Safeguarding patients while implementing mechanical restraints: A qualitative study of nurses and ward staff’s perceptions and assessment

Abstract Aims and Objectives To explore nurses’ and ward staff's perceptions and assessments of patient care while implementing mechanical restraints. Background To prevent the risks associated with the use of restraints in psychiatry and ensure safe mental health care, it is necessary to know more about how the nursing staff experiences, comprehends and intervenes in managing patients subjected to coercive measures. Design This study employed a qualitative descriptive design, in accordance with the COREQ guidelines. Methods Semi‐structured interviews were conducted with 18 nurses and ward staff aged between 22 and 45 years old, who had experience implementing mechanical restraints. Data were digitally audio‐recorded and transcribed verbatim. Inductive thematic analysis was conducted using NVivo 12. Results The participants believed that mechanical restraints should be used as a last resort and that safeguarding patients during implementation is important; however, their assessments of the patients’ physical and mental conditions varied. A clear difference emerged in how management qualified professionals handled situations prior to and during the implementation of mechanical coercive measures. Conclusions The findings emphasise the need to focus on the assessment of patients prior to and during restraint, ensure the quality of safe implementation in a risk‐of‐harm situation, prioritise competence in education, and practice, and improve management. Relevance to clinical practice The findings highlight the importance of assessing the physical and mental condition of patients while implementing restraints, as well as aiding the management, nurses and ward staff in tailoring safety procedures.


| INTRODUC TI ON
The use of coercion in mental health involves ethical, legal and clinical issues. Hence, reduction of restraint and safe mental health care are listed as objectives in the European Mental Health Action Plan (WHO, 2015). Recently, the use of restraints in mental health services has been included as one of the priority areas in the Norwegian patient safety program 'In Safe Hands 24-7' (Meld. St.10, 2012-2013. Although the primary purpose of using restraints is to prevent unsafe behaviour, violence or self-harm (Psykisk helsevernloven, 1999), it could cause additional risks to patients' safety during the intervention (Thibaut et al., 2019). The use of mechanical restraints, pharmacological restraints or a combination of these two may potentially have deleterious physical and psychological consequences for patients (Chieze et al., 2019;Kersting et al., 2019), putting nurses and ward staff in a paradoxical situation, due to the risk of harm from the patients themselves.
In the international context, Norway has a relatively high rate of involuntary admissions in inpatient psychiatric departments (Salize & Dressing, 2004;Wynn, 2018). The Norwegian Directorate of

