International variations in mental-health law regulating involuntary commitment of psychiatric patients as measured by the Mental Health Legislation Attitudes Scale

Introduction

Involuntary detention and treatment for mental-health problems is a potentially controversial procedure, which is often justified for both therapeutic reasons and public protection. Although coercion can be beneficial when risk to self or others is a serious issue, it can also adversely impact upon a person’s state of mind and severely impair their psychological well-being. The legislation that governs the admission and management of mentally ill people in each country should comply with the standards set by the United Nations Convention on the Rights of Persons with Disabilities ¹ in order to ensure balance between patients’ human rights and their need and right for treatment, and public safety. However, despite the efforts of the World Health Organization to standardise strategies for the delivery of mental health-care internationally, previous research has shown that the legal frameworks in this area are diverse across countries, ² even when the countries are culturally and geographically similar. For example, the procedures for involuntary commitment and involvement of stakeholders in the initiation and decision-making process vary across countries. Also, while most legal frameworks include dangerousness criteria in various forms, the need for treatment in the best interests of the patient, regardless of dangerousness, is sufficient to detain individuals irrespective of risk in some countries^3–5 such as Sweden. An overview of the varying legal frameworks in the 11 countries included in this study, based on consultation with national experts, is provided in Table 1.

OPEN IN VIEWER

Table 1.

International experts’ views on differences in mental-health legislation.

Country	Q1. What are the criteria for involuntary placement?	Q2. Diagnoses are legally defined, excluding conditions not sufficient for involuntary placement	Q3. Who is responsible for the initial assessment before transfer to psychiatric facility?	Q4. Who has the authority to decide on involuntary placement and how many experts are involved in the assessment?	Q5. Involuntary placement and treatment are legally defined as different modalities¹ (yes or no)	Q6. Detailed regulation of coercive measures (yes or no)	Q7. Compulsory outpatient treatment possible (i.e. community treatment orders; yes or no)	Q8. Priority of less restrictive alternatives² is explicitly mentioned in the legislation (yes or no)	Q9. Patients are transferred to psychiatric hospital by…	Q10. Independent review of patients’ detention is legally required (yes or no)
Ireland	T or D	Wide; PD, A	GP	Two psychiatrists; 2	No	No	No	Yes	P, F, OPS	Yes
Sweden	T	n.d.	GP, P	GP and psychiatrist; 2	Yes	Yes	Yes	Yes	F, P, E, OPS, IPS	Yes
Germany	D	Wide	Any doctor, police	Judge and psychiatrist; 2	Yes	Yes	No	Yes	No clear regulation	Yes (judge)
Denmark	T or D	Not defined	Doctor	Psychiatrist and GP; 2	Yes	Yes	Yes, if certain criteria are fulfilled	Yes	P	Yes
England and Wales	T or D	Wide, PD included	Doctor, AMHP, GP	Doctor & AMHP, 2 or 3 depending on type of detention	Yes	No	Yes	Yes	P, AMHP	Yes
Slovenia	D	n.d.	Doctor, P	Judge and psychiatrist, 2	No	Yes	No	Yes	P, E, IPS	No, but patient can appeal against detention
Romania	D	n.d.	GP, P	Two psychiatrists and one doctor from other specialty, 3	Yes	Yes	No	Yes	P, F, E, OPS, IPS	Yes
Iceland	T or D	Wide	Any doctor, GP, police	Psychiatrist, 1	Yes	No	No	Yes	P, F, OPS	No, but patient can appeal against detention
Norway	T or D	Wide	GP, P	GP and psychiatrist; 2	Yes	Yes	Yes	Yes	F, P, E, OPS, IPS, any	Yes
India	T or D	n.d.	GP	Psychiatrist and medical officer,³ 2	Yes	No	No	No	F, P, E, OPS, IPS	No, but a board of visitors inspects IPS every month

¹Involuntary placement or treatment legally defined as different modalities: indicates only the legal separation of the modalities, regardless of whether in routine care, persons placed involuntarily must accept treatment.

