A Principlist Justification of Physical Restraint in the Emergency Department

The ethics of physical restraint in the Emergency Department (and elsewhere) has always been an emotive and controversial issue. Recently a vanguard of advocacy groups and regulatory agencies have been aiming to reduce and optimize its use, resulting in new guidance around physical restraint. This article considers prevailing opinions surrounding physical restraint in the Emergency Department using a Principlist model of medical ethics (specifically that of Beauchamp and Childress’ four pillars). It also examines the ethical underpinning of the new guidance on the usage of restraint. Ultimately, examination from a Principlist perspective suggests the use of physical restraint in the Emergency Department is justified, as long as it is used carefully. Despite this, physical restraint can have severe physical and psychological consequences for patients, and work needs to be continued into its reduction and optimization.


Introduction
The use of physical restraints within medical practice has long been considered an emotive and controversial issue (Cheung and Yam 2005). It is certainly not a new practice; as Masters (2017) notes, the restraint of patients for the safety of themselves or others dates back at least three centuries and probably before Roman times (Farina-Lopez and Estevez-Guerra 2011).
Physical Restraint's (PR) dark history informs the present where advocacy groups and regulatory agencies around the world are pushing to reduce physical restraint use (Knox and Holloman 2012) and to make sure its use is 'careful and appropriate' (American College of Emergency Physicians 2020). With change on the horizon comes a need for an in-depth examination of the ethical arguments surrounding PR and how we arrived at this new prevailing opinion about its use.
The Emergency Department (ED) is a unique environment with many challenges, including the management of complex and challenging patients. As such, it is one department that may particularly benefit from this examination of ethical calculi concerning PR. The acute nature of care provided in the ED frequently lends itself to ethical conflicts, both concerning PR and generally (House et al. 2015). Furthermore, the ED is a clinical environment where decisions may need to be taken quickly or risk not being taken at all. Therefore, to prevent harm to patient or practitioner, it is imperative to have restraint policies with a robust ethical underpinning, so as to be able to enact them at short notice if required.
This article takes a Principlist approach to examining the ethics of PR in the ED. The reasons for this choice are some of the same that have made Principlism the dominant framework for addressing many quandaries in medical ethics. Principlism's lack of consistent underpinning moral theory or perspective means that it comprises ideas from several different ethical schools, including deontological, consequentialist and virtue ethics (Hain and Saad 2016).
Principlism also has a correlation with basic human rights (Beauchamp and Rauprich 2016) which makes it ideal for this article. This is because we do not focus here on restraint usage in any particular country, but rather discuss the ethics of PR usage in developed countries generally. Restraint is something that is heavily influenced by the medical culture of the country it is being employed in, and so an ethical model which is 'universally applicable' (Beauchamp and Rauprich 2016) is required.
For the purposes of this article, the ethics of PR as it applies to critical care ED patients (those who may require future transfer to the Intensive Care Unit) will not be discussed. The ethical calculus surrounding restraint in this patient cohort is distinct as in many cases, critical care interventions (e.g. intubation) could not be performed without some level of restraint.
What has changed with PR? PR can be defined as any manual means by which a person is prevented from conducting an action (Masters 2017). Seclusion (which is a more passive form of PR), is the involuntary removal of a person from a situation to a room where they are physically prevented from leaving (Knox and Holloman 2012).
In the past PR referred to the use of leg-irons and manacles (Science Museum 2020). In many areas of the world (even prior the recent push for reform to PR), the medical community has progressed significantly from these draconian measures and modern PR usage in developed countries may comprise holding a patient, moving them to a place of safety or use of rails to prevent a patient leaving their bed.
There is evidence to suggest use of restraints is decreasing with an increased awareness of the importance of their appropriate use in preventing morbidity and mortality (Rakhmatullina et al. 2013). Even so, there is a vanguard pushing for PR to be reformed further. With respect to the ED, the aim is not to remove PR completely, but rather to optimize and reduce it as much as possible. Such optimizations may be codified into a set of rules, which summarizes prevailing opinion on ethical restraint usage in the ED.