| BACKG ROU N D
The concept of 'restraint' involves several techniques and degrees of coercion and implementation. Internationally, 'restraint' in psychiatry is also known as 'physical restraint', which could be either manual or mechanical; chemical restraints are also addressed (Negroni, 2017).
A review highlights that no jurisdiction combines the full suite of laws, policies and practices that together might further the goal of eliminating coercion (Gooding et al., 2020). A study by Steinert et al. (2010) reports the incidence of mechanical coercive measures in psychiatric hospitals of countries such as Austria, Finland, Germany, Japan, the Netherlands, Norway, Spain and Switzerland. Mechanical restraints are forbidden for ethical reasons in countries such as the UK and Iceland (Steinert et al., 2010). Even though many countries allow the use of coercive measures, many countries work towards specific health policies that promote alternatives to coercive practice (WHO, 2021). In this context, several studies point to international initiatives to reduce the use of restrictive practices, such as the Six Core Strategies program, the Safe Wards Model (Fernández-Costa et al., 2020;Goulet et al., 2017;, de-escalation techniques and risk assessment (Fernández-Costa et al., 2020). A Danish review (Bak et al., 2012) mentions three interventions that most likely reduce the number of mechanical restraints: cognitive milieu therapy, combined interventions and patient-centred care.
In Norway, coercive measures are not regarded as part of treatment or therapy but as an extraordinary intervention in mental health care, which can only be applied to prevent patients from injuring themselves, assaulting others, or damaging buildings and physical objects and can be used regardless of patients` legal status when they are admitted to a mental healthcare facility. However, less restrictive interventions must first be proven unsuccessful. Mechanical restraint, according to the Norwegian Mental Health Care Act (1999, §4.8.), refers to different kinds of belts (i.e. restraint in a bed or outside of a bed for arms and feet only). According to the Norwegian legislation (Psykiskhelsevernloven, 1999), healthcare staff should continuously inspect patients subjected to mechanical restraints.
While the legislation does not specify the competence level of the person who continuously assists the patient being restrained, the nursing staff mainly comprises mental health nurses, enrolled nurses, social educators, social workers and healthcare assistants.
The use of mechanical or pharmacological restraints-or a combination of the two-may have deleterious physical and psychological effects and outcomes in patients. A review found that the restraint position (e.g. the prone position used in Norway) can impede lifemaintaining physiological functions (Barnett et al., 2016). Physical injuries can cause respiratory and circulatory problems as well as nerve damages, more specifically through impaired consciousness, acute pulmonary oedema, respiratory arrest, loss of consciousness, aspiration and the release of stress hormones that may affect heart rate, rhabdomyolysis and thrombosis (Berzlanovich et al., 2012;Jönsson et al., 2018;Mohr et al., 2003;Nelstrop et al., 2006). Psychologically, coercive measures can activate emotions and have a retraumatising and traumatising effect (Cusack et al., 2018;Kersting et al., 2019).
Mental health nursing has traditionally focused on 'interpersonal relations in nursing' (Crenshaw & Peplau, 1952). Crenshaw and Peplau redefined the role of mental health nurses based on a more client-focused perspective on what occurs in the interplay between What does this study contribute to a wider global clinical community?
• The quality of safe implementation of restraints is lacking.
• Health authorities should establish national guidelines for the follow-up of patients during mechanical coercive measures to ensure that nurses and ward staff are better prepared. There is a need for greater value and recognition of this aspect in practice. the patient and the nurse and the therapeutic use of the professional self (Crenshaw & Peplau, 1952). The use of the professional self and assessments of patients is illuminated in a Danish study by Nielsen et al. (2018), who reported on forensic mental health clinicians' experiences and assessments of the patients' readiness to be released from mechanical restraints. The results highlighted the clinicians' assessment of the quality of the alliance based on two parameters: (a) the patients' insight into or understanding of the present situation and (b) the patients' ability to maintain a good, stable relationship and cooperate with clinicians. These evaluations are included in the bigger picture of the assessments of patients' readiness to be released from mechanical restraints (Nielsen et al., 2018). A recent qualitative study reported the impacts on the professional relationships when patients stopped engaging positively with healthcare professionals after restrictive practice, exacerbating the professionals` feelings of guilt (Mooney & Kanyeredzi, 2021). Today, the field of mental health nursing suggests different levels of assessments related to basic physiological needs, symptoms experienced (Barker, 2004), and elements such as risk, physical and mental status, symptomatology and effects of medication (Gamble & Brennan, 2005).
A focus group study (Muir-Cochrane et al., 2018) reported on the changing role in acute settings, focusing more on risk assessment and medication while attempting to practice trauma-informed care.
A Canadian review of mental health care and practice revealed the lack of a systematic approach to patient safety (Brickell & McLean, 2011). Marcus et al. (2018) emphasised the need for research that examines the physical, emotional and psychological outcomes associated with patient safety events, stating that patient safety should become a key part of inpatient psychiatry (Marcus et al., 2018). In the Norwegian patient security program 'In Safe Hands 24-7', the use of restraints in mental health services was one of the priority areas that aimed to reduce the use of restraints in psychiatry. Countering action in the reduction and safer use of coercion, a national comprehensive training program 'MAP'-Model  Cusack et al. (2018) exploring available evidence about the physical and psychological impact of restraints concluded that further research is required regarding mental health settings. Moreover, the researchers assumed that mental health nurses were in a prime position to use their skills for addressing the identified issues to eradicate the use of restraints and effectively meet the needs of those experiencing mental illness (Cusack et al., 2018). A systematic review by Douglas et al. (2021), reporting on the qualitative evidence from the perspective of patients experiencing physical restraints, identifies a bio-psychosocial impact, in addition to an impact on therapeutic relationships and patients' needs. The studies included mostly focused on the psychological impact, which was negative and connected to trauma or retrauma. A few studies have reported on the biological impact of pain and injury expressed through fear and distress. Conversely, studies addressing patients' needs during the use of restraints, which highlighted the need for attention to patients' comfort and safety, are sparse due to the practice of the restraint position and of asking the patient to sit up in order to breathe more easily (Douglas et al., 2021).

