Compulsion and “coercion” in mental health care

Abstract

"Compulsion" in mental health care is a reasonably straightforward notion: the use of force, one hopes always governed by law, to make a person accept treatment that has been refused. The term "coercion" is usually taken to include "compulsion", but encompasses a broader range of practices. Sometimes it is used almost synonymously with treatment pressures, including "interpersonal leverage", and even "persuasion". I believe that, if we are to take our thinking — including research ideas — forward, we need a more precise understanding of "coercion" 1. Most accepted is the definition proposed by Wertheimer 2, who includes "threats" as coercive. A "threat" is a conditional proposal ("if …, then …") that, if rejected by the person, leaves him/her worse off according to a "moral baseline" ("if you refuse the medication, you will be detained in hospital"). The moral baseline is that one is normally entitled not to be deprived of one's liberty. That is not to say that it can never be justified, but a special case needs to be made. Wertheimer contrasts a "threat" with an "offer" (or inducement). An example: "if you take the prescribed medication, you will receive a payment" 3. Here a rejection of the proposal does not leave the person worse off, as he/she is not entitled to a payment. Nevertheless, such an inducement can be problematic, for example, by undermining the patient's sense of agency or through corrupting the value of the treatment 4. I take coercion to cover both compulsion and threats. A further consideration is the difference between "objective" coercion and "perceived" coercion. The former follows the definitions given above. The latter is a person's perception of threat, even where no threat may be intended. A disquieting problem facing psychiatry is its "coercive shadow", the fear many patients have that non-compliance may lead to the use of compulsion. Patients may agree to treatment, including admission to hospital, "voluntarily" to avoid the humiliation and stigma of a compulsory order. Research shows this is very common, even though in most places threats are regarded as ethically unacceptable. There has been little discussion of this topic, but I suggest that, given the uncertainty of whether a proposition is a threat or not, we might look at ways of "regulating" threats: for example, making them transparent – their only being made in "good faith" (that is, the threatener really means it)— and clarifying practice in codes of practice or professional ethics. Apart from the problem of definition, research on coercion is dogged by the problem of context. There is a large variation in the rates of compulsory admission to hospital, both between countries (even without outliers, 3- to 4-fold) 5 and within countries 6. The use of seclusion, restraint and forced medication may vary hugely 7, even 10-fold from hospital to hospital in the same country 8. The sources of variation can be attributed to different service configurations, different mental health laws, different social policies (for example, the rate and extent of bed reductions; the degree of emphasis on risk and public protection), and, crucially, culture. In some countries physical restraints are regarded as unacceptable and are rarely or not at all used; in others chemical restraints are thus regarded. Furthermore, the use of compulsion may change significantly over time according to changes in policy and practice. In England, there has been a doubling over the past 20 years 9. Even if well-designed self-report or interview measures are used, ones that are interpreted similarly from place to place, the results of any one study on coercion will likely have limited generalizability. Thus, perhaps more than in any other field of health services research, international collaborative studies are needed. The EUNOMIA programme is a good example 10. Further points to be considered are where (in the community or in the hospital) and when (before discharge or after) the assessment is made, and by whom. Service user researchers may get different responses from conventional researchers. Variations here may lead to different results. Research ethics committees often struggle with research in this area. It is sensitive, and there may be concerns about consent and the "voluntariness" of participation, which may lead to bias due to the exclusion of important subgroups of patients. With careful thought these problems can be overcome. A huge challenge to involuntary treatment comes from the United Nations (UN) Convention on the Rights of Persons with Disabilities 11. By April 2015, 159 states were signatories. The elimination of discrimination by ensuring that rights may be enjoyed "on an equal basis with others" is a fundamental aim. Persons with serious mental illness are considered by the UN Committee for the Rights of Persons with Disabilities, the authoritative body set up by the UN to interpret and monitor compliance with the Convention, to fall under the characterization of "disability" (sometimes referred to as "psychosocial" disabilities). Article 14 states that "the existence of a disability shall in no case justify a deprivation of liberty", meaning that "mental disorder" or "mental illness", even if it represents only one of a number of criteria for involuntary detention in a mental health law, renders such a law non-compliant with the Convention. Article 12 recognizes that all persons enjoy "legal capacity" in all aspects of life on an "equal basis with others". The Committee, in a recent "General Comment" on this article, states that "substitute decision-making", where someone decides for the person with a disability (as opposed to "supported decision-making"), is non-compliant 12. Over twenty "concluding observations" made thus far by the Committee, following its monitoring of reports on progress from States in implementing the Convention, conclude that they must "take action to develop laws and policies to replace regimes of substitute decision-making by supported decision-making, which respects the person's autonomy, will and preferences" 13. It is hard to imagine a society in which it