Exploring the Utility of the Model for Understanding Inpatient Aggression for General Mental Health Settings

ABSTRACT

A comprehensive understanding of factors contributing to aggression within mental health inpatient units is necessary to generate timely preventative interventions. In practice however, the focus is dominated by consideration of factors that are intrinsic to patients. The Model for Understanding Inpatient Aggression (“the model”) was developed for use in a forensic mental health service to understand and help prevent aggression by exploring distal and proximal intrinsic patient‐level ‘personal’ factors and extrinsic ‘contextual’ factors'‐ structural, organisational and interpersonal. The model also incorporates exploration of ‘near miss events’ when aggression does not occur, to learn about aggression and support early use of effective intervention strategies. This study explored whether the model is suitable for use in civil/general mental health inpatient settings. Fifteen people (n = 14 nurses, and n = 1 lived and living experience expert) participated in one of four focus groups. Data were analysed using thematic analysis. Three themes were constructed from the data: (1) Focus remains on intrinsic factors and post event exploration and not early intervention and prevention; (2) “Turning the mirror” on ourselves: The importance of considering and addressing contextual factors; and (3) ‘For us it might fill a gap’: Benefits of using the Model for Understanding Inpatient Aggression. Participants also suggested additions to the model so that it aligns with practice in general mental health inpatient units. In conclusion, understanding and preventing aggression remains a constant challenge. Intervention efforts may be broadened when a wide range of contributing factors is considered, beyond proximal intrinsic/personal patient characteristics.

1. Introduction

Despite mental health services reform and the introduction of various preventative strategies, aggression remains a ubiquitous problem in mental health services (Parakkal Kurian et al. 2023). While all staff play a role in the prevention and management of aggression, mental health nurses' role is significant (Maguire et al. 2023) and they are more commonly exposed to aggression, in various forms (e.g., verbal and physical), than other staff. Aggression can result in the use of restrictive practices such as restraint and seclusion, and the use of these practices is often experienced negatively by patients. Their use also places considerable stress on nurses (Jalil et al. 2017; Jury et al. 2019). Despite ongoing concern regarding aggression in mental health inpatient units ¹ , there are only a few models commonly used to assist practice and inform risk assessment and intervention practices.

Models in this instance are a construct used to assist in understanding why inpatient aggression occurs, as opposed to frameworks, which usually consist of an overarching set of principles or guidelines to provide a foundation for synthesising and understanding information (Partelow 2023). Several models have been developed to assist clinicians in understanding inpatient aggression (for some examples see Bowers 2014; Cutcliffe and Riahi 2013; Malpass and Carroll 2024; Nijman et al. 1999). One of the more recent models that have been introduced into a number of settings and countries is the Safewards model, developed by Len Bowers and colleagues. Safewards is a model of care that supports the reduction of conflict (events that threaten staff and patient safety such as aggression, self‐harm, suicide and absconding) and containment (things that staff do to prevent conflict or minimise the impact of conflict) (Bowers 2014).

Many models, including Safewards, have similar features in that they encourage exploration across patient‐related factors, unit environment, the wider health system‐related factors, and staff‐related factors. These models also acknowledge that a multi‐pronged approach may have the biggest impact on reducing inpatient aggression (Bowers 2014; Cutcliffe and Riahi 2013; Malpass and Carroll 2024). However, most models do not encourage examination of near miss events. To assist mental health nurses with this important task, the Model for Understanding Inpatient Aggression was developed, originally for use in a forensic mental health inpatient units to underpin an organisational understanding of aggression and assist in the prevention and management of inpatient aggression (Daffern et al. 2015).

2. Background

The Model for Understanding Inpatient Aggression was developed in 2012 and has been periodically revised. The most recent revision considered its application to understanding aggression in bed‐based prison mental health units (Maguire et al. 2021). The model (see Figure 1) includes key constructs designed to understand how aggression may or may not occur by prompting the consideration of a range of distal (long‐term) and proximal (short‐term) intrinsic (personal) risk and protective factors, as well as long‐term and short‐term extrinsic (contextual) factors (see Table 1 for definitions of the key contributing factors). These factors are considered with regard to how they can influence a person's state of mind. If a high‐risk state emerges then a person may be susceptible or primed to act in an aggressive manner. There is also the possibility that in this heightened risk state that the patient can be successfully de‐escalated, and no aggression occurs. Of note, this model suggests that ‘near miss’ events are important. Investigating these may help learn about what contributes to aggression and which factors helped prevent aggression (Daffern et al. 2015; Maguire et al. 2021).

FIGURE 1.

The model for understanding inpatient aggression (Maguire et al. 2024).

TABLE 1.

Key constructs in the model for understanding inpatient aggression (Daffern et al. 2015).

