Long-Term Forensic Mental Health Services: An Exploratory Comparison of 18 European Countries

ABSTRACT The objective of this study was to explore current provisions within forensic mental health inpatient services for people who require longer-term care within Europe. We used a structured questionnaire and follow-up semi-structured interviews with experts in forensic psychiatry in 18 European countries. All experts interviewed acknowledged the issue of ‘long-stay’ in forensic psychiatry with patient characteristics including chronic mental disorder, treatment-resistance and violent behavior. Formal and informal definitions of ‘long-stay’ varied widely between countries. Eight experts stated that long-stay services are currently available in their country. Of the countries without long-stay services, five experts expressed a need develop them. Improved quality of life and promotion of wellbeing were emphasized as the fundamental treatment philosophy. Even without an agreed definition of ‘long-stay’, it is clear that a proportion of mentally disordered offenders (MDOs) are ‘stuck’ in ‘the system’. Experts shared common concerns in terms of political pressures to contain dangerous MDOs for ensuring public safety as well as ethical debates regarding long-term forensic mental health care. Further research is required to promote dialogue between and within countries to address the balance of patient' rights and public safety, and to produce longitudinal and economic analyses of existing long-stay forensic service provisions.


Scope of the problem
The appropriate care and risk management of mentally disordered offenders (MDOs) poses a number of complex treatment, moral and ethical challenges (Boyd-Caine, 2012;Buchanan & Grounds, 2011;Konrad & V€ ollm, 2010;Mullen, 2000;V€ ollm, Bartlett, & McDonald, 2016). Though a rapid reduction of beds in psychiatric hospitals generally has been witnessed internationally, there has been a significant increase in demand for forensic services (Hodgins, M€ uller-Isberner, & Allaire, 2006;Jansman-Hart, Seto, Crocker, Nicholls, & Cote, 2011;Priebe et al., 2005Priebe et al., , 2008. While the length of stay (LoS) in forensic inpatient mental health services has fallen, at least in some countries, and recovery principles have been applied to MDOs (Sugarman & Oakley, 2012), a number of patients still experience lengthy stays in forensic services, potentially at inappropriately high levels of security (Shah, Waldron, Boast, Coid, & Ullrich, 2011;Sharma, Dunn, O'Toole, & Kennedy, 2015;Shaw, Davies, & Morey, 2001). This is of concern for two reasons; firstly, low-volume inpatient forensic services are cost and resource intensive, and secondly, the quality of life in these restrictive environments may be poor (Joint Commissioning Panel for Mental Health, 2013;Vorstenbosch, Bouman, Braun, & Bulten, 2014). Data from a previous comprehensive European comparison highlights the wide variation across Europe regarding total numbers of forensic cases (ranging from 100 in Ireland to 5,400 in Germany in 2002) and prevalence rates per 100,000 population (ranging from two in Greece to 21.7 in Denmark (Salize & Dressing, 2005)). Costs are also high, with a general increase in LoS in medium to high secure hospitals in England and Wales (Rutherford & Duggan, 2007) and an average per person cost of £200,000 in medium secure settings per year (Walker, Craissati, Batson, Amos, & Knowles, 2012).