²Priority of less restrictive alternatives: underlines that coercive measures is an ‘ultima ratio’, prerequisite hereto is the availability of facilities offering less restrictive.

³Medical officers only have a medical degree, and they replace psychiatrists sometimes due to the severe shortage of psychiatrists in India.

Q1. Criteria. D: dangerousness to other or self; T: need for treatment; D or T: dangerousness or need for treatment. Q2. Diagnosis. n.d.: not defined; wide: diagnostic categories mentioned with no restriction to specific diagnoses; psychosis: restriction to psychosis or conditions similar to psychoses; PD: special regulations excluding personality disorder; A: addiction is excluded. Q3. Initial assessment. GP: general psychiatrist; P: psychiatrist; AMHP: approved mental-health practitioner; CN: community nurse. Q9. Transfer. F: family; P: police; E: emergency department general hospital; OPS: outpatient psychiatric services; IPS: inpatient psychiatric services.

Compulsory admission rates per 100,000 population vary remarkably across Europe, ranging from six per year in Portugal to 218 per year in Finland. ⁶ This strongly suggests that differences in definitions, legal backgrounds or procedures contribute to variations in detention rates and mental-health service delivery. ⁷

In order to improve national legislation globally and to prevent the excessive application of compulsory procedures in some countries, an evidence-based international debate is needed to facilitate shared learning and opportunities for service improvement. This could lead to the development of a consensus across countries on the best legal practices and then to harmonisation of legislation and practice across the European Union and worldwide to reflect these benchmarks. A recent report issued by the European Union Agency for Fundamental Rights confirmed this need for a renewed discussion about compulsory placement and treatment in the region. ⁸ In line with this, Mulder ⁹ has also suggested that international collaborative working groups should be established urgently to investigate and compare legal practices across countries to support these initiatives.

Despite these appeals, international comparative studies in this field are rare. A previous study comparing attitudes of mental-health professionals and lay people towards involuntary admission and treatment from England and Germany by using scenarios of potentially detainable patients found that the different legal frameworks did not influence attitudes much. ¹⁰ Another comparative review questioned whether various European laws on compulsory commitment to care in relation to substance use disorder or misuse problems comply with international ratified conventions concerning human and civil rights. ¹¹ However, international comparative studies among stakeholders with direct experience of the process, such as professionals and family members, across several countries, covering different legal aspects and procedures are still lacking.

Therefore, this study was designed to compare the opinions of professionals and family members about the operation of the national mental-health law regulating forcibly admission and treatment of psychiatric patients in 11 countries: Ireland, Iceland, England and Wales, Romania, Slovenia, Denmark, Sweden, Germany, Norway and India.

Methods

Results

Level of satisfaction

In order to estimate the level of satisfaction with the mental-health legislation per country, we recoded the responses as either approving the mental-health law (i.e. ‘agree’ or ‘strongly agree’ with the statement) or not approving (i.e. ‘neither agree nor disagree’, ‘disagree’ or ‘strongly disagree’). We found that the question which received highest approval rate across all countries was about the care benefits of the law (Q3), and the lowest approval rate concerned the implementation of the legislation (Q6; see Table 2). Responses clearly varied per question and country. The average overall approval rate for the current mental-health law across all 11 countries was about 68%. The countries with the highest overall satisfaction were England and Wales (76%) followed closely by Denmark (74%), and the most dissatisfied country was Ireland (64%). The actual average scores for each item varied from 2.9 for Q6 (‘implementation of the legislation’) to 3.8 for Q3 (‘care benefits’), with an overall average score of 3.4 out of 5.

OPEN IN VIEWER

Table 2.

Level of satisfaction per question.

Country	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Overall satisfaction
1. Denmark	53	68	74	65	78	52	80	61	51	74
2. Iceland	67	63	60	45	53	41	53	32	51	68
3. India	60	60	81	53	53	27	32	53	49	65
4. Germany	44	56	66	72	54	34	52	53	41	66
5. Norway	56	49	69	65	55	36	44	37	39	67
6. Romania	62	55	87	69	53	24	31	43	54	66
7. Slovenia	27	42	79	73	65	22	69	78	76	69
8. Sweden	61	64	58	59	60	48	70	43	47	70
9. England and Wales	73	75	85	69	85	64	72	50	61	76
10. Ireland	53	38	77	56	64	36	51	23	44	64
Total	54	54	70	62	59	38	54	44	47	68

Table shows percent of respondents approving (i.e. ‘agree’ or ‘strongly agree’) each aspect of the current mental-health law by country.

