protections for people requiring crisis mental health care ¹ , and there may be minimal enforcement.

Low community awareness about mental health, high levels of mental health stigma and, in some countries, a preference for religious and traditional healers contribute further to low levels of help‐seeking from formal services ¹⁸³ . In this section, we focus principally on those countries where specialized mental health services other than large psychiatric hospitals are not available, applicable to most low‐income and some middle‐income countries.

Crisis presentations are often not framed as mental health problems in LMICs. Community responses to mental health crises may focus on overt manifestations of a problem, including acute behavioural disturbance or distress, suicidal behaviour and self‐harm, severe physical consequences (e.g., dehydration in severe depression or exhaustion linked to mania), and sudden loss of sensory or motor functions as part of conversion disorder ¹⁸⁴ . Non‐overt indicators of a mental health crisis, such as suicidal ideation, may not be prioritized for intervention.

An individual’s family often drives the response to a mental health crisis, drawing on informal support from communities. Responses to acute behavioural disturbance could include involvement of the police or religious or traditional heal­ers ¹⁸⁵ , complementary or homeopathic reme­dies, abandonment of the individual to the streets ¹⁸⁶ , some form of restraint ¹⁸⁷ , or emer­gency presentation to psychiatric services. Involvement of the police places the indi­vidual at risk of exposure to physical abuse, excessive force, restraints and detention ¹⁸⁸ . Restraint in the context of families is often seen as a last resort in the absence of accessible and effective care ¹⁸⁹ .

Stigma and taboos associated with self‐harm and suicidal behaviour may result in family concealment or punishment of the individual. Physical treatment for consequences of self‐harm or suicide attempts is not usually accompanied by any form of mental health assessment or treatment.

Community responses may frame acute distress in terms of a spiritual crisis or as the understandable consequence of severe social adversities (e.g., intimate partner violence, an acute life stressor) and mobilize resources accordingly. These responses may include mediation of relationship difficulties, material supports, or providing meaning to adversity ¹⁹⁰ .

A 2015 systematic review of mental health interventions for crises in non‐specialist settings in LMICs found a lack of evidence‐based guidelines for crisis care ¹⁸⁴ . Only one intervention study was identified. In a recently published guidance, the World Health Organization (WHO) set out recommendations for rights‐based, recovery‐oriented responses to mental health crises ¹⁹¹ . In developing the guidance, the WHO sought to identify case studies of good practice that were compliant with the 2006 United Nations Convention on the Rights of Persons with Disabilities, meeting five criteria (use of non‐coercive practices, community inclusion, participation in care, recovery approach, respect for legal capacity). Identifying good practice case studies from LMICs was a priority, but none was found.

An integrated mental health response to crisis presentations is rare in many LMICs. Referral to specialist mental health services may occur, but cost, inaccessibility and non‐acceptability are potent barriers to uptake. Involvement of people with mental health conditions in decisions about crisis care is very limited ¹⁸² . Consequences of the existing responses include violations of human rights, prolongation of severe mental illness linked to heightened vulnerability and poorer prognosis, risk of acute physical ill‐health and premature mortality, and more coercive mental health care (if accessed at all).

The WHO mental health Gap Action Programme (mhGAP) includes an intervention guide (mhGAP‐IG) comprising evidence‐informed algorithms for the provision of crisis care for acute psychosis or mania, suicidal behaviour or self‐harm, as well as acute behavioural disturbance in the context of dementia or developmental disorders ¹⁹² . However, it does not provide clear guidance on key components of rights‐based care (including supported decision‐making, informed consent for treatment, and non‐coercive practices) and evaluation for people with crisis presentations has been limited ¹⁹³ .

There have been small‐scale efforts to provide alternatives to hospitalization for people in acute crisis in Somaliland ¹⁹⁴ and Jamaica ¹⁹⁵ , but these models of care have not been rigorously evaluated and have limited potential for scalability, due to reliance on specialist mental health professionals. An adapted form of the crisis in­tervention team model, used widely in the US, has been piloted with law enforcement officers in Liberia, with preliminary evidence of beneficial impacts on knowledge, stigmatizing attitudes, and engagement with mental health clinicians ¹⁸⁸ .