Factors associated with length of stay
There is currently no generally accepted definition of 'long-stay' in forensic settings, and little is known about the LoS of these patients in different countries. In England, research has found an average LoS in high secure care of eight years (Dell, Robertson, & Parker, 1987), and for medium secure care it is a little over two years (Edwards, Steed, & Murray, 2002). However, some authors have described a trend for patients to stay for five years or more (Rutherford & Duggan, 2007;Shah et al., 2011;Jacques, Spencer, & Gilluley, 2010) with around a third of medium secure patients deemed to need long-term care (Melzer et al., 2004). A more recent cross-sectional study identified that around 16% of patients resident in high secure settings in England had been resident for more than 10 years, and around 3% for more than 20 years (V€ ollm, 2015). The Netherlands and Germany have also experienced increasing lengths of stay in forensic inpatient mental health services and increasing numbers of patients in need of longer-term care (Giesler, 2012;Nagtegaal, van der Host, & Schonberger, 2011).
Some previous research has been directed towards identifying characteristics of patients who stay in forensic inpatient mental health services for excessive time periods (Alderman, 2001;Long et al., 2010;Wheatley, Waine, Spence, & Hollin, 2004;Yorston, 1999). Based on discharge samples, comparing those with longer versus shorter lengths of stay, severity of index offence was identified as most important in personality disordered, and psychopathology in patients with mental illness in UK high secure settings (Dell et al., 1987). In UK medium secure settings, research has identified severity of psychopathology, psychiatric history, seriousness of offending, being on a restriction order (requiring Ministry of Justice permission for transfer and discharge), non-engagement in interventions, dependency needs and lack of step-down facilities associated with longer stay populations (Brown & Fahy, 2009;Jacques et al., 2010;Kennedy, Wilson, & Cope, 1995;Long & Dolley, 2012;McKenna, 1996;Shah et al., 2011). Recent research in Sweden of a high-risk forensic cohort has highlighted that violent index criminality, among other factors, is an important factor associated with longer stays in forensic psychiatric treatment (Andreasson et al., 2014). Research has also highlighted severity of the offences committed (Baldwin, Menditto, Beck, & Smith, 1992;Green & Baglioni, 1998), neuropsychological impairment and low IQ as factors associated with LoS (Colwell & Colwell, 2011). Research from the USA has highlighted problematic behavior and increasing physical health problems associated with long-stay (Fisher et al., 2001), and research from Ireland associated severe mental illness and violent offending with increased LoS (O'Neill et al., 2003). However, little is known about the patient characteristics of those who remain in secure care, how to formally identify them, or how to best meet their needs in existing services when they may require longer-term care than other patients.

International perspective
At the international level, complex differences in patient populations, diagnoses, legal frameworks, differing concepts of criminal responsibility, service provision as well as cultural, political and public expectations lead to heterogeneity in MDOs admitted to forensic care and mental health services providing this care (Edworthy, Sampson, V€ ollm, 2016;Salize & Dressing, 2005). Such differences impact on service provision and treatment outcomes for individuals; for example, certain countries provide various forensic inpatient and outpatient mental health services, while others either do not recognize forensic psychiatry as a separate specialty, or do not possess the sufficient resources or training in order to ensure satisfactory service provision for MDOs (Salize and Dressing, 2005). Some countries have developed policies and services specifically designed for long-stay patients and it is this service provision that is the focus of exploration in our study. There exists a patient population who, due to a perceived long-term risk, spend their entire lives in secure forensic settings. As such the question needs to be asked whether this population's needs are currently being catered for in mixed populations (that is, with 'shorter-term' patients leaving the system quicker, but who may be more acutely unwell than the long-term patient population). With some individuals spending their entire lives in secure settings, restrictions on personal freedoms become more apparent, including restrictions on patient' rights to family life and sexual expression. For example, Tiwana and colleagues (2016) found that many countries lack national policies on sexual expression for patients in forensic mental health services, with the UK in particular being most prohibiting. For people subject to such restrictive settings, it is relevant to explore whether designated long-stay services are able to address fundamental rights and needs of this patient population.

Objective of current study
This study sought to explore current service provisions within forensic mental health inpatient services for those who require longer-term care within Europe. In order to put provision for this patient group in context, we also describe briefly the legal framework governing forensic mental health services in each country, as well as availability and access to services for MDOs. We then investigate the availability of long-stay services in 18 countries within Europe, with a focus on definitions of long-stay, legal frameworks, service configuration, patient populations, quality of life and ethical issues.

Context
All but three included countries are members of the European Union (the exceptions being Switzerland, FYR Macedonia and Serbia), and all countries in this study are state parties to the European Convention on Human Rights and Fundamental Freedoms (ECHR, Council of Europe, 1950), which provides a common-ground that 'legitimizes international scrutiny of mental health policies and practices within a sovereign country' (Salize & Dressing, 2005). Furthermore, all included countries, as of 2015, have signed and ratified the European Convention for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (CPT, Council of Europe, 1987) with most also ratifying the Convention on the Rights of the Persons with Disabilities (CRPD, United Nations, 2006; Finland, Ireland and The Netherlands are all signatories but have yet to ratify the CRPD). Each of these conventions place responsibilities and obligations on state parties to ensure and promote certain human rights and fundamental freedoms for all people without discrimination, particularly when deprivation of liberty has been ordered by a public authority after the commission of a crime or in the context of a person's mental disorder or other disability.

