Problem

Inpatient secure services use a range of restrictive practices to keep patients safe from experiencing or causing harm to self or others. These include the use of restraint, seclusion and rapid tranquilisation, as well as wider practices, such as imposing blanket bans that restrict a person’s liberty and other rights, for example, stopping them from accessing outdoor spaces.1 While the intended purpose of these restrictive practices is to keep patients and others safe, their use must be balanced against the impact that they may have on a patient’s privacy and dignity.

Seclusion is defined as ‘the supervised confinement and isolation of a service user, away from other service users, in an area from which the service user is prevented from leaving, where it is of immediate necessity for the purpose of the containment of severe behavioural disturbance’.2 Although it is difficult to ascertain the exact figure, seclusion is prevalent across mental health settings, with approximately 520 individuals secluded in mental health settings in October 2021.3

The project took place on an acute medium secure forensic inpatient ward for male service users admitted for assessment or treatment under the Mental Health Act.4 5 It accommodates up to 15 service users with a wide range of complex needs. Being a forensic unit, the service users on the ward have coexisting forensic needs alongside mental health conditions.

Our project formed part of a trust-wide positive and safe Quality Improvement (QI) programme, focusing on reducing seclusions (frequency and/or duration) across selected wards in Oxford Health NHS Foundation Trust. This was part of a focus of a national health safety improvement programme and a priority for mental health services.6

Members of the public were not included in this research. Inpatients were involved at several stages of the trial, including feedback relating to their experiences. We received input from inpatients who had lived mental health problems in gathering our baseline information and following the tests of change. We carefully assessed the burden of the participation on the inpatients involved in the research. We intend to disseminate the main results to the inpatients involved using an appropriately decided method of dissemination.

Background

Seclusion, the practice of isolating a service user away from the main ward, is considered a restrictive practice in several countries including the UK. Seclusion is enacted to keep everyone on the ward safe during incidents where a service user might be emotionally distressed or dysregulated. Best practice guidance states that the use of restrictive practices should not be used to ‘punish, inflict pain, suffering or humiliation, or establish dominance’.7 However, evidence into the use of seclusion indicates that it can traumatise both service users and staff; moreover, it can hold little therapeutic benefit.8–12 The experience of seclusion can make service users feel angry, lonely, sad, hopeless, punished and vulnerable.13 14 Such emotional reactions are counterintuitive to the aims of a recovery-orientated and trauma-informed practices, which both aim to enable agency, self-direction and hope.14

Service user attitudes towards restrictive practices demonstrate that many feel seclusion is used too frequently and is related to environmental factors (eg, boredom) and poor communication between staff and service users.15–17 Restrictive practices can disempower service users who feel their voice is lost in the process.18 Often the use of seclusion is experienced negatively, viewed, deemed as controlling and should be stopped.18 19 However, evidence shows that some service users agree that when a service user is violent, it can be one of the more effective approaches in managing such risky behaviours.14

Secure inpatient services typically use the Trinitarian Model when carrying out safety planning, including the use of restrictive practices such as seclusion. A whole system approach to using seclusion encompasses a balance between considering the environment (safety vs therapeutic) and the balance between a reactive approach to procedural security versus a proactive approach to relational security.20 21

There are multiple factors that can impact whether staff focus on procedural or relational approaches to using seclusion. Staff attitudes towards restrictive practices are frequently used by staff as a means of safety for both patients and staff, while also providing therapeutic benefits.15 Staff were also more likely to seclude service users because of their internal perceptions of the patient’s heightened risk, harm to self, harm to others, agitation, substance misuse, psychotic symptoms and damage to property.15 22–24

Broader internal factors and ward culture can also influence a tendency to endorse a more managerial approach to using seclusion. For example, nurses who feel more anger typically sanctioned the use of seclusion, whereas those who felt higher levels of guilt were less likely to sanction its use.25 Contextual factors such as team confidence, climate, culture and organisational instability influence decisions regarding seclusions and seclusion rates.26–28 Taken collectively, research demonstrates the importance of a multilevelled approach that considers patient, staff and organisational factors when exploring reducing seclusions.

Research into strategies which facilitate a reduction in the use of seclusion highlights the importance of education and raising awareness of the impact of seclusion on both service users and staff. Suggested topics for education packages include early assessment, trauma-informed care, effective communication, de-escalation skills and co-production.22 24 A focus on the holistic overview of a patient’s care (incorporating internal, external and situational factors) and good management oversight was also considered effective in reducing the use of seclusion.16 22

Further strategies identified in the research highlight the need for robust communication before, during and after a seclusion event,19 the application of structured risk assessments,29 use of alternative restrictive measures including intermuscular injection and time-out practices13 30 and the introduction of a policy ending night confinement where patients are allowed to exit the seclusion room overnight.31

Measurement

The NHS England National Service User Safety Improvement Programme3 6 actuated an audit process which explored the rationale and use of seclusion alongside the outcomes of restrictive practices from January until July 2021. For our ward, we gathered service user and staff feedback, alongside descriptive statistics relating to the time and duration of seclusion episodes using the Trust Online Business Platform (TOBI). This data was subsequently used to inform the QI multidisciplinary team (QIMDT) change ideas.