| Aim
This study aimed to explore nurses` and ward staff's perceptions and assessments of patient care during the implementation of mechanical restraints.

| Design
This study explored the clinical practice of mechanical restraint among nurses and ward staff using a qualitative approach with a descriptive design inspired by Sandelowski (Sandelowski, 2000). A qualitative description is amenable to obtaining straight answers to questions of special relevance to practitioners and entails the presentation of findings in the practitioner's everyday language.
A qualitative descriptive design aims to produce findings close to the data (Sandelowski, 2010) and facilitates an exploration of the participants' experiences and views that are relevant to glean rich descriptions of the phenomenon being studied (Polit & Beck, 2020).

This study complied with the Consolidated Criteria for Reporting
Qualitative Studies (COREQ) checklist (Tong et al., 2007; File S1).

| Sample/Participants
We recruited a purposeful sample (Patton, 2015) of nurses and ward staff working at two acute psychiatric wards in two Norwegian hospitals. The inclusion criterion was having experience in the follow-up of restrained patients. The participants ( Table 1) were recruited by two nurse managers, with the aim of including those with maximum variations in terms of profession, specialised education in mental health care, sex, age and experience in the follow-up of patients during the implementation of mechanical restraints. We estimated a sample of 15-20 participants as sufficient to gain information power and guided by Malterud et al.'s (2016) items and dimensions. The final sample was determined during the data collection process, with 18 participants assessed as having sufficient information power to match the aim of the study and the sample being dense enough for specificity (Malterud et al., 2016).

| Data collection
Data were collected through semi-structured individual interviews conducted between August and October 2020 by the first author, LB. As an introduction, LB presented herself as a PhD student with previous clinical experience in psychiatry. Each interview started with the question: 'What types of situations can lead to the implementation of mechanical coercive measures?' Participants were also asked about their experiences with observations and assessments of the patient under mechanical coercion and how they took care of the patient in such a situation. The interviews were digitally recorded and transcribed verbatim. Due to the COVID-19 situation, interviews were conducted using Zoom, so that the interviewer and the participants could see and hear each other. The interviews lasted between 28 and 50 minutes. The interview guide was developed by the research team, which included all females with previous clinical experience in psychiatry. The interview guide was piloted in collaboration with educators from a university college who had previous experience in the follow-up of patients being restrained within the context of acute psychiatric care.

| Ethical considerations
The Norwegian Data Protection Official for Research (NSD) approved this study (reference number 503297). All participants received written and oral information about the requirements for participation in the study and the depersonalisation of the data; in addition, they were informed that participation in the project was voluntary, and they provided written informed consent to participate. Consent could be withdrawn at any time without giving any reason and without any negative consequences, and this would not influence the participants' relationships with the