Risk Factor: Any factor that increases the likelihood of aggression. Protective Factor: Any factor that decreases the likelihood of aggression. Long‐Term (Risk and Protective) Factors: Any factor that is enduring across an extended period; will tend to be present for a time period of weeks to months. Short Term Factors: Any factor that is likely to fluctuate over shorter periods of time‐ anywhere from minutes to days. Contextual Factors: Factors in the environment, including structural (e.g., fabric of building) and organisational (policies and procedures) and interpersonal (milieu/culture). Personal Factors: Factors intrinsic to the individual. State of mind: A temporary psychological state of being, that influences readiness to respond in a particular way; influenced by the long/short term, contextual/personal risk and protective factors. High risk state: A temporary psychological state of being indicating a high likelihood of acting aggressively soon. Low risk state: A psychological state of being indicating a low likelihood of acting aggressively soon. Escalation: A stage marked by increasing arousal and signs and symptoms that aggression is imminent. De‐escalation: A reduction in level of arousal and movement to a calmer personal space using self‐soothing strategies such as listening, discussing, negotiating, diverting, and deflecting. Aggression: Any type of behaviour that is directed towards harming or injuring another person who is motivated to avoid such treatment (Baron and Richardson 1994). Reconciliation: The person can regain control (see above) of behaviour and emotions, vigilance is decreased. Post event: A return to the pre‐high‐risk state, a time when the person may be receptive to exploring strategies to prevent recurrence of aggression.

The model was constructed to foster collection and consideration of factors related to the patient, staff, unit, and organisation, by promoting investigation of short‐term and long‐term patient risk and protective factors (see Tables 2 and 3), and short‐term and long‐term contextual risk and protective factors that may have influenced aggression (Daffern et al. 2015; Maguire et al. 2021). Considering a range of factors discourages narrow conceptualisations of aggressive behaviour, such as ascribing the causes of aggression exclusively to personal characteristics of the patient rather than also considering contextual factors (Cutcliffe and Riahi 2013; Daffern et al. 2015; Maguire et al. 2021) as this limits opportunity for prevention and may lead to negative appraisals of patients.

TABLE 2.

Personal factors.

	Long term		Short term
	Risk	Protective	Risk	Protective
Personal factors	Psychopathy Anti‐Social Personality Disorder History of Aggression (Facility with violence) History of abuse or mistreatment Attitudes supportive of violence Problematic early maladaptive schema Tendency to ruminate angrily and to think about violence towards others Anger experience, management and expression problems Major mental illness Hostile attribution bias Cognitive impairment Age related Neurocognitive disorder Dementia Acquired brain injury Learning disorder Intellectual disability Neurodiversity	Supportive relationships Good interpersonal skills Long‐term pro‐social goals Distress—tolerance skills Well‐developed problem solving and coping skills Pro‐social involvement (in education, employment etc.) Support attachments and bonds Connection to culture	Intoxication or disinhibition/withdrawal from substances Persecutory delusions Violent command hallucinations Hostility Anxiety/fear Victimisation Delirium Pain Unmet need Anger/humiliation Boredom Frustration Active symptoms of trauma Negative attitudes Poor relationships with others Strong need for interpersonal dominance Irritable Emotional and behavioural instability	Constructive/meaningful activity Cooperativeness/engagement with staff Regular sleep Strong therapeutic alliance Insight into behavioural problems and mental illness Positive attitude towards intervention and authority Engaged in leisure activities Motivated to participate in treatment Good response to treatment

TABLE 3.

Contextual factors.

	Long term		Short term
	Risk	Protective	Risk	Protective
Contextual factors	Poor communication within treating team or between unit staff and senior clinical managers and hospital administrators Inconsistency of treatment Low staff morale Structural problems with the unit (e.g., lack of light, space, unclean, poor temperature control) Problematic mix on the unit Poor organisation ethos Lack of appropriate policy or procedure Poor leadership Staff shortages recruitment and training issues (e.g., lack of training and access to clinical supervision) High staff turnover Inadequate staff levels, experience and mix on the unit Limited validated instruments and treatment/intervention for aggression Risk aversion and lack of positive risk taking Negative attitudes towards patients/and or the environment Extended length of stay due to social factors, including homelessness	Provision of a therapeutic milieu Structured unit program providing a range of educational, therapeutic, occupational and spiritual activities Well‐maintained environment Good care planning processes Staff competence Positive collegial relationships and psychological safety Healthy ‘safety culture’ Good communication from staff to patients regarding upcoming events such as transfers Effective senior leadership Therapeutic alliance and optimism Adaptability to review systems and implement change Individualised violence assessment and intervention Positive family/carer contact Collaborative/co‐produced risk management approach by the treating team/patient/family/carer De‐institutionalised/homely environments with private rooms	Inflexible, interpersonal interactions for example, authoritarian limit setting style, miscommunication ‘High demand’ period resulting in staff inaccessibility/distraction/lack of presence on ward ‘Bad blood’/subacute conflict if not addressed Lack of access to facilities Coercive or perceived coercive practices	Staff presence/‘being with’ the patient/s Utilisation of effective de‐ escalation techniques Staff with high level of individual knowledge/therapeutic alliance Family/carer contact/social connectedness Violence hot spots recognised and there is a coordinated approach implemented to reduce risk Ability to modify the environment Sensory/calming equipment available Proactive and non‐ coercive approaches for conflict resolution Meaningful occupation

