Safewards: a new model of conflict and containment on psychiatric wards


  • L. Bowers RMN PhD

    Professor of Psychiatric Nursing, Corresponding author
    1. Section of Mental Health Nursing, Institute of Psychiatry, London, UK
    • Correspondence:

      L. Bowers

      Section of Mental Health Nursing

      Institute of Psychiatry

      De Crespigny Park

      London SE5 8AF



    Search for more papers by this author

  • ‘This article is published with the permission of the Controller of HMSO and the Queen's Printer for Scotland.’ ‘The copyright line for this article was changed on March 19, 2015 after original online publication’.


Accessible summary

  • Rates of violence, self-harm, absconding and other incidents threatening patients and staff safety vary a great deal by hospital ward. Some wards have high rates, other low. The same goes for the actions of staff to prevent and contain such incidents, such as manual restraint, coerced medication, etc.
  • The Safewards Model provides a simple and yet powerful explanation as to why these differences in rates occur.
  • Six features of the inpatient psychiatric system have the capacity to give rise to flashpoints from which adverse incidents may follow.
  • The Safewards Model makes it easy to generate ideas for changes that will make psychiatric wards safer for patients and staff.


Conflict (aggression, self-harm, suicide, absconding, substance/alcohol use and medication refusal) and containment (as required medication, coerced intramuscular medication, seclusion, manual restraint, special observation, etc.) place patients and staff at risk of serious harm. The frequency of these events varies between wards, but there are few explanations as to why this is so, and a coherent model is lacking. This paper proposes a comprehensive explanatory model of these differences, and sketches the implications on methods for reducing risk and coercion in inpatient wards. This Safewards Model depicts six domains of originating factors: the staff team, the physical environment, outside hospital, the patient community, patient characteristics and the regulatory framework. These domains give risk to flashpoints, which have the capacity to trigger conflict and/or containment. Staff interventions can modify these processes by reducing the conflict-originating factors, preventing flashpoints from arising, cutting the link between flashpoint and conflict, choosing not to use containment, and ensuring that containment use does not lead to further conflict. We describe this model systematically and in detail, and show how this can be used to devise strategies for promoting the safety of patients and staff.


Conflict (aggression, self-harm, suicide, absconding, substance/alcohol use and medication refusal) and containment (as required medication, coerced intramuscular medication, seclusion, manual restraint, special observation, etc.) are important matters for hospital management and nursing practice. Violent incidents can lead to injuries, sometimes serious, to staff or patients (Langsrud et al. 2007). Suicides, by definition, involve the death of a patient, and absconding is associated with suicide risk (Appleby et al. 2006). Self-harm is also injurious, and its management and prevention tax nursing skills, as well as self-harm being an indicator of increased suicide risk (James et al. 2012). The use of force and coercion that can be involved in containment arouses staff ambivalence and can result in unintended injury to patients, or spoil cooperative staff–patient relationships. Reducing the frequency and severity of these events is clearly very important for wards, the patient who reside there and the staff who work there.

The idea that different events (aggression, self-harm, absconding, etc.) can be grouped together as conflict, and different management methods (as required medication, seclusion, manual restraint, etc.) grouped together as containment, is supported by two main arguments. First, patients who exhibit one sort of conflict behaviour are likely to exhibit others, i.e. these behaviours cluster within patients (Bowers et al. 2000, 2003). Second, different conflict and containment rates cluster within wards, i.e. wards that have high rates of aggression also have high rates of absconding, or wards that have high rates of coerced intramuscular medication of patients also have high rates of special observation use (Bowers 2009). Studies of community samples of young people have also found evidence for a common factor between different problem behaviours (Cooper et al. 2003, Kingston et al. 2011). The implication of these commonalities is that the different events and actions have common causes, and that making an attempt to delineate these in a single model is a sensible thing to do.