Multivariate analysis for factors affecting participants’ overall satisfaction

Scale reliability

We then treated the nine questions as a scale (MHLAS). Its internal reliability was very high (Cronbach’s α = 0.808). Based on this good scale reliability, we constructed an overall total score as the sum of all nine questions and transformed it into a 100-point scale by dividing the total score for each respondent by the maximum possible score. An overall score of 100 here would mean that all nine questions were answered with the maximum individual score of 5 (i.e. ‘strongly agree’). A high score indicates strong satisfaction, and a low score indicates poor satisfaction. In order to preserve the integrity and meaningfulness of this measure for all further analyses, we included only respondents with valid answers to all nine questions, which resulted in 1444 respondents for the multivariate analysis.

Country comparison by overall levels of satisfaction

The mean MHLAS total score by country is presented in Figure 1. Respondents in England and Wales and Denmark expressed the highest approval for their national legislation, and those in India and Ireland expressed the lowest approval. We compared these mean satisfaction levels for all 10 countries with the analysis of variance (ANOVA) test for significant differences and Tukey’s post hoc adjustment for multiple comparisons. This analysis confirmed that there are statistically significant differences in the satisfaction with the current mental-health law between the 11 countries (ANOVA F-test = 7.3, df = 9,14, p<0.001).

OPEN IN VIEWER

Figure 1.

Overall satisfaction with the current mental-health law (percent). 100%=fully satisfied with current mental-health law.

Country classification

We built a classification and regression tree (CART) model to determine the most significant breaks in the trend and to create meaningfully distinct groups.^13,14 The CART model discovered four different groups of satisfaction with the current mental-health law: group 1 (with the lowest satisfaction): Ireland and India; group 2: Germany, Norway and Romania; group 3: Iceland, Slovenia and Sweden; and group 4 (with the highest satisfaction): England and Wales and Denmark.

Factors for overall satisfaction with the current mental-health law

We then performed initial statistical testing on factors potentially affecting the degree of overall satisfaction (MHLAS total score). Demographic characteristics (sex and age) and overall work experience had no association with overall satisfaction. However, the respondents’ experience with involuntary admissions did have a significant effect (ANOVA F-test = 26.5, df = 2,14, p<0.001) in that the more experience the person had with admissions, the more satisfied they were overall with the mental-health law in their country.

Respondents’ professions also had a significant effect on satisfaction (ANOVA F-test = 37.2, df = 3,14, p < .001). Doctors and nurses were relatively more satisfied with the legislation (68% and 71%, respectively) compared to the police (63%) and family members (63%). The difference between doctors and nurses on the one hand and police and family on the other was statistically significant.

We then adopted the overall satisfaction (MHLAS total score) as the outcome variable and used the available independent variables to build a multivariate model (see Table 3). Overall work experience was excluded from the model a priori because family members were not asked this question.

OPEN IN VIEWER

Table 3.

Multivariate model of overall satisfaction (0–100%).

Factors	Coefficient	p-Value
Constant	56.0	<0.001
1. Denmark	7.7	<0.001
2. Iceland	6.5	<0.001
3. India	–5.2	0.003
4. Germany	–2.0	0.111
5. Norway	3.6	0.012
6. Romania	–1.0	0.58
7. Slovenia	1.4	0.435
8. Sweden	2.7	0.036
9. England and Wales	8.2	<0.001
10. Ireland	Reference
Age (ordinal 1–5)	0.1	0.752
Sex (female = 1)	–1.0	0.172
Experience with admissions (ordinal 1–3)	–1.9	<0.001
Profession
Doctors	8.4	<0.001
Police/tribunal members	–1.7	0.189
Nurses	4.5	<0.001
Family	Reference

When controlling for demographics, the number of involuntary admissions experienced and profession, some differences between the countries remained intact, while others became or remained not significant.