| Data analysis
Anonymised and verbatim transcribed interview data were uploaded to NVivo 12. An inductive, data-driven thematic analysis process was chosen, based on four analytical phases, which were routed into a thematic analysis using six steps (Terry et al., 2017).
In Phase 1, 'familiarizing yourself with your data' became important due to the data collection process, with several interviews re-read and discussed among the research team members (LB, SV and IPM). In this phase, the nurses and ward staff's descriptions of a highly complex and challenging situation and the unanimous view that implementing mechanical restraints 'is the last resort we make' became visible.
In Phase 2, we focused on 'generating initial codes' systematically across the entire data set, collating data relevant to each code.
This phase was completed twice, ending with 176 semantic codes, implying that they captured the surface meaning of the entire data-  Lincoln and Gubas's (1985) criteria for trustworthiness were used to establish rigour. The first author, LB, led the analysis and engaged in peer-debriefing with co-authors SV and IPM from the 'familiarizing phase' and throughout the analytic process to establish credibility (Lincoln & Guba, 1985). The researcher provided transferability through detailed descriptions of findings and verbatim quotes, aiming for readers from other care settings (Lincoln & Guba, 1985) to evaluate the extent to which the findings apply to new situations (Polit & Beck, 2010). Maintaining a research log of the study processes established dependability and confirmability (Lincoln & Guba, 1985).

| FINDING S
The study participants from two Norwegian acute psychiatric wards (N = 18) comprised nine nurses, four social educators and five ward assistants aged between 22 and 45 years old, whose experience of implementing mechanical restraints ranged from 1 to 16 years ( Table 2). This analysis generated four main themes ( Figure 1).

| Mechanical restraints: A safety measure but to be used as a last resort
Many participants, regardless of profession, stated that while mechanical restraints can be used as a safety measure in situations that escalate into violence, aggression or self-harm, they should be used as a last resort. In most cases, there is a long duration between the starting point of agitation and the point where a patient must be placed under mechanical restraints. The participants expressed differing opinions regarding how different situations were handled, stating that each situation was assessed individually. One of the nurses explained mechanical coercion as being based on legal requirements for implementation.
It is very individual, but you must note the basis for the decision in terms of being dangerous to others, danger to oneself, or attack on others. Therefore, it is a kind of last resort; every option must be tried first.
(Jordan, Nurse 8) The participants also had a perception that mechanical coercion was not only negative, but also a possible measure that contributed to making the nursing staff experience security.
Some participants mentioned stricter norms for the implementation of coercive measures in the psychiatric treatment system; therefore, much effort was expended to avoid the use of mechanical coercion. The participants described it as a step-by-step approach, with a strong focus on identifying situations and the active use of Actively involving the patient in de-escalating options could also be a successful approach: I remember a patient who acted on inventory (furniture and fixtures) and was threatening the staff.
He was told to go to his room and calm down. When the staff experienced failure in the verbal process, the next step was to implement short-term detention (manually holding the patient), isolation or the use of conveyor belts. Several participants described that the duration for the more lenient coercive measure-short-term detention-had changed from a short timeframe of a few minutes to an entire hour. This led to discussions among the staff who experienced the situation and was perceived as a theme to be questioned. Sometimes you know that the patient does not profit from short-term detention. Maybe he has been abused or gone through many bad things-then we would rather go and get the mechanic restraint quickly.
Apart from the descriptions of many elements that came into play in the assessment, the patients' own wishes could affect the choice of coercive measure in terms of whether they prefer being 'held' for a long time or being directly restrained. The participants also stated that in their medical record, some patients explicitly conveyed their desire to use mechanical coercion instead of short-term detention, which was cleared with ward management.