Wards vary significantly in their rates of conflict and containment, sometimes by a tenfold margin (Bowers 1998, 2009, Bowers et al. 1998). Rates also vary internationally (Bowers et al. 2005, Nijman et al. 2005), and containment methods used in some countries are not used in others (Bowers et al. 2007). Explanations for these differences have not often been sought or described in a systematic way. Where they have been offered, they are restricted to specific types of conflict, most often aggression (Nijman et al. 1999). The Safewards Model represents our attempt to fill this gap.

The Safewards Model in simple form

The most basic form of the Safewards Model is shown in Fig. 1, which summarizes the factors that influence the rates of conflict and containment in wards, and explains why some wards have much more conflict and containment than others. The terms in the model have the following meanings:

Figure 1.

Safewards Model (simple form)

Originating domains. Psychiatric wards are social and physical locations, separate from patients normal residences, and provide 24/7 mental health care on a basis of mixed voluntary and legal coercion. As such they have six aspects or collections of features that can influence the frequency of conflict and/or containment.

Staff modifiers are features of the staff as individuals or teams – or the ways in which the staff act in managing the patients or their environment, initiating or responding to interactions with patients – that have the capacity to influence the frequency of conflict and/or containment.

Patient modifiers are ways in which patients respond and behave towards each other that have the capacity to influence the frequency of conflict and/or containment, and which are susceptible to staff influence.

Flashpoints are social and psychological situations arising out of features of the originating domains, signalling and preceding imminent conflict behaviours.

Conflict collectively names all those patient behaviours that threaten their safety or the safety of others (violence, suicide, self-harm, absconding, etc.).

Containment collectively names all the things that staff do to prevent conflict events from occurring or seek to minimize the harmful outcomes (e.g. p.r.n. medication, special observation, seclusion, etc.).

Our model indicates that there are a set of conflict-originating factors that can give rise to specific flashpoints which can then trigger a conflict incident leading to containment. The model also indicates that containment is in a dynamic reciprocal relationship with conflict, and that sometimes the use of containment can itself give rise to conflict rather than successfully prevent it. Finally, the model shows that staff can influence the rates of conflict and containment in their wards at every level: by reducing or eradicating the conflict-originating factors; by preventing flashpoints from arising out of them; by cutting the link between the flashpoint and conflict, i.e. the flashpoint occurs but does not lead to a conflict event; by influencing the patient modifiers of those same processes; by judiciously choosing not to use containment on occasions when it would be counterproductive; and by ensuring that containment use does not lead to further conflict when it is used.

The Safewards Model expanded

The full form of the Safewards Model can be found in Fig. 2. Six domains identify the key influences over conflict and containment rates: the patient community, patient characteristics, the regulatory framework, the staff team, the physical environment and outside hospital. The outermost ring summarizes the key features within those domains that can give rise to conflict and containment events. The next ring indicates the patient modifiers, what patients can do together that influences the way in which the features of the six domains give or do not give rise to conflict and containment events. The next ring indicates the staff modifiers in a similar fashion. Where arrows exist between this ring and the outmost one, they indicate that staff also have the power to directly modify or alter the features of the domains so as to reduce the risk of conflict or containment events. The innermost ring identifies the flashpoints most closely related to the domains within which they sit, flashpoints being those events or social circumstances that are most likely to trigger a conflict or containment event in the very short term. Conflict and containment are at the centre of the model, linked by a bidirectional arrow representing the fact that while conflict can trigger containment, containment use can itself trigger conflict.

Figure 2.

The Safewards Model (full form)

Staff team domain

The internal structure of the ward is asserted by the staff team, and is composed of the rules of patient conduct, the daily and weekly routine as to what happens when and where, and the overall ideology asserted by the staff either overtly or implicitly by their behaviour as to the purpose of the ward and what it offers to patients. Also included in internal structure are the efficacy and efficiency with which that ideology is put into practice, as shown by the timely and responsive way that the ward as an organization for delivering inpatient care operates. One common and highly visible signifier of an efficient organization is overall cleanliness and tidiness, hence its inclusion here. Finally, the custom and practice among the staff team as to what happens when patients behave in ways incompatible with or disruptive of the internal structure also form part of this domain, as choice of containment method is highly locally determined and very variable among wards, hospitals and countries.

