Introduction

Coercion is practiced in psychiatry throughout the world and has been the subject of a long-standing ethical debate [16]—a debate that has recently intensified following a United Nations (UN) report [18] describing some forms of forced treatment in psychiatry as ‘tantamount to torture’ (p.1). When coercion is exercised within the regulations of the given mental health legislation, it is usually referred to as formal coercion [19, 20, 22]. However, various interventions that fall outside formal coercion might also infringe upon patients’ voluntary and autonomous decisions.

The terminology used to describe and define this type of intervention varies widely. It has been referred to variously as quasi-formal coercion [20], techniques to encourage adherence [1], and treatment pressures [32]. A commonly used general term is informal coercion [12, 19]. For simplification, this term will henceforth be consistently used in this paper. It is meant to be descriptive and not to determine a particular understanding of the nature of the interventions. A widely used description of informal coercion is Szmukler and Appelbaum’s [32] hierarchy of treatment pressures (Table 1), ranging from the least to the most coercive measures, encompassing persuasion, interpersonal leverage, inducement, and threat.

Table 1 Hierarchy of treatment pressures Szmukler and Appelbaum [32]
Full size table

While there has been considerable international research on formal coercion over the past 30 years [23, 27], research on informal coercion is mainly from the last 10 years. Much of this research has concentrated on delineating the topic and examining perceptions of patients who may have experienced informal coercion in psychiatric care [13]. Considerably fewer studies have examined the attitudes of mental health professionals to the use of informal coercion [1, 8, 28, 30], and virtually none has explored the varying impact of sociocultural contexts on attitudes towards the use of informal coercion.

Sociocultural contexts have been suggested to influence the remit of psychiatry in society in general and mental health legislation specifically [24, 25]. They may also be important for how informal coercion is used and experienced.

Purpose of this paper

Against this background, this study aims to explore attitudes towards and experiences of informal coercion of mental health professionals in countries with different sociocultural backgrounds. We aim to explore attitudes and experiences in a systematic manner using a qualitative methodology across countries with different psychiatric traditions and societal backgrounds. For the purposes of this study, we refer to ‘psychiatry’ as an umbrella term for mental healthcare services provided by a range of professions, including nursing, social work, clinical psychology and occupational therapy.

Methods

A focus group methodology was used to elicit participants’ attitudes towards and experiences of the use of informal coercion. Countries were selected for reflecting broad sociocultural divide. We aimed to include two countries of each of five distinct cultural regions: Anglo-America, Central/Eastern Europe, Latin America, Scandinavia, and Southern Europe.

Sample

Depending on the organisation of the national healthcare system in each country, potential participants were recruited in hospital departments and community services. Lead investigators used personal contacts and/or mailing lists of clinicians working in mental health services. Inclusion criteria for participants were: having worked in mental health for at least one year after qualification, currently working with patients with severe mental illness who might be subject to informal coercion, and aged between 18 and 65 years. Purposive sampling was used for gender and inclusion of one person from each major mental health profession per group. It was estimated that four focus groups per country would be sufficient to lead to within-group and between-group data saturation, based on previous literature [26].

Procedure and data collection

Focus groups were carried out between January 2013 and April 2014. Groups were led by experienced facilitators who had received training in focus group methodology and used a specifically developed topic guide (see Supplementary material). One facilitator per country carried out all four focus groups. Facilitators had varying backgrounds in psychology, psychiatry, and bioethics. A second member of the research team was present to observe and take notes.

After introductions, the facilitator opened the discussion with a general question about participants’ experiences of using coercive measures in clinical practice. The facilitator then introduced the topic of the focus group and explained the differences between formal and informal coercion using Szmukler and Appelbaum’s [32] hierarchy of treatment pressures (Table 1). Case vignettes (Tables 2, 3) adapted from examples of informal coercion as depicted by Molodynski, Rugkåsa and Burns [19] were then presented. The vignettes illustrated hypothetical cases with examples of how informal coercion might be used to encourage adherence to treatment; participants were asked to identify which level(s) of interventions appeared acceptable to them. The vignettes were deliberately vague to encourage debate and invite comparisons with the participants’ own experiences in practice. Facilitators used the topic guide to explore attitudes and experiences using standard probe questions, using Socratic enquiry to clarify opinions and invite further detail.