Sex and age had no effect on overall satisfaction with the current mental-health law in this model. Experience with admission did have an effect though in that as in the univariate analysis, greater experience was associated with higher satisfaction when the other variables in the model were held constant. Doctors and nurses tended to have significantly greater satisfaction than family members, while the satisfaction level of criminal justice professionals was not distinguishable from that of family members. The key finding from the multivariate models was that most inter-country differences remain statistically significant, even after controlling for demographics and other important factors (see Table 4).

OPEN IN VIEWER

Table 4.

Significant differences between countries based on multivariate models for total MHLAS score (overall approval).

Factor	Reference country
	Denmark	Iceland	India	Germany	Norway	Romania	Slovenia	Sweden	England and Wales	Ireland
1. Denmark	Ref.
2. Iceland		Ref.
3. India	Less	Less	Ref.
4. Germany	Less	Less	More	Ref.
5. Norway	Less		More	More	Ref.
6. Romania	Less	Less	More		Less	Ref.
7. Slovenia	Less	Less	More	More			Ref.
8. Sweden	Less	Less	More	More		More		Ref.
9. England and Wales			More	More	More	More	More	More	Ref.
10. Ireland	Less	Less	More		Less			Less	Less	Ref.

‘Less’ means less satisfied than the reference country. ‘More’ means more satisfied than the reference country.

Evaluating satisfaction based on the individual questions

Another way to investigate differences in satisfaction with the mental-health law between countries is to compare responses to the individual MHLAS items using ordinal regression. ¹⁵ We built nine separate ordinal regression models (one for each item), using Ireland as a reference group in all analyses. This approach was used due to the prior reporting of Irish data as noted above. ¹⁶ In each model, we included a control block of sex, age, profession and experience with admissions. The main results with reference to the country differences compared to Ireland are summarised in Table 5.

OPEN IN VIEWER

Table 5.

Multivariate ordinal regression models (MHLAS individual items).

Outcome	Q1. Treatment efficacy	Q2. Admission criteria	Q3. Care benefits	Q4. Consent to treatment	Q5. Detention review	Q6. Implementation of the law	Q7. Information about the law	Q8. Transfer to hospital	Q9. Reciprocity principle
Factor	Odds ratios (significant only, p<0.05)
1. Denmark		3.1			1.9	2.0	4.9	5.0
2. Iceland	2.6	4.9				2.3	2.1	3.1	1.7
3. India		1.7	0.6	0.6	0.3	0.5	0.3	3.0	0.4
4. Germany	0.4	1.5	0.4	1.9	0.4	0.7		3.5	0.4
5. Norway	1.4	2.1		1.5	0.7	1.6		2.7
6. Romania		1.7			0.5		0.4	2.7
7. Slovenia	0.3			2.4	0.5	0.5	2.0	9.8	2.9
8. Sweden		2.8	0.3		0.7	1.6	2.0	2.5	0.7
9. England and Wales	2.5	4.5			2.0	3.3	2.3	3.5
10. Ireland	Reference
Age (ordinal 1–5)	0.9
Sex (female = 1)								0.8
Experience with admissions (ordinal 1–3)	1.2	1.2	1.4	1.1	1.2	1.4	1.3		1.2
Profession
Doctors	1.8	2.1	2.3	1.6	2.4	2.2	2.5	1.9	3.1
Police/tribunal members			0.6				1.7
Nurses	1.8	1.6	1.4		1.5	1.7	1.9	1.8	1.9
Family	Reference

For example, looking at the multivariate ordinal regression model for treatment efficacy of mental-health law (Q1) in Table 5, we can say that Iceland, Norway and England and Wales tended to approve of this aspect significantly more strongly than Ireland. On the other hand, Germany and Slovenia had significantly lower satisfaction compared to Ireland. India, Romania and Sweden were not statistically different from Ireland on this question.