We determined that the baseline measure for the project was a reduction in the duration (outcome measure) and frequency (balancing measure) of the use of seclusion by 10% over 6 months which was the agreed percentage throughout the Trust. The rationale behind these baseline measures was that a reduction in the use of seclusion would likely also result in a reduction in the frequency. As duration and frequency of seclusion are both important, we focused our change ideas to consider how we may reduce both with the overall vision of reducing the use of seclusion. The baseline measures would be complemented by further feedback from the staff team and service users following the implementation of the project.

The team held regular meetings throughout this period to discuss the data and any particular cases that may have skewed the findings. This was not the case during this period, but we have discussed an experience more recently in our lessons and limitations on the topic of sustainability.

Design

Our project team (QIMDT) was made up of professionals representing a range of disciplines on the ward including medics, nursing, occupational therapy and psychology. We followed the Insitute for Healthcare Improvement model32 which provided us with an opportunity to understand our problem, shape a SMART (specific, measurable, achieveable, realistic, timebound) aim and appropriate measures and take a systematic approach to identifying the different change ideas that we could consider. Supported by Oxford Healthcare Improvement, the QIMDT met once a week on the ward in the initial stages of the project to determine the parameters of the project, create our driver diagram(online supplemental file 1) and define the change ideas. We used the feedback from our audit and the baseline data from TOBI to determine the driver diagram and change ideas. This data indicated a high use of seclusion in the preceding 6 months on the ward. These high figures were attributed to misunderstandings regarding the rationales for the seclusion or de-seclusion of service users among the staff team. Additionally, consultants were often anticipated to be part of the decision to seclude or de-seclude leading to longer than necessary periods of restriction over periods such as weekends and bank holidays when access to consultants was lessened. Feedback from service users indicated that they felt that seclusion periods were too long and that they would often feel ‘bored’ following seclusion due to a lack of stimulating activities offered.

Alongside the use of our own data, we also ensured attendance of members of the QIMDT at a twice monthly ‘Positive and Safe meeting’ (as part of the Patient Safety Improvement Programme6) targeted at all wards using seclusion across our Trust, during which the rationale for seclusion and outcomes were reviewed. We were also able to access specific bespoke support for our project through the Forensic QI hub including a specific training and discussion event for our project with two members of the QI hub. The wider ward team were briefed throughout the project’s duration (starting in February 2021) through various outlets including weekly staff team meetings and a dedicated QI board placed in one of the offices on the ward. This was part of the first test of change in relation to raising awareness of the project.

Strategy

Our driver diagram, created by the QIMDT, guided the structure and establishment of change ideas and these were tested using PDSA (plan, do, study, act) cycles. Alongside the data related to the use of seclusion, we also continued to gather staff and service user feedback. We broke our approach down into three categories: raising awareness, reviewing seclusion processes and enhancing de-escalation skills.

Lessons and limitations

Our key points of learning from the project include the importance of MDT working, ward team ownership of the seclusion process, improvement in communication between staff and service users and the need for a range of activities to occupy service users held in seclusion. Providing the space and opportunity for the MDT to coalesce around this project and create a shared impetus for change provided a platform for psychologically safe discussions, innovation through change ideas and coherence with the direction of the project. The empowerment of the ward staff team was a crucial element in the project as it does involve making changes to existing practices. Bolstering the skills and confidence to shift towards therapeutic risk taking, such as more frequent use of de-escalation is essential for promoting staff–patient relationships and promoting better patient outcomes and experiences. As well as increasing skills and confidence, part of empowerment can be about giving permission to act. Reiterating to ward teams that they did have the skills to make decisions to end seclusions gave them a sense of agency to come together and take ownership in that aspect of service user care.

The limitations of our project include the short period of time we implemented our change ideas, and this was partly due to the enthusiasm of the team to make a change. This made it challenging to discern which tests of change (and PDSA cycles) were more effective than others and, with hindsight, we acknowledge that we could have been more systematic and slowed down the rollout of the tests of change. We also identified that further improvement is needed in collecting feedback from service users regarding the use of seclusion. As with every busy clinical environment, workplace pressures are likely to affect some of our effectiveness of measures of recording incidents. We are also aware that further efforts should be made to improve the service users’ access to activities while in seclusion.

We acknowledge that sustainability can be an issue when working in environments with high staff turnover. To mitigate some of the impacts of this, we ensured that the aims of the project were imparted to new and temporary members of staff through induction, regular discussions during staff meetings and seclusion reviews. This helped to facilitate a culture of empowerment among the staff team in managing seclusions in a challenging environment.