would be seen as right that persons who are seriously incapable of exercising autonomy or expressing their will and preferences would be allowed to act so as to incur grave harms, including death. Where the UN Convention is valuable, apart from its clear articulation of a host of other rights for people with disabilities, is in making us scrutinize in depth our justifications for coercive interventions. Together with colleagues, we 14, 15 have argued that conventional mental health law discriminates against persons with a mental disorder since it does not respect such persons' autonomy (or rights to self-determination or self-governance) in the same way as in the rest of medicine. In the latter, considerations such as impaired "decision-making capacity" and treatment needing to be in the person's "best interests" justify the over-riding of a treatment refusal. In the mental health field, a diagnosis of a "mental disorder" – usually vaguely defined – and the presence of some kind of risk to self or others comprise the criteria. The rules are entirely different. Furthermore, the "protection of others" permits the preventive detention of persons with mental disorder on the basis of the risk they are deemed to pose before they have actually committed an offence. This group is unique in this regard. The many more persons without a mental disorder who are equally or more risky are not liable to such detention. In this regard, non-discrimination means either having generic "dangerousness" legislation equally applicable to all who present an unacceptable level of risk, or no preventive detention for anyone. Thus we 14 have argued for a non-discriminatory, generic, "fusion law" that would apply to all persons, whatever their diagnosis — medical, surgical or psychiatric — and whatever the setting. Involuntary interventions would only be justified for those who lack decision-making capability (unable to understand and retain the relevant information, to appreciate its pertinence to their situation, to reason with it in the light of what is important to themselves, and to evidence a choice) and only where it would be in their "best interests" (essentially what that person would have chosen if he/she had retained capacity in the current circumstances). Advance statements or directives (see 16 in this issue of the journal) could play an important role here. Northern Ireland is currently proposing to legislate along these principles. Bach and Kerzner 17, attentive to the "legal capacity" standard of the UN Convention, have proposed three levels of "decision-making capability". The first is "legally independent", having full decision-making ability as outlined above. The next level is where varying degrees of support — informal or formal — would be required to assist the person to arrive at a decision based on the person's will and preferences. The third level, "facilitated" decision-making, would represent a last resort and would be restricted to instances where it is impossible to arrive at a settled understanding or interpretation of the person's will and preferences and where decisions are made by another person. However, as part of this action, the facilitator would continue to work with the person to establish with time what are the person's will and preferences. An approach that combines both of the above could be developed. "Decision-making capacity" and "best interests", both terms criticized by the UN Convention Committee, can be helpfully reconceptualized in terms of the person's "real" or "authentic" will and preferences 15. The huge variation in rates of involuntary treatment suggests that in many countries there is considerable scope for a reduction. From an ethical point of view, a randomized controlled trial (RCT) of involuntary inpatient treatment is hardly possible. We accept that it can be morally justified, indeed obligatory, to treat people involuntarily under certain circumstances. However, there have been three RCTs of involuntary outpatient treatment (or community treatment orders). While each has its flaws, none has shown a clearly significant improvement in any of a range of outcomes 18. I have argued that an alternative approach, consistent with the "fusion" proposal, would conceive of community treatment orders in a different way and would look for different, individual, patient-preferred, outcomes 19. There is reasonably consistent evidence, even when involuntary treatment has been authorized, that "perceived coercion" is less when the relationship between patient and clinicians is good, and when patients believe their "voice" has been heard 20. A promising means of reducing the need for coercion at times of crisis, especially a relapse of illness, might be an advance directive, or the less legally formal "joint crisis plan" (see 16 and 21 in this issue of the journal). There is evidence in the case of the former that, when helped by a facilitator in drawing up the directive, in the short term at least, patients may experience their care as better 22. Joint crisis plans have been more extensively studied. An earlier, sizeable, RCT found a significant reduction in involuntary admissions when a joint crisis plan had been agreed between patient and clinical team. However, a much larger RCT involving 569 patients found no difference in involuntary admissions or any other outcome 23. A lack of treatment fidelity or clinician "buy-in", a problem for any multicentre complex intervention, may have been responsible. A joint crisis plan pilot study for patients who self-harm also found no benefit 24. However, 85% of patients who had a joint crisis plan said they would recommend it to others. Perhaps this reflects the respect accorded to the patient's "voice" in the joint crisis plan negotiation. In conclusion, there are considerable conceptual and practical difficulties in understanding and researching compulsion and coercion. Nevertheless, it is hugely important to our patients and, indeed, for the status of psychiatry that we do all that is possible to reduce recourse to these measures to a minimum.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 147 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.