It is striking that on Q2 (admission criteria) and Q8 (transfer to hospital), almost all countries had a more positive attitude in comparison with Ireland, whilst on the other items the direction of relative satisfaction varied.

Age and sex had an effect on only two items: older people tended to be less satisfied than younger people on Q1 (treatment efficacy), while women tended to be less satisfied than men on Q8 (transfer to hospital).

Experience with admissions had a strong effect on most item responses: the more experience the respondent had with involuntary admissions, the more satisfied they were on all items apart for Q8 (transfer to hospital) which was not significant.

Professional role had an impact on the level of satisfaction on each item, with doctors and nurses being more satisfied in most of the cases than criminal justice professionals and family members.

Discussion

This study is the first to examine attitudes towards key aspects of mental-health legislation in a large international sample with substantial groups of professionals and family members in many countries. Whilst there are clear limitations in terms of self-selecting participation and variations in the proportion of stakeholders in each group across countries, the findings highlight areas of potential good practice. As such, they could be used to inform the improvement of legal practices across countries to protect the fundamental rights better of people with mental-health problems.

With regard to international variations, respondents in three countries (England and Wales, Denmark and Iceland) were relatively satisfied with the operation of their legislation, whilst those in Ireland and India were relatively dissatisfied. These national differences cannot be reduced to regional variations across Europe, however. Iceland and Denmark are both ‘Nordic’ countries which might explain their similarities, but Norway is also in that region of Europe. It shares many legal and cultural institutions with the other two countries but consistently rated lower satisfaction. This dissatisfaction could be partly explained by the relatively high rates of involuntary admission in Norway. ¹⁷ Even more starkly, Ireland and England and Wales have similar ties but are at the opposite end of the scale in terms of satisfaction. India is a geographical and cultural outlier in the set of countries studied here, and the relative lack of resources for mental-health care in many parts of that country ¹⁸ may go some way to explain the consistently low rate of satisfaction there in comparison with all of the European countries. It should also be noted that during the study period, the Mental Health Act (MHL) 1987 was in force in India, while a new MHL that takes into account the UN Convention on the Rights of Persons with Disabilities (UN-CRPD) has come into effect since 29 May 2018. Further, our findings of German stakeholders being significantly less satisfied with their legislation in comparison with other countries (i.e. Iceland, Slovenia, Sweden, Denmark and England and Wales) are in line with recent conclusion that the German federal state laws are still extremely heterogeneous and do not fully comply with the requirements of the UN-CRPD. ¹⁹ It is hoped that analysis of the qualitative responses from each country included in the survey will enable these patterns to be more clearly understood.

The total score of the MHLAS provides a composite measure of attitudes in this area, and the analysis here indicates relative dissatisfaction amongst respondents in some roles and some countries. Criminal justice system (CJS) professionals (i.e. tribunal members and police) and family members were relatively dissatisfied compared to health-care professional (nurses and doctors) on the overall use of the relevant mental-health law in their country and on all but one of the individual items. This presumably reflects the various relationships each group has with the unwell person who is being detained and with the systems available to provide the person with care and treatment. The wishes of family members in the crisis situation are likely to vary between those who want more or less robust intervention than they actually receive. They may feel the intervention was excessive, so their dissatisfaction is on behalf of their relative’s human rights or they may conversely feel the intervention was ‘too little too late’. Police officers are likely to aim for the most efficient transfer of the person into the hospital and again will probably vary according to the degree of sympathy they have towards people with mental-health problems. Those who are unsympathetic may regard any involvement as an inappropriate use of their time which could be devoted to other activities seen as more relevant to policing.

Individual item responses enabled us to identify some of the specific areas of relative dissatisfaction. The lowest average score of 2.9 on the way MHLs are implemented into practice shows that most of the countries are struggling with how mental-health services are organised in order to follow legal requirements. All countries except for Slovenia had a more positive attitude in comparison with Ireland on the admission criteria for involuntary placement. In comparison with other countries, Irish legislation explicitly excludes people suffering from personality disorder or those who are addicted to drugs or intoxicants from involuntary treatment, even if these people pose a danger to themselves or others (see Table 1). These unmet treatment needs may explain why dissatisfaction levels are highest among Irish professionals and family members regarding this aspect of the law.