The staff modifiers of the internal structure include the following:

  1. Staff anxiety and frustration, or rather the degree to which staff can regulate their normal emotional responses to the disruptive behaviour of patients that threatens the internal structure of the ward. Staff anxiety accentuates patient anxiety and self-control ability, as well as hinders nurses' ability to respond in the most effective and socially skilled way. Staff frustration and anger have the capacity to amplify patient anger, or alternatively trigger catastrophic loss of self-esteem, either of which responses can trigger further or more extreme conflict behaviours.
  2. Moral commitments, particularly to honesty (even when it was difficult or costly), bravery (being willing to confront patients and risk violence when necessary), equality (demonstrating, through a variety of ways, a lack of superiority), non-judgmentalism (eschewing large-scale moral valuation of the patient), universal humanity (expression of an inclusive picture of the human race and a valuing of people despite their diversity) and individual value (an appreciation of the value of the individual person).
  3. Psychological understanding, meaning being able to deploy a range of alternative explanations for the difficult behaviour of patients, derived from psychological models, studies or psychotherapeutic approaches, instead of judging patients to be morally bad and worthy of punishment. These psychological understandings, thus, generate different ways for staff to respond to such behaviours, as well as aid with emotional self-regulation.
  4. Teamwork and consistency refer to the way in which the staff support each other practically and psychologically so as to aid emotional regulation, specifically in allowing ventilation of emotions ‘off stage’ and in sharing the burden of face-to-face contact with challenging patients. In addition, the team produces consistency in asserting and applying the internal structure to patients, consistency over time, between nurses, and between patients. This aids in legitimizing the internal structure in the eyes of patients, supporting self-control, and dampening any sense of injustice and therefore anger.
  5. Technical mastery refers to the range, depth and quantity of social and interpersonal skills and responses available to the staff in order to deal with patient challenges to the internal structure, including bringing comfort to the distressed and the de-escalation of those becoming agitated, as well as skilled exercise of power and control.
  6. Positive appreciation indicates the degree to which the staff like and enjoy being with patients, affording them respect, compassion and companionship.
  7. The two-way arrows (Fig. 2) indicate that in the case of the staff team domain, the internal structure itself is under the control of the staff, who determine the content of the rules and routines, or who operate efficaciously or not. Therefore, the domain itself can be regarded as a staff modifier.

The flashpoints for internal structure are those moments where power and influence are exercised by staff, either when denying or refusing a patient's request, asking a patient to do (or stop doing) something, communicating unwelcome news to a patient about a staff decision taken elsewhere, or when ignoring a patient's overt or implicit requests for assistance or support.

Physical environment domain

The features of the physical environment that influence conflict and containment rates include its quality (better quality environments evoke greater care, are more comfortable and express greater respect for patients) and complexity (more difficult to observe environments make supervision by the staff harder, and supervision suppresses suicidal impulses and enhances self-control). Other features of the physical environment relate more directly to containment, for example whether the door to the ward is locked to patients trying to exit, or whether a seclusion room or a psychiatric intensive care unit is available.

The staff modifiers of these features include the maintenance of the environment, such as speedy repairs, frequent redecorations, and regular furniture replacement, and the staff's own respect for the physical environment and their caring attention to it, as well as keeping the environment clean and tidy so that it looks its best. Other staff modifiers reflect the degree to which the physical environment can be adjusted to patient choices regarding colour and decoration, from choices of bed coverings and curtains through to the availability of posters and the potential for personalizing bedrooms or bed spaces. A further element of staff modifiers are the ways in which staff adjust the way they operate so as to provide good patient supervision, from the use of checking routines through to being caringly vigilant and inquisitive. This refers to the staff taking an interest in patients, observing them, responding to indications of distress, and/or noticing their absence, as well as being inquisitive to the degree that they will respond to unusual noises or unsatisfactory responses and inquire as to what is going on in an assertive manner.