• Rule 1: Only use restraint as a last resort (Chapman et al. 2016, Beysard et al. 2018) • Rule 2: All reasonable attempts must be made to maintain the restrained patient's privacy and dignity (American College of Emergency Physicians 2020) • Rule 3: The method of restraint should be the least restrictive necessary (American College of Emergency Physicians 2020) • Rule 4: Practitioners restraining patients must have adequate training in PR (Cheung and Yam 2005, American College of Emergency Physicians 2020) PR performed by untrained staff can have terrible consequences for the patient (Evans and Strumpf 1990). • Rule 5: After an incident of PR there must be adequate debrief Debriefs are a key part of modern restraint practice, allowing staff reflection (Cheung and Yam 2005) as well as discussion of how to prevent similar events in the future (Chapman et al. 2016). • Rule 6: There must be adequate documentation of PR usage (Chapman et al. 2016, American College of Emergency Physicians 2020) Proper and comprehensive documentation of PR allows for professional accountability and clinical audits to measure restraint usage.
There is evidence to show that if the above rules are applied to restraint policy, they can reduce physical restraint without a significant increase in negative outcomes (O'Keeffe 2017). For example, Enns et al. (2014) saw a 13-27 to 7-14% decrease in restraint usage as a result of intervention involving education and training. The focus of this article is to explore the ethical underpinnings of the above rules as well as of the use of PR in the ED generally.
Autonomy and non-maleficencethe importance of using of restraint as a last resort The best-known model for applying Principlism to medical ethics is that of Beauchamp and Childress (2012) and their four pillars of medical ethicsautonomy, beneficence, non-maleficence and justice.
Autonomy (and how best to respect it) is a good place to start our ethical examination of PR as it is perhaps the most frequently discussed of the four principles. Some believe that it is the predominant principle (Kokiko and Watts 1995), or that it is an integral part of the other three (Gillon 2003) whereas others hold the four principles as of equal importance (Hain and Saad 2016).
If autonomy may be loosely interpreted as a patient's freedom, then one may take the hypothetical standpoint that restraining a patient would by definition never be ethical -'restraint: something that limits the freedom of someone' (Cambridge English Dictionary 2021).
However, the regulatory implications of such a standpoint would be eliminating restraint in the ED completely, which is of course an impossibility (Fisher 1994). Furthermore, many would argue that there are situations where PR is ethically justified. However, PR's conflict with the autonomy of the patient does support rule 1 of PRthat it should only be used as a last resort. Prior to employing restraint, a patient may be calmed using methods such as pain relief or attending to sensory deficits e.g. by increasing light (Evans and Strumpf 1990).
The principle of non-maleficence also supports rule 1the use of PR only as a last resort. Non-maleficence can be thought of as not causing harm, and with regards to medical interventions specifically, not causing harm disproportionate to the benefits of any treatment (Beauchamp and Childress 2001). In its most severe forms, PR can have severe consequences for the patient's physical health, such as skin tears, pressure ulcers and exacerbation of pre-existing delirium (Cheung and Yam 2005, McBrien 2007, Chapman et al. 2015. This is especially true if it is incorrectly applied by someone who is not trained in PR (Rakhmatullina et al. 2013)thus supporting rule 4.
Aside from physical consequences of PR, reported patient perspectives document significant psychological consequences as well. Strumpf and Evans (1988) assessed the subjective impact of physical restraint on older patients and found that, 'patients vividly described anger, discomfort, resistance, and fear in response to the experience of restraint' (p.132). Direct quotes from patients in another study by the same authors show an emotional and overwhelmingly negative patient perspective on restraint: • 'I felt like a dog … that I was dirt' (Evans and Strumpf 1990, p. 126).
• 'I haven't forgotten the pain and the indignity of being tied' (Evans and Strumpf 1990, p. 126).
The physical and psychological consequences of the use of PR (even when correctly applied) not only support the use of restraint as a last resort but are an obvious argument for using the least restrictive necessary method of restraint (rule 3).
The possible adverse consequences also support the need for a team debrief after any PR incident (rule 5), where it can be discussed if it is possible to prevent such an event in the future. Some even argue for inclusion of the patient themselves in such debrief sessions, aiming to reduce concomitant psychological sequelae from the restraint experience (Cheung and Yam 2005).