Table 2 Case vignette 1
Full size table
Table 3 Case vignette 2
Full size table

Facilitators were instructed to allow between 60 and 120 minutes for completion of each group. Groups were digitally recorded and transcribed verbatim. Groups in the United Kingdom, Canada, Chile, Germany, Italy, Mexico and Spain were carried out and transcribed in the original languages. Groups in Croatia and Sweden were conducted and transcribed in their official language and subsequently translated into English; the Norwegian groups were carried out in Norwegian and directly transcribed into English. All translations were done by experts with fluency in both languages. This approach ensured that all transcripts were in languages spoken by the members of the core team for the purposes of coding and further analysis (EV, CB, SP).

All participants provided written informed consent. The study originally received ethics approval from Comité Ético de Investigación Clinica del Hospital Universitario Fundación Alcorcón (Spain). Following this, ethics approval was granted in the United Kingdom (Queen Mary Research Ethics Committee, ID: QMREC2012/80), Canada (Ontario Shores Centre for Mental Health Sciences, Research Ethics Board, ID: REB # 13-011-B), Chile (Director General, Hospital Clínico de la Universidad de Chile; ID: No. 63; 14th November 2013) and Sweden (Regionala Etikprövningsnämnden Uppsala, ID: DNR 2013/011). National ethics approval was not required in the remaining five countries.

Analysis

Data collection and analysis were done concurrently. Transcripts were imported into the qualitative software package QSR NVivo 10. Two researchers (EV,CB) independently performed line-by-line coding on 16 transcripts to generate initial codes. Based upon this, a preliminary coding framework was developed. To assess its reliability, eight researchers external to the study used the preliminary framework to code one transcript each. The results were compared against the same eight transcripts, three coded by EV and five by CB; rates of agreement ranged from 80 to 98.5 %. Minor changes were made, and the final coding framework was used to code all transcripts.

The analysis was undertaken between October 2013 and July 2014 by the core research team, consisting of an academic bioethicist (EV), a research psychologist (CB), and a clinical academic psychiatrist (SP). Thematic analysis was carried out using an iterative process as described by Braun and Clarke [5]. Codes were compared and linked into themes, informed by regular discussion within the core research team. Following identification of candidate themes, the core research team re-read each transcript to ensure that themes were firmly grounded in the data, to further refine the candidate themes, and to locate illustrative quotes. Revision of themes and overall content were informed by repeated formal discussions within the Unit for Social and Community Psychiatry (Queen Mary University of London) which includes about 25 researchers and clinicians from different professional backgrounds.

Results

Two countries from each of the five target regions were included: Canada, United Kingdom (Anglo-American); Croatia, Germany (Central/Eastern Europe); Chile, Mexico (Latin America); Italy, Spain (Southern Europe); Norway and Sweden (Scandinavia). In each country, four focus groups were completed, forty in total. Exact response rates to recruitment in each country could not be determined. Some groups were pre-existing teams, some were composed of individuals from the same larger institution, and some were composed of individuals from different institutions. Therefore, the degree to which participants knew each other varied.

A total of 248 mental health professionals (163 women; 65.7 %) took part in the focus groups. Between 4 and 13 participants attended each group (M = 6.2). Group make-up by profession was largely heterogeneous (reported in Table 4), consisting predominantly of psychiatrists (n = 78), nurses (n = 73), clinical psychologists (n = 46) and social workers (n = 36).