Furthermore, all countries had a more positive attitude in comparison with Ireland on the way people are transferred to psychiatric hospitals. If we take Norway as an example, Norwegian stakeholders’ more positive attitude may reflect recent service innovations there. ²⁰ Although still in its experimental stage, the University Hospital of Bergen implemented the first mental-health service ambulance in 2005 aiming to replace local police as the means for transporting patients to inpatient services. Initial evaluations have showed positive results with this service such as less stigmatisation and better access to relevant information.

In addition, we found that access to information about the procedure for involuntary commitment and treatment differs between countries, with stakeholders from Romania, India and Ireland being the least satisfied with this aspect. We share the opinion of Wyder et al. ²¹ that providing patients and their relatives with free and full access to information about the involuntary admission and treatment process, their rights to appeal and rights of access to their own clinical records may help them to regain personal control over their lives after the disempowering and intrusive experience of involuntary admission and treatment.

Furthermore, it is worth noting that health-care professionals here (nurses and doctors) had significantly more experience with involuntary placements than CJS professionals and family members. We found that the more experience the person had with admissions, the more satisfied they were overall with the mental-health law in their country, which could be explained through a process of a cognitive dissonance. ²² Health-care professionals who have been repeatedly exposed to the mental discomfort of acting against a patient’s expressed wishes may have transformed this distress into a more positive view in order to reduce internal tension. This will allow them to execute their professional duties in circumstances they are not able to change by experiencing less psychological stress and internal conflict. From a clinician’s point of view, it is also natural to look at the process from one’s own perspective. Previous research has shown that attitudes of psychiatrists and nurses regarding compulsory admission are in keeping with those of the general population in four surveyed European countries, but they are not in keeping with clinicians in those countries who are not involved in the compulsory admission process. ²³ This indicates that clinicians who participate in the compulsory admission process seem to be in line with the public when it comes to attitudes, but interestingly not necessarily with the views of family members. Alternatively, mental-health professionals may have more positive attitudes towards the use of legislation because they have experience of its benefits, especially in comparison with the suffering, lack of recovery and life disruption caused by not admitting and treating people in a timely fashion.

We propose that the MHLAS described here can be used as a brief, practical tool for surveying the attitudes of busy people involved in the formal and informal care of people with mental-health problems. It can be completed quickly, and there is evidence of good internal consistency. We recommend that it is used in a wider range of countries not included in this sample to enable comparisons to be made and a global picture of this important topic to be constructed.

A number of study limitations must be acknowledged here in order to judge the meaningfulness of the findings. In particular, the participants were a self-selected group who may have particularly strong views (positive or negative) on the issue and do not necessarily represent the wider population of equivalent professionals and family members in each country. We therefore fully acknowledge that our samples may not be representative of the entirety of the professional groups surveyed in each country. In addition, we realise that the law does not stand apart from the range and quality of services provided in the various countries surveyed, and neither do sociocultural differences in attitudes towards people with a mental disorder and the involuntary admission process, with previous studies showing differences across different EU countries and respective professionals. ²³ The sample in each country reflected the networks of the lead researcher there, leading to highly skewed representation of professional groups across countries. So, there may be a country/profession interaction effect in the sample which must be considered when interpreting the findings on either of these variables. This issue was addressed in the regression analyses by including a variable indicating the professional group for each participant. We were unable to survey patients because of the way the study was designed, but future research should specifically take into account patients’ views.

Finally, it should be noted that high satisfaction does not necessarily equate to best practice. Satisfaction with the process may be based on priorities far removed from the needs of the patient such as bureaucratic simplicity or personal interests. The degree to which it can be inferred that the relative approval for mental health law in England and Wales, Iceland and Denmark indicates these countries have the ‘best practice’ is questionable and should not be interpreted simplistically.

Conclusion

There are significant variations across Europe and beyond in terms of approval for how the national mental-health law framework operates in each country. The MHLAS can be used to study these variations and to contribute to improved practice in this challenging area.

Supplemental Material