Design
We approached experts associated with the EU-funded COST action IS1302 (Cooperation in Science and Technology) 'Towards an EU research framework on Forensic psychiatric care'. 1 A national selection process is required to join this network, which ensures all COST country representatives are leading clinicians and researchers with expertise in forensic psychiatry and a particular interest in long-term forensic psychiatry. A written questionnaire was developed by the core group of the COST action (grant holder, chair, co-chair, scientific advisor; see Supplementary Materials). Topics addressed included system and definition elements, admission and discharge, patient characteristics and service provisions. This questionnaire 2 was designed to provide context, prepare for and direct the content and structure of the subsequent semi-structured interviews 3 . Semi-structured interviews were conducted by telephone with the experts that had provided the questionnaire for their country (see Supplementary Materials). The interviews focused on service provision for long-stay populations in each country, characteristics and practice of service provision, key challenges and hindrances in their implementation and outcomes. Information about practical aspects of such services (e.g. size of wards, level of security) as well as clinical and risk factors, treatment pathways, legal frameworks, perspectives on quality of life and ethical considerations of long-stay facilities were also discussed. Interviews were recorded with consent of the participant and were subsequently transcribed and analyzed using thematic qualitative analysis. Participant data were anonymized, with all names removed from within transcripts, and stored securely as per the Data Protection Act 1998. The study was part of a larger, national, multi-center project, sponsored by Nottinghamshire Healthcare NHS Trust, which provided Research and Development (R&D) approval. Due to the nature of the study (expert interviews) separate NHS research ethics approval was not required.

Procedure
We gathered data from 18 separate countries in total. Out of the 19 countries participating in the ISCH COST action IS1302, we were able to interview 17 representatives from 16 countries (one per country, with two from The Netherlands); 4 all interviews took place via teleconferencing (in the English language) between June 2013 and November 2014. For the remaining two countries (Croatia and Macedonia), we were only able to use data from the initial structured questionnaire.

Analysis
Data analysis was conducted using thematic analysis (Braun & Clarke 2006) to identify common themes, and was coded using NVivo qualitative data analysis software by one researcher, with 20% (i.e. four interviews) double-coded by the senior author (NVivo, 2014). Data were analyzed both inductively (with themes that emerged from interview content) as well as deductively via the use of coding determined by the themes explored in both the initial questionnaires and subsequent semistructured interviews (Fereday, 2006). All participants checked the full interview transcripts for accuracy of their statements before analysis.