Flashpoints include patient secrecy or solitude, spaces, and times in which the lack of staff supervision allows the surfacing and acting upon of suicidal or self-harming instincts, or which allow abuse or bullying between patients. The degree of admission shock experienced by patients is also likely to be increased if the ward is in a deteriorated and unkempt condition. The point at which the exit is discovered to be locked may prompt either anger/resistance, or a slump in self-esteem and potential self-harm.

Outside hospital domain

Stressors from outside hospital largely relate to the patient's friends, family or home. Contact with friends and family, if hostile, argumentative or upsetting in other ways (for example, the patient's absence from important events, or an expressed need for support from the patient that cannot be provided, or the conveying of bad news of some sort, such as illness, death or other loss), can give rise to distress and conflict behaviours. Some relationships with family members may be toxic or extremely stressful for patients, for example demanding parents who show no understanding of the effects of mental illness; or a major relationship with a partner that is breaking down, financial and childcare agreements after divorce; or childcare difficulties, poor bonding or even abuse, and the involvement of social services. Contact with friends and family can occur via phone, email, social networking channels, letters or during visits. Other stressors from outside hospital relate to home and accommodation; for example, there might be requirements for home care that patients have difficulty in coping with while in the hospital, such as bills, repairs, maintenance, as well as worries about burglary during their absence. Alternatively, moves of accommodation are common during an admission, and if that move is to a less desirable place in the eyes of the patient, as discharge approaches, stress and conflict behaviour are more likely.

Staff modifiers relate to acquiring and developing a fully rounded knowledge of the patient's friends and family network, coupled with an appreciation of the meaning, nature and significance for the patient of his or her relationships with them. Such full knowledge allows either the effective involvement of friend and relatives in care provision, or a fully therapeutic approach to dealing with any problems or issues, potentially involving a range of different therapeutic approaches, from parenting training through marital or couple therapy, through to family therapy provision. Active patient support in these relationships by the staff, assisting them to manage and regulate them, offers further possibilities for modification of their potential to lead to conflict behaviour in the ward.

Flashpoints include the occurrence of an argument with a friend or family member, receipt of bad news from outside the hospital, a loss or disappointment on the part of the patient, and a home crisis of some sort (fire, burglary, actual or threatened loss of tenure, major reminders of bills and indebtedness).

Patient community domain

Conflict arising from the patient community has its roots in contagion or discord. Contagion arises either because patients copy the disruptive or risky behaviour of each other, or because such behaviour on the part of other patients arouses anxiety and uncertainty, triggering certain conflict behaviours as coping mechanisms or defences. Alternatively, the anxiety aroused may lead to more frequent or intense psychiatric symptoms that themselves give rise to further conflict behaviours. The other origin of conflict in the patient community is discord between patients, who are essentially living in close proximity with others they did not choose, and whose behaviour can be difficult, unpredictable, irritating or obnoxious.

In this case, there are patient modifiers that influence whether contagion or discord actually gives rise to conflict behaviour, and these parallel the staff modifiers relating to internal structure, for example patients' ability to regulate their own normal emotional responses of anxiety and frustration towards the behaviour of their fellow patients; their psychological understanding of such behaviour in order to avert judgement and condemnation; their technical mastery of social skills and repertoire of graceful social responses; their moral commitments to honesty and equality; and the degree to which patients in the ward, as a group, offer each other mutual support in tolerating the difficult behaviours of those who at any one time are extremely disruptive.