Multidimensional autonomyan argument for making all attempts to respect the patient's privacy and dignity Valero (2019) argues it is possible to consider autonomy from a multidimensional perspective, in the context of PR he would see autonomy as more than just the physical freedom of a patient. This more nuanced understanding may allow practitioners to preserve a patient's dignity (rule 2). For example, a patient may require seclusion due to an acute exacerbation of the mental health condition (restriction of their physical autonomy), but their 'decisional' autonomy (Valero 2019) can still be respected e.g. asking them if they would like to change out of dirty clothes. Gillon (1994) makes a similar point, suggesting that respect for autonomy may not need an all or nothing approach. A patient may be unable to make decisions about their care (e.g. that they need to stay in the ED instead of leaving), but able to decide what they eat or wear. In these cases, ED staff should allow the patient to make the decisions they are able to in order to preserve their dignity (rule 2).
Beneficencean argument for PR in those lacking capacity However, the larger ethical quandary of PR results not from it conflicting with an individual principle (e.g. autonomy or non-maleficence), but from conflict between ethical principles (e.g. autonomy vs. beneficence).
A real-world example can readily show this. Few would argue that ED staff were acting unethically if they prevented a young child from moving whilst setting a cast for a broken bone. In this scenario, one ethical principle has taken precedence over the otherstaff are restraining the child (restricting his autonomy) in order to better fix his broken bone (beneficence). This conflict of principles is one of the things that makes the ethical calculus around use of PR particularly controversial and difficult.
Applying the ideas of WD Ross (2002) may allow us to navigate conflicts between Beauchamp and Childress' principles by allowing them to be considered as prima facie. Prima facie refers to self-evident primary principles of moral practice (Petrini 2013); they are binding until meeting conflict with another moral principle, at this point a decision must be made to prioritize one principle over another (Gillon 1994). In the above situation, beneficence has taken priority over autonomy.
Of course, it must be noted that due to their age, the child is not likely to have capacity and that influences our ethical calculus. The authors would posit, if the patient were capacitous (for example if they were an adult of sound mind) that ED staff would have no choice but to respect the patient's unusual request not to have their broken bone cast, thus giving autonomy precedence over beneficence. So, we must examine how does capacity (and more importantly those who lack it) factor into the argument on PR and autonomy in the ED? Beauchamp and Childress (2012) state that in order for a patient to be autonomous, they must have agency (the capacity for intentional action). This echoes the work of Matthews (2007) with his argument that autonomy can only be respected if it exists. Valero (2019) extends this argument to say that before respecting autonomy, healthcare staff must work to restore it.
We can take a second real world example to examine how this applies to PR. Delirium can lead to patients losing their agency (or capacity) but is often reversible. Imagine an example where staff employ bed rails to stop an older patient who is delirious from leaving the ED, their intention is to treat the patient's delirium with an antipsychotic. Once again, staff have prioritized beneficence (the return to sound mind of the patient) over the patient's autonomy (respecting her desire to leave).
Of course, there is another side to this argument, as some believe that restraint in those lacking capacity is not beneficent. Veatch (1984) argues that even a patient who lacks agency still has a right to respect of their autonomy. Meanwhile some argue for contextual decisions around PR usage in those lacking capacity. For example, an older person with dementia may have long periods of lucidity where they are capacitous; Cheung and Yam (2005) argue that the autonomy of the patient should be respected wherever possible and restraint not used in these periods of lucidity. This scenario is particularly relevant to the ED as 40-50% of all admissions via ED (McBrien 2007, Clevenger et al. 2012 are in older people with pre-existing dementia.
Justice as an argument for using PR to protect other patients and staff In modern health care, when PR is employed, it is often when the threat (or perceived threat) of harm to staff and other patients is high (Enns et al. 2014, Chapman et al. 2015, Tan et al. 2015, Masters 2017, Tropea et al. 2017; but how do Beauchamp and Childress' four principles apply in this scenario? This is an important question, as evidence suggests one third of nurses globally face assault at work (Spector et al. 2014). Aggressive behaviour is recognized to be a particularly frequent challenge for those working in the ED (Beysard et al. 2018), perhaps because ED is the usual place of presentation for those who are acutely ill.