Table 4 Counts of professions in sample
Full size table

Themes

The analysis identified five distinct but related themes that were commonly found across countries: (A) belief that informal coercion is effective, (B) widespread unease regarding informal coercion, (C) explicit dissonance between attitudes and practice, (D) implicit dissonance between attitudes and practice, and (E) paternalism and responsibility vs. autonomy as underlying values.

Informal coercion is effective

Agreement on the effectiveness of informal coercion was almost universal. Notwithstanding ethical or practical concerns, the majority of participants believed that techniques of informal coercion were effective for a range of outcomes; most commonly, treatment adherence. Specific measures in current use were also mentioned.

I think that we can assume that [informal coercion is] quite effective, considering that we, in fact, use it.

FGSW104, social worker, Sweden

I feel that all of them can be effective and my position is that the four [measures of informal coercion] are useful.

FGCH302, nurse, Chile

If a young man…doesn’t want to take medicine […], then the course of therapy would be to discuss it with him, and try to explain the reason for him taking this medicine […]. In most cases, one way or another at the end they accept.

FGCR104, psychiatrist, Croatia

Alongside adherence, participants viewed informal coercion as associated with improvements on a number of outcomes, including physical health, hospitalisation rates, and psychosocial circumstances.

FGCA304: Objectively, fewer admissions, fewer medical health problems, more:

FGCA302: More stable housing.

FGCA304: To be involved in the community, more stable housing, more likely to have re-engagement with family. The list of positive measures is very significant.

Psychiatrist (FGCA304) and social worker (FGCA302), Canada

Many groups described specific incentives that were used to promote adherence. These were often directly linked to receiving medication, and included financial incentives, offering food and drink, and making social outings with patients.

One of the main things we do in terms of incentivisation, give people money to take their medication or to take their depot injection really, and it has worked in the sense that it has kept some individuals very well for a very long period of time and their admission rate has fallen tremendously.

FGEN401, social worker, United Kingdom

Generally, participants stated that those who use coercion as a therapeutic strategy are more effective than those who do not.

Facilitator: Do you think that informal coercion is more effective…, the same clinical case and two therapeutic strategies: a clinician who uses coercion and another one who does not?

FGMX103: I believe the one who uses coercion is more effective.

Psychologist, Mexico

Widespread unease regarding informal coercion

While acknowledging that informal coercion can be effective, there was widespread unease about the term and its usage. Many participants felt that informal coercion was ‘ethically’ unpalatable. The language used to describe coercion was debated; understanding of what constitutes coercion varied. Participants stated that the sense of unease can be reduced by framing their actions in less severe terms.

FGNO302: By all means, [inducement] might be effective, but, I agree with you.

FGNO308: Oh yes, I am sure that it can, but it just tastes, ethically, a little bad.

Psychologists, Norway

From a moral point of view, I struggle to be aggressive towards another person. Because any form of coercion is a form of aggression towards someone else.

FGIT209, psychiatrist, Italy

Fundamentally I find coercion ethically not very justifiable.

FGDE404, psychologist, Germany

It’s like extorting something from the patient; if you give them money to take medication, you become completely omnipotent in making life decisions for that person, as if we were gods.

FGSP101, nurse, Spain

The unease was accompanied by a debate on the notion of informal coercion itself. At the outset, there was consensus in many groups that informal coercion is not regularly employed in clinical practice. As groups progressed, however, participants negotiated the language, used novel terms or eventually became accustomed to the term ‘coercion.’ There was gradual acknowledgement that acts of coercion seem more acceptable when portrayed in another light, and that informal coercion is regularly employed in clinical practice.

It depends on what we call it, you know. […] “We will see you outside the office, OK, we’ll go to your home instead of your needing to come here.” And so we call this offering something or a service and that we are flexible, but the flip side is that it is really a form of coercion, isn’t it? There are some nuances here.

FGNO308, psychologist, Norway

Unlike formal coercion, there is an absence of clear professional guidelines on the use of informal coercion. With this in mind, many participants acknowledged differences on how informal coercion is defined in practice.