The model is based on three core principles to understand inpatient aggression, indicating that the approach needs to be (1) comprehensive (uses a broad range of systems, assessments and interventions to improve and sustain clinical practice and ensure a safe working environment to aid personal recovery); (2) participative (all parties must consider it meaningful to work together and take an active role in designing and implementing initiatives that prevent or reduce aggression); and (3) systematic (collection and use of data is crucial to define the problem, design, implement and evaluate interventions for prevention and/or reduction of aggression) (Daffern et al. 2015).

Having a model can help guide clinical practice, understand aggressive acts, structure efforts to prevent aggression at multiple levels (person, unit and organisational), and contribute to the promotion of a safer environment for staff and patients that can facilitate recovery (Daffern et al. 2015). Given the ongoing issues related to aggression in forensic and mental health inpatient units (also often referred to as civil mental health or psychiatric inpatient settings, Dickens et al. 2013; Huitema et al. 2021; Meehan et al. 2016) it is possible that the Model for Understanding Inpatient Aggression may also be a helpful addition for mental health inpatient units. Therefore, the objective of the current study was to explore the utility of the Model for Understanding Inpatient Aggression for mental health inpatient units. The aims were to: (1) explore key stakeholder perspectives of the model, and (2) identify potential modifications/additions to the Model for understanding Inpatient Aggression for mental health inpatient units.

3. Methods

3.1. Design

A qualitative design was used to explore the different perspectives of participants about the Model for Understanding Inpatient Aggression, and to elicit any potential modifications/additions that might be necessary for mental health inpatient units. This design was selected to gain an in‐depth exploration of the participants perspectives on the model for use in practice outside of forensic mental health settings.

3.2. Participants and Recruitment

Purposive sampling was used for this study with the inclusion of the following groups of participants: (1) members from the Victorian Prevention of Clinical Aggression Forum. The purpose of this forum is to assist services in the state of Victoria Australia to identify, implement, and sustain evidence‐based practice in relation to the prevention and effective management of clinical aggression across mental health and acute healthcare settings. Each service is encouraged to have representation in the forum; (2) Clinical Nurse Consultants in the role of preventing aggression and reducing restrictive practice in mental health units, (3) consumer lived experience workers in the mental health improvement program in Victoria (one of the key initiatives of the improvement program is to work towards the elimination of restrictive practices), and (4) mental health Nurse Practitioners from across Victoria, who are involved in the prevention of inpatient aggression. The inclusion criteria were determined based on participants needing to have a sound understanding of the prevention of aggression in mental health inpatient units and the work towards eliminating restrictive practices. We also wanted to include lived and living experience expertise related to the prevention of aggression and elimination of restrictive practices. The Victorian Prevention of Clinical Aggression Forum members have experience in teaching the prevention and management of aggression, the Clinical Nurse Consultants are employed in roles to reduce aggression and use of restrictive practices, the Nurse Practitioners are the highest level of clinically skilled nurses and have responsibility to prevent aggression and use of restrictive practices. The Consumer consultant staff are employed to provide lived and living perspective and expertise in the mental health improvement program regarding the work towards eliminating restrictive practices.

An email invitation was sent to eligible participants with a summary of the study and a request to indicate their interest to participate by responding to the email. Consent to participate was gained through participants returning signed consent forms by email to TM. A total of 15 people participated, n = 10 Clinical Nurse Consultants, n = 2 Nurse Practitioners, n = 2 Nuse Unit Managers and n = 1 Consumer Consultant Lived Experience. See Table 4 for the composition of focus groups. Participants were from 14 different services, and out of these services, two used the Safewards model to under pin the understanding of inpatient aggression, and three services used locally derived templates or prompts, the rest did not have a dedicated model to understand aggression. None of the services were reported to be exploration near miss scenarios.

TABLE 4.

Composition of focus groups.