The staff modifiers are, thus, largely about how the staff support and help patients respond positively to each other. Role modelling of equanimity and of skilled responses to challenging behaviour potentially equips those patients who witness it with greater skills. Giving explanations about behaviour and information of psychiatric symptoms and conditions (including formal education packages) fosters patients' psychological understanding of each other. In addition, the possibility of copycat events can sometimes be prevented by the immediate removal of the means to carry them out, for example removal from the ward of all plastic bags following a patient's attempted suicide using one. The presence of staff and their good relationships with patients (presence+) allow intervention at an early stage of potential arguments, with diplomatic negotiation or other action averting irritations that may otherwise later turn into violence.

As the origins of conflict in this domain are contagion and discord, flashpoints include any occasion in which patients are brought into close proximity with each other, so any assembly, joint activity or crowding in the ward prompts interactions that can be difficult, induces misunderstandings between patients, or fosters the witnessing of conflict by other patients. Those misunderstandings may be further fostered if communication between patients is made more challenging by the stress of queuing or waiting, or by a high level of noise, making hearing more difficult. Bullying, stealing and property damage between patients are also incendiary and likely to lead to conflict if not managed or dealt with. Finally, staff and patient turnover increases anxiety and uncertainty in the patient community, making conflict more likely.

Patient characteristics domain

A large variety of patient characteristics can give rise to conflict behaviour, and these fall into three groups:

  1. Symptoms, for example paranoia resulting in defensive aggression or absconding; specific delusions motivating irrational behaviours; auditory hallucinations, such as voices instructing the patient to behave in certain ways; depression, leading to suicide attempts or irritability; or use of alcohol or drugs, resulting in irritability or disinhibition.
  2. Personality traits, perhaps especially features of antisocial personality disorder leading to instrumental aggression, or borderline personality disorder linked to self-harm.
  3. Demographic features, particularly being younger and male.

The staff modifiers of this are, therefore, the delivery of the most effective and efficient treatments, which may involve pharmacotherapy and/or psychotherapy. The speedy resolution of symptoms means reduced risks of conflict behaviour. One specific version of psychotherapy would be the functional analysis of conflict behaviours that the patient does exhibit, coupled with the appropriate behaviour treatments to extinguish them (Emerson & Einfeld 2011). Finally, general nursing support and intervention in terms of responding to patient symptoms, providing reassurance and minimizing the impact of those symptoms of patients' behaviour all provide opportunities to reduce the risk of conflict behaviours occurring.

Relevant flashpoints in this domain include exacerbations or sudden increases or expressions of severe symptoms or illness, or any occasions in which patient freedom, liberty and independence are curtailed – issues of particular salience to young men and sensitive for those with personality disorder traits. In this way, the flashpoints of the patient characteristics domain link to those already described under the staff team domain.

Regulatory framework domain

The external structure of the ward includes those constraints on patient behaviour dictated largely from outside the ward itself. These range from the operations of the mental health act and the coerced detention of patients in hospitals against their will (resulting in patient hostility, anger, aggression and absconding), through national policy on mental health care as it impinges upon patients' journey through the psychiatric system (what is or is not provided and under what conditions, treatment, accommodation, financial benefits.), to hospital policies around complaints, appeals and prosecutions of patients for assaults or other criminal behaviour.

With the exception of hospital policy, which may be influenced by the staff delivering direct care in wards, the rest of these things are not under staff control. However, the way in which they are executed can be modified by the staff. Respect for patient rights, attention to due process, the provision of accurate information particularly in relation to appeals and advocacy, expressions of hope and positive planning for the future, and support in utilizing the complaints process all enhance the patient's perceived legitimacy of the external structure, reducing the frustration and hopelessness that can lead to conflict behaviours. Increasing the liberty or choices of the patient in areas where this is still possible may also compensate for restrictions that have to be applied.

Flashpoints in this domain are those moments in which power is exercised by the psychiatric system, potentially resulting in aggressive rebellion or collapse of self-esteem and depression on the part of the patient. These include the refusal to allow a patient to leave the hospital, the enforcement of treatment, and the failure of a complaint or appeal. These moments bring the patient's situation into sharp relief and can trigger conflict behaviours.