There’s a huge difference in how patients and different personnel identify [informal] coercion. Is it coercion when you have to go to therapy X number of times? Is it [informal] coercion or is it an offer?

FGSW304, psychiatrist, Sweden

When an action does appear to be coercive, clinicians reduce their sense of unease by adopting a different definition for their actions—in this example, redefining a ‘threat’ to a ‘warning about the consequences.’

There is a thin line sometimes between what constitutes a threat and what constitutes a warning about the consequences. I sometimes tell them that, if their behaviour puts those around them in danger, the emergency service and police will intervene. That means that, if they threaten others and are aggressive, that’s not acceptable and their family has the right to call the police. So, that’s how I distance myself from it but, at the same time, warn the patient of the consequences.

FGCR403, psychiatrist, Croatia

I believe threat is very common, because in fact it does not sound so much like threat to me […] for example, one does not threaten, but rather points to the facts.

FGCH102, social worker, Chile

Explicit dissonance between attitudes and practice

Tensions between attitudes and practice were observed in almost every group. Many participants considered forms of informal coercion sometimes necessary, despite regarding them as unethical. They reported engaging in measures that they feel uncomfortable with, yet which they consider to be justified by therapeutic need. This inconsistency tends to arise in clinically complex situations. Participants mentioned a number of justifications for using informal coercion in spite of their concerns, including severe symptom levels and the potential for development of patient insight.

I feel like I have differing views even within myself. So, I don’t like the thought of using coercion, because I believe autonomy is very important where people have capacity, but at the same time I can’t imagine not using what are some of the most…the strongest clinical tools I have.

FGEN404, psychiatrist, United Kingdom

Attitudes toward the acceptability of coercion are sometimes overridden where therapeutic need is deemed more important. Sometimes decisions are made that were unforeseen or not ideal.

I usually think that if you have the patient’s best interests… At some point you can look at it, crazy as it may have been but if I’m doing [informal coercion] so that the patient doesn’t end up in even more trouble or embarrass themselves, as long as I can justify my decision, why I acted this way, I tend to be quite confident in my decision. But yes, it does happen that you make crazy decisions.

FGSW402, nurse, Sweden

Attitudes about the acceptability of coercion may be affected by emotional responses, such as annoyance and frustration. The emotional responses of the clinician to an escalating clinical situation can result in more extreme measures being used, where they previously would have been unacceptable.

You don’t want to use coercion but I’m wondering if the frustration in dealing with this over and over and over again leads you down the path to using more coercion. To say “You know what, we’ve seen this a thousand times, we need to push this guy into this.”

FGCA407, nurse, Canada

There are situations in which one is quite stressed in the moment, and tends to try and steer the patient.

FGDE204, psychiatrist, Germany

It is a sign of my defeat…which corresponds more to us being exasperated than to the severity of the patient him/herself. It is more an indicator of how much we struggle to manage the situation in some other way.

FGIT401, psychologist, Italy

Notwithstanding ethical considerations, participants reported that the use of informal coercion may be justified in certain situations, or with certain patients. These are linked to ‘extremes’ of behaviour, e.g. acute psychosis, mania, lack of insight. Coercive methods are implemented more often in particular situations.

Extremes of behaviour are responded to with extremes of action.

FGCA304, psychiatrist, Canada

The most common behavioural ‘extreme’ noted by participants relates to severity of psychiatric symptoms. If a patient is in acute psychosis or mania, for example, it was seen as more acceptable to use more coercive measures.

FGEN305: You know, the more unwell they are, the more likely you’re going to use the bigger, more questionable forms of persuasion. “Do this, or you know…” In the hope that they get some insight and their mental health stabilises. If someone is stable, and you are using threats or inducement and things, that’s really not on. But if they’re floridly psychotic, then:

FGEN306: Then threaten them.