Focus group number	Participant role
Focus group one	Two Clinical Nurse Consultants and one Nurse Unit Manger
Focus group two	Two Nurse Practitioners
Focus group three	Three Clinical Nurse Consultants and one Consumer Consultant Lived Experience
Focus group four	Three Clinical Nurse Consultants and one Nurse Unit Manger

3.3. Data Collection

Data were collected via four focus groups between May and July 2024. Prior to the focus groups, participants were sent a copy of the model and asked to familiarise themselves with it. Focus group questions related to the current model used in practice, thoughts about the Model for Understanding Inpatient Aggression, and what adjustments might be needed for application to mental health inpatient units. The focus groups were conducted online using Microsoft Teams, and a focus group guide was used to structure the focus group discussion. With the group's permission, the focus groups were recorded, and the audio files were professionally transcribed. The focus group lasted from 45 min to 1 h and 8 min.

3.4. Data Analysis

Data from this study were analysed using the recommended six‐stage approach from Braun and Clarke (2019). The phases were (1) becoming familiar with the data; to achieve this, TM listened to the audio files several times and checked the recording against the transcripts to ensure accuracy and took notes. (2) TM and CD separately developed codes by finding similar statements and giving them suitable codes. (3) TM and CD then placed the collected codes into possible themes. Thematic maps were then generated on a word document. Phase (4) involved the early review of themes by the entire research team, which resulted in further refinement of the themes. (5) following this team review, transcripts were revisited by TM to check that coding supported the themes, and key data were reflected in the themes. The final step was writing this paper.

During the focus groups, various additions to the personal factors and contextual factors tables were elicited from the participants. Before making any changes to the tables, to ensure any suggestions from participants were supported by evidence in the literature, all suggestions were extracted from the transcripts, and a broad search of the literature was conducted to determine if there was any empirical evidence to support their inclusion. This was because all previous factors presented in the tables have been checked to ensure they are supported by evidence.

3.5. Rigour

To ensure rigour, a reflexive approach was employed. Invited participants were selected based on their knowledge and experience of working to prevent and manage aggression and lessen the use of restrictive interventions. All researchers have experience researching aggression, and all have qualitative research experience, and TM, CD, AC and MD are mental health clinicians, so they bring applied expertise to the interpretation of findings. The checking and correcting of audio files against the transcribed data was conducted to ensure accuracy of the transcriptions and to increase familiarity with the data. The data analysis was also conducted in a collaborative manner. Reporting for this study used the Enhancing the QUAlity and Transparency Of health Research (EQUATOR) network recommendations for qualitative research, the Consolidated criteria for REporting Qualitative research (COREQ) checklist (Tong et al. 2007).

3.6. Ethical Considerations

Approval to conduct this study was granted from the Swinburne University of Technology (Project ID: 7611). All ethical requirements were met. Confidentiality was retained by de‐identifying data and using participant numbers.

4. Findings

There were three themes that were interpreted from the data: (1) Focus remains on intrinsic factors and post event exploration and not early intervention and prevention; (2) “Turning the mirror” on ourselves: The importance of considering and addressing contextual factors; and (3) for us it might fill a gap: Benefits of using the Model for Understanding Inpatient Aggression in a mental health inpatient environment.

4.1. Theme One: Focus Remains on Intrinsic Factors and Post Event Exploration and Not Early Intervention and Prevention

This theme relates to participants reflections on current models, or the absence of models used in practice to conceptualise aggression, and their absence may contribute to a superficial or biased consideration of potential contributors to aggression. This is reflected by the following quotes.

It's still not that sort of discussion around what else could it be? What else is the aggression telling us other than just going for the most obvious sort of thing. I think that's where we miss a lot of opportunities to intervene earlier. (P1)

Participants reflected that the focus primarily involves a review of incidents after an act of aggression with scant focus on prevention. Without a model in practice to guide review, focus may remain on review following aggression.

Maybe it's because we are so focused on the post bit, and we should be focusing on the contextual stuff and the personal stuff (contextual factors) beforehand. (P3)

I think that we focus on aggression management training, but it's focused a bit more on what happens when they become aggressive, not what can we do to prevent that in the first place. (P2)

Without a model there is also risk that incident reviews are brief and limited.

What they're doing now is a rapid incident review. So rather… we'll have basically a quick and dirty review. (P1)

We don't do any kind of experience of care afterwards or look at what was contributing to the aggression. (P6)

4.2. Theme Two: “Turning the mirror onourselves”: The Importance of Considering and Addressing Contextual Factors

While acknowledging that it is potentially confronting for some staff to explore the contextual factors (which can include poor communication within the team, poor leadership and inflexible interactions, see Table 3 for more examples), there was consideration as to how this is necessary to ensure a rounded and robust conceptualisation of the issues.