Original contributions by the Safewards Model

The Safewards Model seeks to explain all conflict behaviours and all containment methods collectively. The model is, thus, more comprehensive than separate models for aggression, absconding, etc., and acknowledges the now strongly empirically substantiated relationships between them. By depicting the bidirectional link between conflict and containment, the model shows that the use of containment motivated by the desire to prevent future conflict can, in some cases, trigger that conflict. It also, thereby, allows for separate interventions to reduce containment without necessarily influencing conflict rates, such as the many seclusion and restraint reduction initiatives in a variety of countries around the globe. Furthermore, in drawing a distinction between originating domains and flashpoints, the model highlights the tensions produced by the normal operations of inpatient psychiatry, that are indeed intrinsic to it, and demonstrates how these give rise to more focused, time-located flashpoints. The separate identification of originating domains and flashpoints allows clearer thinking about what can and cannot be changed by clinical staff working in the ward, and facilitates the generation of ideas for change that have the potential to reduce conflict and containment.

The Safewards Model also brings significant new considerations to the fore. For the first time, patient–patient interactions are seriously considered and included in explanations for rates of conflict and containment. While patient characteristics and symptoms have been widely previously reported as causes of conflict and containment, the Safewards Model identifies treatments as an effective safety-producing strategy, and identifies that the way staff respond to patients' characteristics can significantly impact on their capacity to give rise to actual conflict or containment events. The external structure or regulatory framework and its features are, for the first time, identified as an originating domain for conflict and containment in the ward. There are still things that clinical staff can do in implementing these structures that can reduce the risk of conflict and containment. However, the Safewards Model also demonstrates that there are actions that can be taken at the highest policy levels which can lead to wards that are safer for staff and patients. Additionally the Safewards Model incorporates influences on patients' behaviour from outside hospital, providing new understandings and therefore new ways to intervene. Finally, while the importance of the physical environment has been asserted by others, the Safewards Model moves beyond a simplistic recommendation of quality improvement and to a more precise description of the physical features of wards and their relationship to staff and patient safety.

Implications for conflict and containment reduction

Staff team

The obvious implication is that any intervention that brings about change to psychological understanding, moral commitments, emotional regulation, technical mastery and teamwork skill, building positive appreciation, and effective ward structure is likely to contribute towards reducing the rates of conflict and containment. Two clear routes to bringing about such change are education/training and clinical supervision of frontline workers. However, these are not the only possibilities. Ward managers and qualified nurses can project these values, model these skills, informally instruct their fellow staff members, challenge each other, and review the care of patients in light of these principles. In addition, it may be possible to nudge a ward in the right direction by devising a number of tasks and protocols for doing things, or implementing small procedures that edge people towards change.

An alternative approach is to focus on the identified flashpoints, finding better ways to manage them. The flashpoints are the social locations in the ward that are most likely to trigger conflict, the staff–patient interactions where the ward structure is established, re-affirmed, demonstrated and instantiated. It is not that hard to find ways that nurses can clarify the structure with the patient group, perhaps reduce the number of rules and be more consistent in their application. The ways in which nurses interact with patients over these issues can be changed. To take a simple example, instead of waiting for patients to knock at the office door, requests can be pre-empted by nurses going around the ward and asking patients what they need and want in advance. Similar small changes to routines and usual practices have the capacity to make large impacts on rates of conflict and thus containment.

The connection between structure and containment is clearly mediated by aggression. Limit-setting or patient requests that are denied do not, by themselves, lead to containment. But if a confused, frustrated and belittled patient responds with anger that is met by a combination of anxiety and irritation by the staff, the use of containment may well be the eventual result. Cutting this cycle and others like it may pay a serious dividend.