Social worker (FGEN305), nurse (FGEN306), United Kingdom

When a person with mania spends all their money, it saddens me! That person there, when they get better and see what they have done—disrupted their relationships with their neighbours, the relatives don’t want to see them anymore, they’ve spent all their money—imagining this, I feel pressured to say: “Oh God! If I do this thing [informal coercion] these consequences won’t happen!”

FGIT301, psychiatrist, Italy

A second justification was the longer term goal of developing the patient’s insight. Participants stated that the pre-emptive use of informal coercion can result in later development of insight, and even gratitude.

Later, when they’re healthy, many are grateful that you got them to take their medicine. They maybe didn’t understand at the time because they were too sick to understand, but they still got their medicine.

FGSW102, nurse, Sweden

I would use both persuasion and inducement. If you believe it’s not right, especially inducement, you should always talk to the person, face-to-face, to see how they are really doing. Trying to see why they’re annoyed, what not taking medication means, if it’s a way to challenge the doctor. I would tell them that the medication is not for the doctor, but for themselves.

FGMX304, social worker, Mexico

Implicit dissonance between attitudes and practice

Many participants disavowed the coercive measures illustrated in the case vignettes, while later in the session stating that they may practice similar measures.

Participants told anecdotes of using measures that unwittingly fitted the descriptions provided in the case vignettes without acknowledging that they might be coercive. In these cases, clear contradictions between attitudes and clinical practice were evident.

If you have got a good alliance and it is based on a long-standing contact with the patient, then it’s there because the patient wants it to be so. This is nothing you can point to and use. No [interpersonal leverage] feels totally wrong. I’d never do this.

…

If a good alliance has been established, of course you can influence the patient much more. The patient is more inclined to do things. That’s how it is. There’s no getting away from that. But in the end, it will benefit the patient too. The treatment doesn’t usually go well if an alliance has not been formed. But sure, we can use [interpersonal leverage].

FGSW103, social worker, Sweden

Despite fitting the descriptions of informal coercion, participants were generally reluctant to label their practice as coercive. Participants shared their clinical experiences involving informal coercion, stating that in principle it was ‘unprofessional’ but in practice was acceptable.

[The clinician in the vignette is] saying “Well, I will help you if you’re well enough,” and then there’s this left up in the air, “Well the clinician will only help me, she’s trying to bribe me and using something I want to get for my children,” which seems very low, very unprofessional and unkind, actually.

…

Yeah, things like [going for meals with clients] can be very useful. I don’t think there’s any sort of a, abusively coercive element to it. They have a choice whether they turn up or not.

FGEN201, nurse, United Kingdom

I wouldn’t use threat, because it represents punishment.

…

Yes, you could end up in a situation in which threat is a possibility.

FGMX405, psychologist, Mexico

In some cases, there was disapproval of the methods in the vignette, followed later by acknowledgement that the methods are used in clinical practice.

If I am a subtle persuader, I can circumvent you in three seconds. Therefore, in this sense, persuasion becomes an act of psychological violence on the patient.

…

We work a lot on that level, I believe—I mean, a kind of persuasion that does not appear like violence.

FGIT103, psychiatrist, Italy

Many participants became aware of the implicit dissonance as the group progressed. Attitudes towards informal coercion appeared to evolve with the duration of the focus groups. From an initial standpoint of disapproval, participants defined and re-defined informal coercion until eventually accepting and justifying its use in clinical practice.

By saying in the beginning of the conversation that I use no form of coercion (apart from involuntary hospitalisation) and then, at its end, saying that I use all of them depends on the key factor of good intentions and focusing on the patient, whether they’ve experienced something as well-meaning, although they’re not always in a state which allows them to see good intentions as such.