I find sometimes it's a bit difficult to actually bring these things up when you feel that there's being poor leadership and things, but if you've got a matrix or something where you can refer to it might make it a bit easier. (P6)

It's loud, it's noisy, frightening, those sort of things I think get glossed over. (P1)

I think we often focus on the events that happened and what happened just before that as the trigger or the flashpoint to something as opposed to everything else that's under that umbrella beforehand. (P3)

Participants gave several examples of how the contextual factors present in their mental health services were considered to be contributing to a lot of the aggression from the stark or “awful” colours of the walls to the hospital beds and various other factors present in the environment.

Our walls are extremely clinical, they're white, they're aggressive almost. We've adapted stuff around that to make it more colourful. But a lot of the environment actually upsets and distresses patients as well…I would say 70% of the time our environment contributes to their distress. (P3)

We find our actual beds cause a lot of aggression. People just can't get a good sleep, they get cranky, they're restless, they've basically just got a mattress on a hardened base type of thing. So that causes a lot of issues in our ward. We did have a big paint and they went with this awful green, which is just terrible. (P2)

In addition, participants also reflected on staff factors they had seen contribute to aggression, where one participant vividly described the importance of “making sure that when staff go in to give a consumer feedback, they don't go in boots and all and rip someone's heart out of their chest and stick it to curb” (P6). Further, there was also identification that when practice, in particular interpersonal skills may not be ideal, “you as the staff member are actually the risk” (P6). There was also discussion acknowledging that conversations about contextual factors need to occur, however it can be difficult to bring up issues related to contextual factors when needing to address issues such as poor communication and leadership.

I think it'll lead (use of the The model) to a more in‐depth discussion around probably the areas that people don't really want to talk about. (P12)

While considered important to address, participants also highlighted that introducing these sorts of discussions will require a supportive environment.

It might be the patient be the focus again, sometimes the consultant might be the treating consultant and they're not going to throw themselves under the bus so to speak and say, well I didn't do this, or I could have done that.

I think we do need to turn the mirror on ourselves because…I may get things not quite right. So, it's important. I think we do look at that because we don't get things perfect, but I think we need to be able to stand up to that and not be fearful that we're going to get sacked. (P1)

The last quote highlights the cultural changes that may need to be in place to safely explore the contextual factors.

4.3. Theme Three: For Us It Might Fill a Gap: Benefits of Using the Model for Understanding Inpatient Aggression

The final theme relates to the perceived benefits participants identified if the Model for Understanding Inpatient Aggression was introduced to assist in conceptualising and reviewing aggressive behaviours. Participants liked how the model can be used to tease out the impact of contextual factors, which is reflected in the following quote.

I think it's a really good way to put that into practice to think about who is this person in the context of the environment that they're in rather than this person exhibits these factors so therefore may aggressive or something like that. It is a really good way to think about this person and people that are receiving care in the context that they're receiving the care in. (P8)

The model was also seen as filling a current practice gap, prompt consideration of relevant factors and assist in existing meetings where the focus is on reducing restrictive interventions.

That table is (personal and contextual factors tables) great because there's things in there that even I probably wouldn't have looked at or thought of. (P2)

I chair the RRI (reducing restrictive interventions) committee, my plan is to present this model and maybe adopt parts of it that we can use within the service, particularly around formalizing the contextual factors in the way it's been done at the model. I think it'll help the committee to maybe come to even better decisions than what we've been doing at the moment. (P12)

Despite the model being developed in a forensic mental health service, participants did not consider the model to be forensic mental health focused and many participants also spoke of the increasing presentations of patients who have histories of offending behaviour and/or significant risk of violence.

4.4. Suggested Additions to the Model

Participants in this study suggested some additions which would lead to better utility in an inpatient mental health setting. Participants also suggested changes or expansion on existing items to reflect changes to language and practice in the 12 years since the Model for Understanding Inpatient Aggression was developed, especially in light of recent mental health and broader health and social sector reforms (see Table 5).

TABLE 5.

Suggested changes to the model and evidence to support adaptations.