| Working actively through dialogue to secure the patient
Once the decision on using mechanical coercion is made, and the patient is secured with a safety belt, the focus shifts from securing to further de-escalation and gaining control of the situation. Many participants emphasised the importance of conveying information to the patient to 'secure' them and prevent further trauma. One of the participants described the importance of tuning in to the patients` words and employing suitable language: It is in a way useful to try and reflect or describe how the patient appears, like, 'Now you have become so angry, and we perceived you to be very threatening, and we do not want you to get into a situation where you are dangerous to yourself or others'.
(Ariel, Nurse 1) The same participant also described basic behaviours with which to secure the patient, such as meeting the gaze, actively breathing with the patient, being observant of one's own voice and position, and taking control of the situation alongside the patient.
Some of the participants indicated that the patient should not be exposed to a vulnerable situation and be the subject of others` ob- Another observation expressed by several participants was that some patients wanted to go to the toilet immediately after being belted. They revealed the dilemma between releasing the patient in transport straps to go to the toilet and making the patient use a urine bottle or bedpan in the bed. Several of them also expressed that it was important to ensure that the patient received food and drinks.
One said that he was afraid to give the patient food that could get stuck in the throat, and another mentioned that it was important to always observe the patient because they could have underlying diseases and a greater risk of developing a blood clot. One of the participants talked about how he constantly observed and assessed the patient's breathing, awareness and confusion but that the doctor was responsible for prescribing the frequency of measurements: If we monitor vitals, the doctor decides how often we should take blood pressure, pulse, and respiration.
Then, we, on our own initiative, check the patients` skin status related to circulation sometimes.
(Hunter, Nurse 7) Although several participants expressed how they followed up on the patients' somatic condition during mechanical coercion, some particularly contrasting statements were made. When the interviewer summed up the description 'about use of your clinical gaze' and asked whether one of the participants had ever monitored blood pressure, counted pulse or respiration rate, the nurse responded that he never had been in a situation when he had felt it was necessary. Another stated: If someone is restrained for a long time, then you have to take blood pressure and monitor them; but if we do not manage, then it is not that important; it is rare they are restrained for so long.
(Drew, Nurse 4) One of the ward staff explained his task as follows: As soon as they are belted, the patient is checked

| Continuous assessment to discontinue coercive measures
Several participants reported that their observations and assessments of the patient aimed for the coercive measure to have the shortest possible duration. Some described observations of patients' behaviour, while others made assessments about the patient having 'calmed down' and not being angry anymore, with a lower severity of threats. Interactions with patients who were receptive to contact were also a part of their continuous assessment.
While some participants said that they tried to involve the patient by imparting information on how to proceed while being released from the restraints; others said that patients had to provide good contact, speak clearly and display calm body language before the belts could be loosened. Patients who did not provide appropriate emotional contact or voicing threats indicated that it was too early to end the restraint. Knowing the patient from previous admissions and being able to trust the patient had an impact on the assessment of whether participants could release them from the restraints. The lack of prior knowledge of the patient could mean less trust in the patient, which could lead to the patient remaining under restraints longer than necessary. Some participants said that they had previously misjudged whether the patient had calmed down enough, causing them to act out or commit selfharm, which had led to the reintroduction of mechanical coercion. The discontinuing process, which follows a gradual method of releasing the patients by, for example, loosening the belt on an arm and a leg, takes place after close dialogue with colleagues who bear formal responsibility for the actual shift. Some participants described patients who calmed down with medication and eventually fell asleep, which was an indication that they could be released from the restraints. One of the participants stated, 'The challenge is that there will be a kind of subjective assessment of when to end the decision. There is no clear template or objective' (Reese, Ward Assistant 2). Another described how the formal decision of determining of restraints was practised: It is us, the ward staff, who decide when to release the coercive measure when we feel it is safe to do so. We then notify the doctor. Finally, we must write down how long the patient has been in belts.

| D ISCUSS I ON AND IMPLIC ATI ON S
This study sought to explore nurses` and ward staff's perceptions and assessments of patient care during the implementation of mechanical restraints. The findings conveyed their views on restraints as safety measures, but as a last resort. Through participation in the MAP train- measures. The findings also revealed that personnel training in observation and assessment during the implementation was not anchored at the system level but was left to experienced nurses or social educators who were responsible for providing in-house training to their newly hired colleagues. We found variations among the participants in terms of their physical assessment and the extent to which they thought this was important. With regard to determining the coercive measures to be used, the assessment revealed individual discretion.
Different professional competences and experiences related to the assessment process were highlighted as challenges.