Physical environment

That the physical environment has a part to play demonstrates that the hospital, its managers, resources and organization are all likely to have some effect on conflict and containment rates. Ward physical environment quality, including its cleanliness, are all products of current or past managerial action. Whether the ward door is permanently locked or not is also likely to be a hospital wide managerial and policy decision, and as we have already noted this may have gains (decreased absconding) and losses (increased aggression and self-harm).

However, the staff can also act to modify the way in which the environment interacts with patient behaviour. Increased checking routines and the use of intermittent observation can compensate for ward complexity, and monitoring patients by being caringly vigilant and inquisitive can prevent suicides (Bowers et al. 2011). Prompt requests for repairs, attention to décor, supervision of and attention to cleaning services, and keeping the ward tidy can all contribute to a better quality environment that enhances patients' self-esteem, expresses respect and can reduce absconding. If the ward is locked, staff could increase alternative choices for patients, or act in ways that enhance self-esteem or minimize the impact of the locked door.

Outside hospital

In order to influence the capacity of external factors to trigger safety-threatening incidents in the ward, staff need to be aware of and involved in more than patients' life in the hospital. Their financial circumstances are important, and they may need help and support with the benefits system, money management and debts, or simply help with accessing resources they already have. The importance of family problems means that carers cannot be held at arm's length from the ward while the patient is treated. They may need to be engaged with services and supported to resolve problems during the admission. If they are supportive and helpful, patients might need assistance with keeping in contact with their family and friends, and discussion might need to take place over what they perceive to be their family responsibilities. If they are worried about their accommodation, this can be checked on by community workers, or leave visits could be facilitated. Inpatient care has a tendency to focus attention, thinking and action mainly on patients with their problems in the ward, seeing everything else as a task for community services. However, such a focus neglects the important factors outside the ward that are influencing patient behaviour in the ward, potentially in very negative ways. Thus, the distal effects of what goes on outside the hospital for patients can be events that inpatient staff have to urgently deal with, and prevention can entail grappling with those initial causes, even if they lie outside the ward and hospital.

Patient community

The management of patient–patient interactions may be as important as the management of staff–patient interactions. Significant effort here on patient education (in technical mastery and psychological understanding), conflict resolution, role modelling and staff presence might reduce the rates of conflict. Awareness of the risks of contagion could mean that preventative actions can be taken, or pre-emptive reassurance or explanations can be given to the patient community. In some ways, the evidence on the importance of the patient community reaffirms the potential value of a modified therapeutic community approach in inpatient wards.

Management of the level and fluctuation of activity in the ward might also be a means to produce a calmer, quieter environment, with less chance for patients to be put in close contact in potentially tense and ambiguous social situations.

Patient characteristics

The consistent links to younger age and male gender indicate that much conflict and ensuing containment is about rebelliousness, independence and power, all of which are highly salient issues for men and for younger people. This hints that finding ways to enhance patients' choices, freedom and control over their circumstances might help reduce conflict and battles with the staff. Efforts to achieve a mutually respectful partnership might do much to avert conflict arising from these issues. Both point towards the potential value of a modified therapeutic community in wards (Haigh 2002, Mistral et al. 2002), and also indicate that any hint of authoritarianism on the part of the staff will be counterproductive and incendiary. The additional fact that a significant proportion of conflict and containment events are accounted for by a smaller proportion of patients indicates the possible efficacy of (1) staff changing their responses to patients after the first event to avoid subsequent ones, and (2) targeted therapeutic interventions directed towards the most difficult patients.