Change/addition	Personal/contextual risk/protective factor	Participants reasoning and reference to relevant supporting literature
Smoking	Personal risk factor Contextual risk factor	Participants in all focus groups reported that the emotional dysregulation associated with nicotine withdrawal (physiological and psychological) increased risk of aggression. In addition to withdrawal and the loss of a coping mechanism, participants described leave denial or postponing of leave to smoke or vape contributed to aggression. A systematic review by Spaducci et al. (2017) and an interrupted time series analysis by Robson et al. (2019) found no link between a comprehensive smoke free policy and an increase in physical or verbal violence, which is contradicted by Weltens et al. (2021) who found that nicotine withdrawal symptoms are associated with aggression. There is evidence to support the link between denial of requests and aggression in the study by Papadopoulos et al. (2012). There is limited evidence on the association between granting or denial of leave and aggression, however Barlow and Dickens (2019) found through a systematic review that patients should experience a robust, transparent, systematic process with clear decision‐making protocols to reduce the potential for conflict.
Acute intoxication/withdrawal	Personal risk factor Contextual Risk Factor	Multiple focus groups experienced an increase in aggression secondary to intoxication at the time of admission or on return from leave. Another contributing factor experienced by participants was an increase in aggression secondary to withdrawal. Evidence supporting these experiences is mixed. Stewart and Bowers (2014) found that although there is an increase in verbal aggression, there was no association with physical aggression. In contrast, Weltens et al. (2021) found an increased incidence of aggression within the first 3 days of admission associated with intoxication and/or withdrawal. Furthermore, substance use is found to be a primary contributing factor in aggression within emergency departments (Kleissl‐Muir et al. 2018). This aligns with the experience of our focus groups.
Neurodiversity, Intellectual Disability (ID), Cognitive Impairment and Acquired Brain Injury (ABI)	Personal risk factor Contextual risk factor	Multiple focus groups highlighted that the mental health hospital environment is not suited to those with neurodiversity, intellectual disability, cognitive impairment or acquired brain injury. Hospital environments are often loud, unpredictable and lack privacy. The physical layout of mental health inpatient units was also described as contributing to aggression, specifically for those with age related cognitive impairment or decline. One participant explaining, an ‘L shaped’ unit for those prone to ‘wandering’ can lead to aggression when consumers feel are ‘trapped’ in corners. The three most common reasons for neurodiverse people accessing mental health care are anxiety, self‐harm and aggression (Hill 2022). Girasek et al. (2022) found that dementia, mild cognitive impairment and intellectual disability contributes to aggression risk in mental health settings. In the absence of neurodiversity affirming environments and specialist, evidence‐based interventions the risk of aggression in this population remains high (Pantazakos and Vanaken 2023)
Homelessness	Personal risk factor	Participants, especially those from regional settings raised that homelessness contributes to aggression in the inpatient setting due to increased length of stay. Weltens et al. (2021) found no causal link between homelessness and aggression, however Russolilo et al. (2023) and Fletcher et al. (2021) found increased length of stay is associated with an increased incidence of aggression
Staff cognitive bias, being ‘primed’ for a bad shift and fear	Contextual risk factor	Participants described staff being ‘primed’ for a ‘spicy (bad) shift’ though constant communication between staff on shift to those coming on to shift via text messaging and ‘group chats’. Participants drew a link between this and increased incidence of aggression due to the negative bias staff may develop towards the work environment and patients in response to these communications. Lim et al. (2017) found that negative influences on the therapeutic alliance and therapeutic optimism can hinder effective de‐escalation and other interventions aimed at reducing incidents of aggression
Staff distraction/decreased situational awareness	Contextual risk factor	Multiple focus groups with participants representing a diverse range of settings observed staff distraction and lack of presence on the unit increases incidents of aggression. Participants made specific reference to the presence of smart phones contributing to distraction. According to the Safewards model, in the patient community domain “the presence of staff and their good relationships with patients (presence+) allows intervention at an early stage of potential arguments, with diplomatic negotiation or other action averting irritations that may otherwise later turn into violence” See: www.Safewards.net
Shared Rooms	Contextual risk or protective factor	Some focus groups considered shared rooms may contribute to aggression secondary to lack of space and privacy. There is little evidence to support this hypothesis, but Shepley et al. (2016) found that ‘de‐institutionalised’ environments which include having private rooms increase patient and staff safety.
Cultural Diversity	Person risk or protective factor/contextual risk or protective factor	Participants noted that miscommunication and misunderstanding may contribute to aggression and provided the example that some communication of distress by people from certain ethnic and/or cultural groups may be misinterpreted by staff and other patients as aggressive. This may result in restrictive interventions that may lead to aggressive reactions by patients and/or interpersonal conflict and aggression with other patients. By contrast, Lansford (2018) found that culture may help prevent aggression over time, with the example that aggression is less stable from childhood into adulthood when comparing Finland with the United States, hypothesizing that this is due to differing social safety net resources available to families and communities.
Supportive relationships with family, carers, supporters and community	Personal protective factor	Focus groups identified that positive relationships with family, carers and supporters helps prevent aggression. This is supported by Labella and Masten (2018) who found that the provision of warmth, structure, predictability and other adaptive processes by parents and caregivers reduced aggression within young people
Safety Culture and Early Intervention (vs staff avoidant of setting boundaries)	Contextual risk or protective factor	Focus groups highlighted the need for inpatient units to develop and maintain a safety culture, with staff presence and early intervention or boundary setting to prevent aggression. Positive collegial relationships and psychological safety are important enablers of a safety culture within mental health inpatient units (O'Donovan and McAuliffe 2020). Staff presence and engagement with patients was identified by Molloy et al. (2024) as being critical to safety culture, aligning with participant's experience that staff presence is essential for early intervention and the prevention of aggression
Therapeutic milieu	Contextual Risk or Protective Factor	Participants working with young people identified therapeutic milieu as being an important factor in preventing aggression within mental health inpatient units. This is supported in the literature as reducing incidence of not only aggression, but also self‐harm (Belsiyal et al. 2022)
Power dynamics, including intergenerational discrimination and disempowerment experience by Aboriginal and Torres Strait Islander peoples	Personal risk factor	Uneven power dynamics found in mental health inpatient settings between staff and patients was identified as a potentially important contributor to aggression. This is supported by Fletcher et al. (2021) who found that patients said that coercion or the perception of coercion can precipitate aggression. Martin et al. (2015) found that trauma is associated with aggression for people with mental health issues in prison settings; however, their findings were inconclusive when comparing Aboriginal and non‐Aboriginal participants
High staff turnover		Occupation violence and aggression is strongly associated with high levels of staff turnover in many health care settings, including mental health inpatient units (Ose et al. 2023). The participants in this study identified that high staff turnover, resulting in fatigue or burn out in remaining staff and an imbalance between experienced and novice nurses increases the risk of aggression. These risk factors are present in the current version of the The Model for Understanding Inpatient Aggression (MUIA).
Creating a poster and/or infographic version of the MUIA	Not applicable	Participants in multiple focus groups raised the idea of having a more easily accessible version of the MUIA in a poster and/or infographic format to be placed in the nurse's station or offices. There is limited evidence on the use of poster and/or infographics for concepts such as the MUIA, however Kong et al. (2024) found that infographics are not only an effective means of communicating knowledge of change behaviour, but preferred over other mediums (Martin et al. 2018)