| Do we meet health authorities' goals?
The objectives of the European Mental Health Action Plan include a reduction in the use of restraints and safe mental health care (WHO, 2015). The findings in this study revealed a well-integrated view of coercive measures as a safety measure but as an absolute last resort in escalating situations. Thus, the care culture is aligned with international health policy goals that aim to reduce the use of coer- harm (Hem et al., 2018). The conflicting paradox when using coercion could lead to moral distress for healthcare professionals in their everyday practice (Hem et al., 2018).

| Imbalance in prepared competence before and during implementation
We obtained contrasting findings related to the preparedness of nursing staff prior to and during the implementation of mechani- It is paradoxical that although coercive measures are used to prevent harm to patients or ward staff, their use can inflict physical and mental harm to the patient. It is therefore important that nurses and ward staff exhibit broad competence in providing basic care to patients to prevent the risk of harm related to potentially deleterious physical and psychological effects and outcomes (Barnett et al., 2016;Berzlanovich et al., 2012;Cusack et al., 2018;Douglas et al., 2021;Jönsson et al., 2018;Kersting et al., 2019;Mohr et al., 2003;Nelstrop et al., 2006). The use of coercion in mental health involves ethical, legal and clinical issues. In our opinion, if nurses and nursing staff must implement coercive measures for safety purposes, the educational system and targeted mental health care must equip nursing staff with knowledge and skills through education, training observations and assessments of the implementation phase. If nurses and nursing staff gain increased competence, a focus on safe implementation may also contribute to reducing the use of coercion.

| Being unprepared led to scarce overall assessment
Our findings provide a basis for determining when mechanical coercive measures should be used. The participants revealed a focus on de-escalation and on gaining control of the situation. Many also emphasised the importance of information to 'secure' the patient to prevent further trauma. These findings align with the traditional focus on mental health nursing and the therapeutic use of the professional self (Crenshaw & Peplau, 1952), which highlights the im- should be provided to improve competence and skills throughout the entire assessment process, prior to and during the implementation of restraints.

| Limitations
This study has limitations. Considering the limited group of health professionals participating, their different professional roles and the fact that they were drawn from two limited geographical areas in Norway, the findings must be interpreted with caution. Despite these limitations, these nurses and ward staff contribute to the scarce empirical literature through insights into the practice of assessment and caring for patients under mechanical restraints-a subject area that is important for the development of health services in acute mental healthcare settings.

| CON CLUS IONS
This study aimed to explore nurses` and ward staff's perceptions and assessments of patient care when implementing mechanical restraints. The findings emphasise the need for management-in addition to nurses and ward staff-to focus on the assessment of patients, both prior to and during the implementation of restraints.
To ensure the quality of safe implementation in a potentially harmful situation, the focus should be placed on competence and professional responsibility. We suggest that nurses and ward staff should be well educated to contribute to the safer implementation of coercive measures, which also corresponds to an international call for integrated and coordinated holistic care, aiming to meet the patients` mental and physical healthcare needs. Nevertheless, this is related to a paradoxical situation that involves ethical, legal and clinical issues for patients and ward staff. Further research should explore how to develop nurses` and ward staff's assessments and follow-up of patients under coercive measures, which is important for patient safety and clinical practice.

| RELE VAN CE TO CLINI C AL PR AC TI CE
This study has several implications for clinical practice. First, the results highlight the paradox that although coercive measures are used to prevent harm to patients` and ward staff, its use may inflict physical and mental harm to patients`, but also lead to moral distress for healthcare professional in their everyday practice. Moreover, the results call attention to how the management qualifies professionals to handle situations prior to and during implementation of mechanical coercive measures. Finally, we stress the need for awareness of the patients` physical health in such situations.

ACK N OWLED G M ENTS
The authors acknowledge funding to undertake this project through Molde University College. The authors also wish to thank nurses and ward staff who participated in this study for their time and sharing their insights.

CO N FLI C T S O F I NTE R E S T
None.

AUTH O R CO NTR I B UTI O N S
LB, SV and IPM involved in study design. LB collected the data. LB, SV and IPM involved in analysis and manuscript preparation.

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data are not shared.