Acutely ill patients (such as those suffering severe mental illness or intoxication/ withdrawal symptoms) are often the ones in danger of being restrained as they may be at risk of hurting themselves or others (Chapman et al. 2015). Data from around the world supports this association in the ED; for example, two recent studies indicated up to 90% of those restrained in the ED were experiencing mental illness (Gerace et al. 2014, Wong 2020, while other studies found alcohol or drugs were implicated in 33.7% of the analysed restraint cases (Beysard et al. 2018, Wong 2020. Beauchamp and Childress' (2012) principle of justice may be used to form the ethical underpinning of PR use in response to violence or aggression. Justice aims to promote fairness and equality among individuals in their receipt of healthcare, not just at the individual level but also at a systemic one. As such, one could argue that restraining violent or aggressive patients in the ED is ethical, as it may prevent an unwarranted assault on a fellow patient (or member of staff).
Furthermore, were a violent or aggressive patient to be severely ill (which as discussed above is quite likely), ED staff are in the difficult position of needing to treat them urgently despite their violent behaviour. In this scenario, one might suggest that it is more ethical to restrain them and treat them than to remove them from the department.
However, PR is a practice that can have severe deleterious outcomes, and so to make sure that practitioners are just in its application, PR use must be monitored (rule 6adequate documentation). If practitioners were found in any hospital to be biased towards selecting a particular group for restraint more often, it would contradict the principle of justice by not providing equitable treatment to all people.
Data shows there is a natural bias for PR to be used in two specific patient cohortsthose who are young and under the influence of psychoactive substances as well as older patients with cognitive impairment (Beysard et al. 2018). However, this is evidence of patients being treated according to their presentation (due to the increased risk of these patients to themselves and others) and not of any staff biases.

Looking to the future
Although there is a vanguard pushing for reduced and optimized PR in the ED (and elsewhere) there is still much work to be done to improve its universal application. The usage of PR in the ED (and the ethicality of PR usage) varies from country to country and hospital to hospital. Some countries and institutions have implemented guidelines to reduce PR whereas others are yet to do so.
Even if new policies are implemented, this is only half the battle. It is important to check whether these guidelines are being followed; Chapman et al. (2015) found that even when there was clear hospital policy on restraint, staff may be unaware of it. A lack of standardized terminology and definition of PR may also stand in the way of change by negatively impacting documentation (Chapman et al. 2015, Beysard et al. 2018. Obviously change will need to be planned on a per country (and per institution) basis, however qualitative research into staff opinions and motivations for applying physical restraint may provide a good first step into understanding and changing any negative institutional culture around physical restraint (Cheung and Yam 2005).

Conclusion
When considered from a Principlist perspective (specifically that of Beauchamp and Childress), there is significant ethical support for the prevailing opinions surrounding PR usage in the ED. Despite this, physical restraint is not a practice without adverse consequences for the patient, and work needs to be continued into its reduction and optimization.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Notes on contributors
Hugo Hall is in his penultimate year of medical school at King's College London. He hopes to do both clinical work and research when he graduates. His areas of interest are emergency and pre-hospital medicine as well as medical education.
David G. Smithard is a clinician researcher based in the South East of Greater London. He qualified form the London Hospital Medical College in 1986. He obtained his MD in Dysphagia following stroke from London University in 1997. He has published 70 papers in peer reviewed journals as well as giving numerous presentations to Learned Societies (British Geriatric Society, European Society of Swallowing Disorders, UK Swallowing Research Group, Dysphagia Research Society, Society of Acute Medicine). He continues to conduct research into Dysphagia Rehabilitation focussing on sarcopenia, frailty and dysphagia and dysphagia post stroke. David is Chair of the UK Swallowing Research Group and has been a board member of the European Society for Swallowing Disorders and Honorary Secretary of the British Association of Stroke Physicians. He has been awarded an honorary Fellowship by the Royal College of Speech and Language Therapists and is a Visiting Professor at Greenwich University. David has delivered invited lectures (on dysphagia in older people and stroke) in many European countries (including Kazakhstan, Denmark, Switzerland, Germany and Italy). He is a member of various working groups looking into the management of dysphagia in Care Homes and in Acute settings. David is the editor of Triple Helix, the official journal of the Christian Medical Fellowship.