5. Discussion

This study explored the Model for Understanding Inpatient Aggression with staff who work in mental health inpatient units as well as a lived experience expert to determine if the model might be helpful to understand, prevent, and manage aggression in general/civil mental health units. This study has drawn attention to current practice, which tends to focus on personal factors and post‐incident review of aggression with scant emphasis on working to understand a range of contributing factors that could help prevent aggression. Furthermore, findings from this study showed that many of the factors contributing to aggression are extrinsic; the model's elaboration of these factors and encouragement to consider their relevance helps build a more comprehensive understanding of aggression. In turn, this may lead to a greater array of intervention options. Additionally, the current study suggests the Model for Understanding Inpatient Aggression may fill a current practice gap by encouraging exploration of unit and staff factors as targets for interventions (Weltens et al. 2021).

The focus on post‐incident exploration rather than prevention and early intervention may not be surprising, given there remains a tendency for services to focus on the teaching of restrictive practices in the prevention and management of aggression training. A recent study by Maguire et al. (2023) exploring a structured method for nursing intervention following risk assessment found that while nurses considered the teaching and assessment of skills such as limit setting and de‐escalation important, the reality was that services still prioritise the teaching and assessment of skills associated with applying restrictive practices, including physical restraint. A shift in focus to prevention techniques and ways of understanding the array of factors that contribute to aggression may be necessary to reduce the use of restrictive practices and prevent aggression in mental health settings.

Regarding the physical features of the unit, there is evidence that suggests design and architectural features such as a ‘home‐like’ environment in mental health inpatient units can ease patient stress, enhance social interaction, foster autonomy and perceptions of control, ensure privacy and dignity, and prevent under‐and or overstimulation (Bodyrzlova et al. 2024). Home‐like features may not always be recognised in interior design choices (e.g., stark white walls are common in mental health units), and there is little opportunity for patients to personalise the spaces due to policy and procedural requirements.

In terms of participants identifying that intrinsic patient factors and post event exploration are prioritised, other studies have also noted a preoccupation with factors intrinsic to patients, with a need for more focus on the earlier stages of escalation and factors that are responsive to prevention efforts, including unit and staff factors (Welsh et al. 2013; Weltens et al. 2021), hence the need to “turn the mirror on ourselves”. When considering the extrinsic factors that may contribute to aggression, participants in this study identified issues such as problematic communication skills, or inadequate leadership by some colleagues contributing to aggression. Addressing issues such as poor leadership and communication skills can be difficult issues to raise with colleagues. A recent study by Cranage and Foster (2022) exploring challenging workplace situations in mental health nursing identified colleague related factors (in addition to patient‐related, nursing role‐related, and service‐related challenges) as causes of moral distress for mental health nurses. Colleague related factors included circumstances involving poor communication and perceived poor practice/quality of care (Cranage and Foster 2022). Examination of unit and staff factors could assist managers to address staff and unit related issues where indicated (Hamrin et al. 2009; Welsh et al. 2013; Weltens et al. 2021) and the Model for Understanding Inpatient Aggression in this study was seen as a mechanism to assist raising issues related to colleagues practice when this may not be ideal.