To date, there have been two randomized controlled trials of crisis interventions for people presenting to non‐specialist services after suicide attempts in LMICs ¹⁹⁶ , ¹⁹⁷ . Both trials evaluated the brief intervention and contact model, comprising an initial one‐hour psychoeducation session at the time of the attempt, followed by nine phone calls over the next 18 months which assessed suicidality and support needs. The larger, multi‐country trial (Brazil, China, India, Iran and Sri Lanka) demonstrated an impact of the intervention on repeat self‐harm attempts and suicide, whereas the single country study (French Polynesia) showed no impact ¹⁹⁷ .

For the future, improving crisis response in LMICs will require the development and evaluation of contextually appropriate interventions, building on existing community resources and enabling community members to identify and support those in acute crisis, alongside strengthened access to mental health care and changes to policy and legislation. Building on community resources and equipping accessible individuals (e.g., peers, family members, community health workers, traditional and religious leaders, community leaders, teach­ers, police) to deliver psychological first aid in response to a mental health crisis is an important step to improving care ¹⁹⁸ . The crisis intervention team approach that has been used with law enforcement officers ¹⁸⁸ may also be relevant for traditional and religious healers or community leaders, who play an important role in determining community responses to an individual with acutely disturbed behaviour.

The COVID‐19 pandemic has had a significant impact on the availability and accessibility of mental health care globally, including in LMICs ¹⁹⁹ . Use of hotlines and digital technology creates new opportunities to provide crisis support and to identify and respond to those at risk of suicide, although as elsewhere the most vulnerable may also be at high risk of digital exclusion. Ensuring that crisis care is available in local primary and general health care settings is essential. Competency‐based assessments of health workers delivering WHO’s mhGAP‐IG ²⁰⁰ in non‐specialist settings should incorporate de‐escalation techniques, and programmes should be informed by the WHO recommendations for crisis care ¹⁹¹ and ensure supported decision‐making and provision of alternatives to coercive care.

Formal mental health crisis services also need to be able to move outside of facilities – for example, providing outreach to those in crisis who are homeless or restrained at home and unable to access care. The potential contribution of peer support to many aspects of mental health care, including crisis response, is gaining traction in LMICs ²⁰¹ , ²⁰² , but starts from a low base of involvement and empowerment of people with lived experience of mental health conditions ²⁰³ .

Policies and legislation upholding the human rights of individuals experiencing a mental disorder are necessary to the implementation and sustainability of effective and appropriate interventions. The WHO has specified what legislation and regulations need to include, as well as how these might be implemented. For example, current efforts in India to implement these principles through new mental health legislation include strategies to support decision‐making for people experiencing a mental health crisis through advanced directives and nominated representatives ²⁰⁴ .

Much more robust evaluation needs to accompany programmes to improve crisis response within communities, ensuring that unintended adverse consequences do not result, for example, where law enforcement agencies or traditional healers become involved in crisis response. Before adapting existing or developing new interventions, we need greater understanding of what happens at the point of crisis, to identify ways to move towards more rights‐based and person‐centred care. Interventions should be co‐developed with service users, their families, service providers and other key members of the community to increase their appropriateness, acceptability and sustainability.

For the future, while the transfer of high‐intensity, high‐resource, specialist models from high‐income countries to LMICs is likely to be undesirable and ineffective at meeting need, reverse innovation is possible. Where crisis responses are developed that are embedded in communities and service user involvement, as in the voluntary sector responses discussed earlier, they have the potential to serve as a template for collaborative crisis care in high‐income countries.

CONCLUSIONS

Much of the focus in this paper has been on specific acute care models and the potential they hold for improving care and widening the range of options available in a crisis. However, this reflects a clinician rather than a patient perspective. During a crisis, a service user may seek help from and be supported by a range of local agencies and will be affected not so much by the quality of individual services as by the overall accessibility of appropriate types of help and the extent to which an integrated and flexible crisis response is available from helpful and empathic staff ²⁰⁵ .

So far, very little research has focused on the overall patient journey and on crisis care systems ⁴⁷ . A flexible and accessible local area crisis care system that offers a variety of crisis options to meet service user needs and preferences and that integrates sectors appears optimal. However, a relatively complex service system involving multiple crisis service models may also lead to fragmentation and service gaps. We therefore suggest that how best to design integrated local crisis care systems should be a research and policy priority. Co‐production with people who use services and their communities, as well as staff in all relevant sectors, is essential for such redesign to address diverse needs in crisis effectively and acceptably.