The link of conflict and containment to illness and symptoms also carries many lessons. First and foremost among these is that effective and speedy treatment will reduce symptoms and conflict and containment. However, treatment for psychoses does not have to mean drugs only. It can also include elements of cognitive behavioural therapy, functional analysis, social skills training and other psychotherapeutic treatments. However, it does have to be admitted that few of these have been adapted for acutely ill inpatients, or have been tested for their efficacy among acute ward populations. Much more research is needed here. This being so, the origins on conflict in psychiatric symptoms still do not imply that there is nothing that nurses can do. Symptoms are not expressed or even experienced in a vacuum, but in the social context of the ward. Nurses have choices over how to respond, and some responses may be better than others, some may help enhance patients' coping strategies, whereas others might increase patients' stress, and hence elicit yet more symptoms (Zubin & Spring 1977). Nurses can also influence how patients respond to each other's symptomatic behaviour, modelling efficacy, respect, dignity-giving and de-escalatory approaches, or even directly teaching or advising other patients on what to do. They may do the same for patients' friends and family when they visit. In doing so, they can draw upon their experience and traditional psychiatric nursing practice. That psychiatric nursing practice, expertise, creativity and skills in symptom management do exist, but would benefit from further systematic collection, codification and publication (Bowers et al. 2009).

This is not the end of the implications of the link between symptoms and conflict and containment. In addition, this means that all nurses should have expert knowledge of and be able to recognize psychotic symptoms. And this should be beyond just delusions and hallucinations, incorporating detailed understanding of a range of thought disorders (McKenna & Oh 2005) and their implications for care, and in-depth knowledge and understanding of the cognitive difficulties experienced by psychotic patients (Wykes & Reeder 2005). Furthermore, recent research is establishing more and more that psychoses are the result of genetic inheritance, plus childhood adversity, deprivation and abuse (Morgan et al. 2008). This makes approaches, such as trauma informed care (Bloom 1997), worthy of significant consideration.

Regulatory framework

On a national basis, this would seem to imply that a more liberal mental health legislation that gives more rights to patients and provides more scope for challenge and appeal would contribute to more peaceful wards. It implies that attention to procedural justice, patient rights, information giving, facilitation of appeals and factors that are more under the control of local staff would all increase the legitimacy of the psychiatric ward, support greater patient self-esteem, dissipate and diminish anger at detention, and contribute thereby to lower conflict and containment rates. An effective, unbiased, objective, neutral and speedy complaints process for patients might also contribute towards the same ends. In addition, policy targeted at reductions in specific containment use, mandating such things as reviews, time limits for reviews and more careful authorization procedures, might have a significant impact.

In the operation of any mental health legislation, listening to patients' points of view, hearing them out, negotiating with them, and being generous, flexible and willing to compromise might also contribute to reducing conflict and containment. In a complementary way, giving autonomy to patients in other areas might compensate for the restrictions necessary for detention. Choices could be increased around in ward activities, meals, snacks, décor, furnishings and timings, all of which might address detained patients' needs for respect and freedom. Staff may also intervene to address hopelessness and self-stigmatization due to hospitalization in a variety of informal and organized ways.

As an organization, structure may be facilitated by clear and consistent policies on what patients are and are not allowed to do, and effective communication of those policies to all patients and staff. Moreover, the organization can helpfully articulate clearly the purpose of the wards, the value base of care and the general unpinning philosophy, all of which will enhance structure. There might be additional management activities that can ensure that the organization supports the purpose of the wards, and effectively and efficiently provides the services required for that care to be delivered in the ward, for example fast turnaround of any necessary investigations, speedy processing of referrals to additional disciplines with fast response times, administrative support and high-quality continuous cleaning services. Collectively, these endeavours might enhance structure and reduce conflict and containment; when absent, they result in an inconsistent and disrespectful service to patients.

In relation to containment, the evidence is fairly clear that usage can be restricted by policy without adverse outcome.


I have described a new model that explains the differences found between wards in their conflict and containment rates. The model introduces new concepts and describes new domains as part of the causal explanation offered. Most valuably, the Safewards Model allows the generation of a number of different interventions in order to reduce the rates of conflict and containment in wards.


This paper presents independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research Programme (RP-PG-0707–10081). The views expressed are those of the author and not necessarily those of the NHS, the NIHR or the Department of Health.