Results
Legal frameworks and service provision for mentally disordered offenders

Legal frameworks
The majority of countries included in this study operate under civil law jurisdiction, with only England, Wales 5 and Ireland operating under common law jurisdiction. Procedural differences between admission and diversion provisions can be partly attributed to legislative differences. In some countries, e.g. Germany and Switzerland, local variations exist in the various 'states' ('L€ ander' in Germany, 'Kantone' in Switzerland). In Germany, however, state laws define patient rights and staff duties, which vary considerably between states with no standardized procedures beyond admission and discharge of forensic patients.
The majority of countries' legal frameworks relevant to the management of MDOs are found under criminal laws and penal codes as well as dedicated mental health legislation. The seven countries in which MDOs are managed mainly via mental health legislation (i.e. separate to penal codes or criminal codes) include England and Wales ( Croatia (the EU's most recent member state to join in 2013) introduced the country's law on the 'Protection of Persons with Mental Disorder', which came into effect January 2015. This new legislation intends to replace the traditional medical model approach to treatment with a human rights based approach to reflect the principles of the CRPD (Bagari c, Zivkovi c, Curkovi c, Radi c, Bre ci c, 2014). This is an important development in mental health law in Europe, with implications for the way in which the CRPD is acknowledged and implemented by countries that are signatories to the convention. For individuals with disabilities deprived of their liberty, provision of 'reasonable accommodation' (Article 14) is of particular concern to long-stay forensic populations.
Access to forensic mental health services Forensic mental health services and general mental health service provision for prison populations vary widely between countries (see Table 1 for an overview of services, and Table 2 for patient population characteristics, admission and discharge procedures). Most countries included (with the exception of England and Wales, Finland and Latvia) stipulate that a treatment order/ forensic placement and a prison sentence can be imposed at the same time, with the order by which they are served differing between countries (as also detailed by Salize & Dressing, 2005). Most countries' admission procedures allow for MDOs to be 'diverted' from the criminal justice system to forensic mental health services either before or after sentencing (Table 1).
Diversion from prison to a designated forensic or psychiatric hospital is not possible in Croatia, Germany, Italy, Lithuania, The Netherlands or Serbia; if a prison sentence is imposed and subsequently a mental disorder emerges, then treatment is provided for via prison general psychiatric services or a designated prison hospital. After diversion from prison to forensic mental health services for treatment, should treatment be considered complete, most countries require the patient to be sent back to prison to complete their sentence. Finland, Italy, Latvia, Lithuania, Poland and Portugal are the exceptions, where patients are either discharged back to the community or sent to facilities of lower security. Each country expert stated that there was no fixed release date for people admitted to forensic mental health services, with the key criterion of dangerousness as justification for lengthier stays. It is worth noting, however, that since the time of the present study, new legislation has been introduced in Croatia and Italy specifying that patients cannot remain in forensic settings for longer than what would have been their sentence for the same offence had they been healthy and sent to prison. Furthermore, in the case of Portugal, for crimes punishable by imprisonment for less than eight years, length of stay in a forensic mental health service cannot exceed this time. However, However, see Barbui & Saraceno (2015), detailing new legislation that calls for downsizing and closure of these forensic hospitals. 5 Forensic and penitentiary services and admission criteria differ between Catalonia and the rest of Spain, with services more heavily privatised in Catalonia.
should dangerousness not be found to justify continued stay in a forensic mental health facility (as reviewed every two years), then the patient must be discharged. In England and Wales, patients may either be moved back to prison (while their sentence is still active) or remain in the hospital system, depending on the needs of the individual. Each country provides some inpatient prison psychiatric services, with either designated 'prison hospitals', psychiatric wards within prison, or via visiting mental health professionals (psychiatrists or psychologists); however, these services are not standardized throughout prisons in the included countries (Table 1). Admission criteria for forensic and/or general psychiatric inpatient services for MDOs share some similarities across countries, including that a mental disorder needs to be present in order to be admitted, e.g. psychotic disorders, personality disorders (typically associated with another mental disorder), cognitive disorders, learning disabilities and substance misuse (typically associated with another mental disorder). Substance use related disorders and personality disorder (as the sole disorder) might constitute an exclusion criterion for forensic services in some countries (Table 2).
In order to be admitted as a forensic psychiatric inpatient, the majority of countries' laws and regulations stipulate that a person needs to have committed a crime and that there was a relationship between the mental disorder and the criminal behavior. This is not the case in England, Wales or Ireland in which patients can be admitted to forensic mental health services under civil legislation if they are in need of treatment but have not committed an offence.
Although MDOs in Belgium and Slovenia must have committed a crime in order to be admitted as a forensic patient, they are the only other European countries that do not require a relationship between the mental disorder and the criminal behavior. In Germany, the seriousness of a crime (usually a violent crime) and a high risk of reoffending are requirements for admission to forensic services. In most countries, it is necessary for the offender to have diminished or absent criminal responsibility in order to be admitted to forensic inpatient services. England, Finland, Ireland, Portugal, Serbia, Slovenia and Switzerland are the exceptions to this rule. In these countries, admission is typically on the basis of the need for treatment and 'therapeutic security' ( Table 2).