FGCR403, psychiatrist, Croatia

Paternalism and responsibility vs. autonomy as underlying values

In every group, the underlying values of psychiatry were debated. Opinions fell under two broad viewpoints. One view was that professional opinion regarding ‘best interests’ has more weight than the patient’s right to autonomy. Others argued for a less paternalistic approach, questioning whether psychiatry ever has the right to impose its views on patients. The majority of participants, however, acknowledged that paternalism and autonomy are interwoven in practice and that paternalistic approaches are inescapable. This was suggested as being driven by a sense of accountability and a desire to protect oneself and one’s community. This theme relates to themes C and D; paternalism is linked to the ‘best interests’ justification.

Some teams (varying within and between countries) described a softer approach which values autonomy.

If someone’s deemed capable, you’re not doing anything anyway, right? […]We’re letting them going about their business. We’re free to live our folly, right?

FGCA304, psychiatrist, Canada

A key reason for supporting autonomy is the perception that clinicians have no right to force patients.

Just like I go to my GP asking for tablets for a headache, the patient comes here asking for treatment for psychosis. Similarly, if they’re not actually under the Mental Health Act, what right do I have to be persuading them, coercing them, inducing them, whatever you want to call it? What right do I have to be imposing my views on what they need, on that autonomous person?

FGEN404, psychiatrist, United Kingdom

Where does the free will of a psychiatric patient begin and where does it end? Can we really say “You should do this”?

FGDE303, social worker, Germany

Other teams described a more paternalistic approach. Where risk of harm is apparent, value judgements are made regarding the prioritisation of safeguarding over autonomy.

If someone announces they will kill themselves you have to hospitalise them. Why? If they came to you, that means they are seeking help. […] I think a right to life is more important than the right to autonomy. I know I have taken that right away.

FGCR104, psychiatrist, Croatia

A key reason for taking a paternalistic approach is a perceived lack of patient capacity.

Mental disorders in some ways affect capacity and judgement of the person him/herself and, therefore it is almost necessary that there is another person that in some way makes a more or less coercive decision towards a goal that is really, anyway, in the best interests of the patient.

FGIT302, psychiatrist, Italy

They lack capacity and I have responsibility for them because they are at risk, their family is at risk and then there is no other option for me than to go against their wishes.

FGSP105, psychiatrist, Spain

The two approaches were not necessarily seen as mutually exclusive. A gentler approach can mask the underlying assumption that clinicians are expert providers and that their clinical judgement reflects authority based on good intentions.

Our intentions are good and want the best for our clients and want them to recover as much as possible […] But I guess subconsciously we feel like we know what will help them to recover more than they do, so we try to coerce them into […] ‘Take your medication, see your doctor, live in this place’ all these kinds of things. […] I think it is subconscious and motivated by good intentions.

FGCA102, social worker, Canada

Participants reported being often confronted with dilemmas in which value-based paternalistic judgements are made, sometimes because they are the easiest option.

We can very easily fall into being parental, believing it’s our children that we’re working with. We attach our own values on how life should look like.

FGSW304, psychiatrist, Sweden

Safeguarding others against harm is often seen as a major role of psychiatry. This was suggested by participants as being a driving force for the use of coercive measures such as interpersonal leverage, which is justified below.

It is so extremely important that he has these injections otherwise it becomes so hopeless for the 20 other people around […]. So here, we are equally acting out of consideration for the people around them, actually, as for the patient himself.

FGNO402, psychiatrist, Norway

You don’t just treat the patient, there’s also a factor of interacting with society and one has social responsibilities. If a patient with bipolar disorder is relapsing, they could do anything to their child and we have responsibility for the patient.

FGCH102, social worker, Chile

Participants described invisible pressures, based on role expectation, as leading them to using coercive measures.

The unconscious sort of nudge is that […] the client needs to be on medication, and to some extent the nudge is, you’re failing in your duty as a nurse […] if you do not get the client on the medication. That is not said, as such, but it’s kind of a nudge here and a nudge there.

FG307, nurse, United Kingdom

Handing over the power to psychiatrists or psychologists does not imply that they know how to use it in the best way.