Given that the interpersonal behaviour of staff, including initiation of prevention efforts, can either exacerbate or de‐escalate aggression and violence (Fletcher et al. 2021), it may also be important to ensure staff have a sound understanding of how contextual factors can also contribute to violence and, because these factors are within their control, may be easier to modify than intrinsic patient factors (Pelto‐Piri et al. 2020). In addition, services also need to ensure there is an emphasis on sustaining open communication and professional relationships to ensure positive team and unit culture and the identification of, and attending to, problematic practice (Cranage and Foster 2022), as well as ensuring staff are skilled in some of the more complex communication techniques such as de‐escalation and limit setting (Maguire et al. 2023).

The therapeutic relationship is the cornerstone of mental health nursing intervention (Hartley et al. 2020), and is essential to effectively undertaking more complex interventions, including de‐escalation and limit setting, which are essential to reducing incidents of aggression (Maguire et al. 2019). When inpatient units are loud, chaotic, task saturated, and risk focused, they may not be conducive to the provision of personal recovery for patients and successful application of nursing tasks (Jørgensen et al. 2022; Maguire et al. 2023).

Clinical supervision can reduce nurses frustration and increase empathy towards patients (Tuck 2017), supporting nurses in developing and maintaining the therapeutic relationship. It also supports the development of professional competencies, self‐esteem, increasing the effectiveness of care (Butterworth 2022; Buus et al. 2011; Snowden et al. 2017). It also enhances nurses' communication and clinical reasoning skills (Hudays et al. 2014) better equipping nurses to navigate emergency situations (including incidents of aggression), prioritising patient wellbeing and safety (Mann et al. 2009). Ongoing, facilitated reflection on decision making through clinical supervision assists with the re‐establishment of professional self‐esteem and competency, critical following an incident of aggression (Howard and Eddy‐Imishue 2020).

When undertaken in conjunction with education or training, it supports the translation of theoretical knowledge into practice (Heaven et al. 2006) and should be considered when initially introducing the Model for Understanding Inpatient Aggression or similar models. Furthermore, clinical supervision improves communication and teamwork (Hamilton et al. 2023), with group supervision in mental health inpatient settings being found to create a sense of solidarity, whilst giving space for difficult conversations (McDonough et al. 2024). Such approaches could be utilised for examining and reflecting upon contextual factors of the Model for Understanding Inpatient Aggression, as suggested by participants in this study.

5.1. Strengths and Limitations

This study was limited to the exploration of the Model for Understanding Inpatient Aggression with staff and one person with lived experience of mental health problems from across blinded, and therefore may limit generalisability to other services and areas that may differ in practice and population. Although we wanted to recruit more people from the lived experience perspective, there was a staff shortage at the time, limiting recruitment. People in these roles in our state tend to be in high demand for consultation and are therefore stretched in terms of being able to participate in projects such as this; however, we welcome their involvement in future research. A strength of this study was the involvement of a range of skilled nurses in clinical and leadership positions who had expert knowledge and skills in understanding and preventing aggression, as well as lived experience expert input.

6. Conclusion

Understanding and preventing inpatient aggression remains a challenge for nurses, other staff, patients and services, and may prove difficult to address when the focus remains on patient factors and post‐incident reviews rather than ensuring attention is also paid to near miss and early intervention techniques. Models that incorporate contextual factors are important for services to consider as part of an approach to understanding and preventing aggression and may be where positive actions to address aggression can have the greatest impact. Despite the Model for Understanding Inpatient Aggression being derived from a forensic mental health perspective, results from this study suggest that the model may also be suitable for area mental health services and fill a current practice gap.

6.1. Relevance to Practice

Ensuring mental health nurses have a comprehensive model to understand aggression with consideration to the unique patient factors in conjunction with interpersonal, structural and organisational factors might assist movement from a reactivity to proactivity. Additionally, highlighting the complexity and importance of interpersonal skills may prompt organisations to consider the emphasis of these skills in the prevention of aggression training, and how ongoing skills and refresher training might be undertaken to ensure staff are competent. Working with patients to understand their perspectives of the contributors to aggression and the unit setting is also important to ensure their views are captured and to enhance collaboration.

Note: Please see (https://www.linkedin.com/groups/13057210/) link for a copy of the model.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.