FGMX202, psychologist, Mexico

Alongside the perceived responsibility to safeguard others, and the pressure to achieve a favourable outcome, participants reported an overarching concern for accountability. It was viewed that paternalistic judgements are sometimes made because of the clinician’s worry that they might be held responsible for negative consequences of their patients’ non-adherence to treatment.

We kind of have a duty to the public and uphold safety, as much as we want to help them with their recovery, include families, all of that, we still have that hanging over our heads as well. Because if anything were to go wrong, we’d be the first people they’d come to, to find out what went wrong with that care.

FGCA105, social worker, Canada

If we fear for the patient’s life or the life of another person at risk from the patient, we can’t leave it. We have to communicate with someone the patient knows, which sometimes is someone from a community service.

FGMX201, psychologist, Mexico

Contradictory evidence

Our analysis identified mainly commonalities across countries. Yet, there was also evidence in contradiction to the conclusion that the findings were rather consistent across all countries. Family involvement featured more frequently as a theme in Latin countries (Chile, Italy, Mexico, Spain). Participants described experiences of applying pressure to family members, as well as family members bringing pressure on clinicians to make treatment decisions. Participants also described the preservation of family relationships as justification for the use of coercion on the patient. Finally, instances of explicit and implicit dissonance were slightly less prevalent in the Spanish speaking countries (Chile, Mexico, Spain).

Conclusions

Main findings

The study used a consistent methodology across ten countries from five different sociocultural regions. It found commonalities of attitudes towards and experiences of informal coercion. Participants reported that (a) interventions referred to as informal coercion are effective tools to promote adherence. In spite of this, there was (b) widespread unease regarding their use, with ethical issues frequently cited. This unease was related to a general sense of dissonance between attitudes and practice. (c) Explicit dissonance was evident in participants who acknowledged that their unease co-exists with their convictions of the need for informal coercion. (d) Implicit dissonance was also evident, in those participants who were disapproving of the concept and term of coercion, yet described acting in ways that matched the vignette descriptions of coercion. This dissonance reflected an underlying tension of (e) paternalism and responsibility vs. autonomy as core values in psychiatry. The dissonance observed was unidirectional. In theory, participants generally disapprove of the use of coercion. In practice, however, their disapproval is frequently overridden.

Strengths and limitations

To our knowledge, this is the first empirical study to examine attitudes and experiences of mental healthcare professionals regarding the use of informal coercion across countries in different sociocultural regions. The study design was implemented broadly as envisaged, and this was achieved across ten countries.

The research team was multidisciplinary. Facilitators were qualified, and the analysis was iterative involving different disciplines so that the interpretation was informed by different perspectives.

Furthermore, participant make-up of the focus groups was largely reflective of service organisation in each country, e.g. multidisciplinary teams in the United Kingdom and Canada, specialist psychology and psychiatry services in Italy and Mexico.

There are also limitations. While the selection of countries was purposive, the selection of sites within countries was opportunistic. Despite this, the results across most groups were consistent with a high level of saturation.

The definition of the five sociocultural regions may have been arbitrary. All but one (Mexico) of the included countries are currently classified as ‘high-income’ countries, according to the World Bank, and we do not know whether the findings also apply to lower income countries. Much previous research on this topic, however, has only concerned Western, high-income countries, and it is of note that the findings from Mexico were similar to the findings elsewhere.

The terminology used was a point of contention in many groups, prompting discussion on the actual definition of informal coercion and eliciting some strong negative reactions. This negative attitude towards the connotation of the term coercion may have introduced a linguistic variation, and influenced the later discussion of the actual behaviours that were labelled as acts of ‘coercion’ in this category.

Comparisons with the literature

The findings reflect that mental health professionals tend to use informal coercion in practice more than they feel is right in principle, for three possible reasons: that they believe it is effective; that they feel under particular pressure to use it; and that their understanding of paternalism and responsibility, as afforded to their role, justifies or even demands it.

International similarities in attitudes to and experiences of informal coercion have also been observed in smaller scale studies across the United Kingdom and the United States [7, 17, 21]. Perceived effectiveness as a reason for the use of informal coercion has also been noted previously in qualitative studies [11, 30]. The current literature provides little evidence that informal coercion [15] or formal coercion [6, 14] is effective, although there are recent exceptions [29].

The use of informal coercion, in principle, led to participants reporting widespread unease. Theoretical disapproval of coercive techniques has been found among clinicians in similar studies, particularly with regard to inducement [8, 28]. However, other studies have reported clinician endorsement of such measures—though, crucially, only when the measures are not labelled as coercive [1, 30]. As noted by Hoge et al. [10] coercion is a ‘moralized construct’ (p. 170) and therefore subject to disagreement among clinicians about which behaviours it applies to.

Whether labelled as coercive or not, it is evident from the findings that mental health professionals sometimes behave in ways that they would theoretically disapprove of, whether explicitly or implicitly. Attempts to empirically examine the prevalence of informal coercion in psychiatry have been hampered by debate on the semantics of coercion and clinicians’ perceptions of what behaviours are coercive [30]. Interviews with patients have indicated that perceptions of informal coercion are common, and that rates of perceived coercion tend to rise along with severity of illness [2, 7].

The term coercion—and all the translations used in the study—for the actions described in the case vignettes, made many participants feel uncomfortable and were frequently regarded as provocative and inappropriate. The findings of this study suggest that many mental health professionals feel affronted when the term “coercion” is used for their practice and is ambivalent towards using the interventions that have been captured as informal coercion. For most professionals, however, these interventions are part of what they do. This is reflected in the identified dissonance, which may lead to tensions for individuals and within teams, particularly when it is not made explicit.

The dissonance clinicians report feeling in the course of their practice is reflective of an enduring debate about the underlying values in psychiatry. The idea that psychiatry is inherently coercive is not novel [33] although debate has intensified following remarks by the UN Special Rapporteur, who stated that ‘severe abuses continue to be committed in healthcare settings where choices by people with disabilities are often overridden based on their supposed “best interests,” and where serious violations and discrimination against persons with disabilities may be masked as “good intentions” of healthcare professionals.’ (p.14) [italics added]. Such high-profile criticism of psychiatric practice may fuel the unattractiveness of the field as it suggests that professionals may have to enter ethically uncomfortable territory in the course of their work [9]. For clinicians in this study, and perhaps across psychiatry in general, Beauchamp and Childress’s [3] widely used principles of biomedical ethics—beneficence, non-maleficence, justice and autonomy—come into direct opposition [4]. A beneficence or ‘best interests’ approach was related to paternalism in clinical practice. In adopting an approach of beneficence, clinicians believe that autonomy must be compromised.

Implications

The findings of this study indicate that mental healthcare professionals across a variety of countries work with ambivalence and contradictory expectations. This raises a broader question about the underlying values that drive psychiatry and that are likely to have drawn many professionals into working in the field. The widely acknowledged sense of unease could contribute to the portrayal of working in mental healthcare as potentially challenging and/or unattractive. Professions in which people cannot behave in ways they feel is ‘right’ or ‘ethical’ become unappealing [31]. In such professions, a position towards external accusations of coercion as well as internal unease about one’s own practice should be established.

It may be unrealistic to formulate precise guidance that pre-empts all uncertainty about the use of informal coercion in practice. However, there is an obvious need for more explicit and targeted debate in psychiatry around informal coercion [34]. This debate could lead to a code of practice that provides a coherent approach for how to align clinical practice with the principles that professionals believe in. Such guidance may also allow addressing the issue of informal coercion more explicitly in training and ongoing professional supervision.

Future research may explore whether the belief in the effectiveness of different forms of informal coercion can be underpinned by evidence and how positive therapeutic relationships and good communication can help to reduce informal coercion. The findings of this study suggest that such research can be done on an international scale, despite national and sociocultural differences.