Definitions
Seven of the 18 countries were able to offer a formal definition of 'long-stay', either under legislation, regulations or based on national health research, and 13 countries' experts provided an informal observation of length of forensic inpatient stay in secure settings (Figure 1).
The formal definitions do not necessarily reflect what would constitute a 'long-stay' in forensic mental health services in practice in Finland, Ireland, Portugal, Spain and The Netherlands. For example, Finland's law on Social and Healthcare Service Fees (1992) defines 'longstay' as three months of continuous institutional treatment regardless of the reason for treatment (i.e. applicable to forensic psychiatry, general psychiatry and somatic treatment). Subsequently, 'as Finnish law defines long-term as being over three months practically all forensic patients are long-term' (Finland). Finland's expert stated that national data relating to inpatient forensic hospitalizations gathered by the Institute of Welfare and Health (HILMO) estimated a median LoS of forensic inpatients at nine years in 2012, and around five to six years between 2010-2012. Ireland and Spain have formal definitions of 'long-stay' for general mental health services of two or more years.
In The Netherlands, long-stay is defined as a forensic measure lasting for six years or longer (see TBS, as described below). One expert from The Netherlands observed that stays of ten years or more can be seen in designated long-stay services; however, with pressure to reduce LoS in compliance with performance indicators.
[W]e have to reduce length of stay from ten to eight years… you can expect people to stay longer than those ten years, so we have to get a filter for them… (The Netherlands, Expert B) The Netherlands was the only country where a legal definition specific to forensic services exists, and patients may be transferred to specific forensic long-stay facilities once this time has lapsed under a separate legal section.
Other countries that offer a formal definition of 'longstay' include Lithuania, FYR Macedonia and Portugal, legally defined as a lapse of six months, treatment of more than one year, and more than three years respectively. Representatives from these countries were, however, not able to provide national research data regarding LoS in forensic populations. Figure 1 illustrates LoS in forensic care for the 13 countries where such information was provided, ranging from four years in Italy to ten years in Belgium, England and Serbia and The Netherlands. These figures include participant observations of LoS in what they described as a 'longer-stay' population in various forensic services for MDOs, including high-secure populations in England and populations in designated long-stay facilities in The Netherlands.
Generally, between four to eight years was considered 'long-stay' (at varying security levels) in eight of these  Since the time of the present study, new legislation has been introduced in Croatia specifying that patients cannot remain in forensic settings for longer than what would have been their sentence for the same offence had they been healthy and sent to prison. 2 Since the time of the present study, new legislation has been introduced in Italy specifying that patients cannot remain in forensic settings for longer than what would have been their sentence for the same offence had they been healthy and sent to prison.
3 For crimes punishable by imprisonment for less than eight years, length of stay in a forensic mental health service cannot exceed this time (however, in practice this sometimes happens). However, should dangerousness not be found to justify continued stay in a forensic mental health facility (as reviewed every two years), then the patient must be discharged.
countries. Some experts identified that LoS had decreased in recent years: The cross-sectional mean length of stay for the 94 secure beds here is about seven years. That has fallen over the last ten years from being in the region of 12-13 years.
With the new medications and new treatments we do not have many situations in which patients stay for twenty years or thirty years. (Portugal) However, the fact that some patients do spend their entire lives (or a vast amount of it) in forensic mental health services was highlighted in some interviews: There are also patients who stay for actually their whole life. (Latvia) I: What are the long-term prospects for people who don't go back to the community? R: To die in prison, something like that. (Belgium) Offenders who won't be dischargedit's clear they can't get out during their lifetime. So they stay till they die. (Switzerland)

Dedicated forensic mental health long-stay services
Representative experts from eight countries stated that specific services are available for long-stay forensic inpatients, either in a separate hospital or specific treatment wards (Table 3). The representative from Croatia stated in the questionnaire that specific services are currently available to forensic patients who are long-stayers; however, because it was not possible to undertake an interview we were not able to obtain further information. Portugal's expert stated that services were available for long-stay patients, but upon closer questioning it emerged that these services do not differ to those for patients with shorter lengths of stays. The remaining countries currently offering some form of 'long-term' forensic inpatient mental health services include: England, France, Germany, Ireland, The Netherlands and Spain (see Table 3).
Only in The Netherlands are admission criteria for these services standardized by law nationally (under a separate TBS long-stay order: Terbeschikkingstelling, translated as 'at the discretion of the state', allocating a prison sentence followed by a psychiatric treatment order for mentally disordered offenders. The prison sentence serves as punishment, followed by a treatment order to promote reduction in risk of further offending). After having been an inpatient at two separate forensic mental health hospitals for six years or more, where a patient has completed relevant treatment programs but with little discernible progress and no foreseeable reduction in risk from further treatment, they can then be transferred to a long-stay facility following review by an independent national panel. Where other countries, such as England, France, Germany, Ireland and Spain, have specified treatment wards within forensic mental health hospitals, there are no national laws or policies to govern these and so the design of such services is left to individual units resulting in inconsistencies. The expert from Ireland provided a detailed definition of forensic longstay, namely having been under forensic care for at least five years but with no recovery pathway to the community in the foreseeable future.
In Portugal, 'long-stay' is understood as 'forensic patients in inpatient safety measures for an indefinite time'. Patient characteristics in long-stay services were described by experts as displaying violent or dangerous behavior, 'therapeutic non-responders' (or treatmentresistance), those who present a 'danger to society' (having committed violent crimes or presenting with continued violent behavior) and those who are in the service for longer than average or 'indefinitely' (Portugal). Treatment within these facilities includes general psychiatric and medical treatment, however, with less focus towards risk reduction and greater focus on 'wellbeing' (Germany), 'quality of life' (England, Ireland and The Netherlands) or preparation for intensive rehabilitation and educational interventions (Spain).
There is greater emphasis on 'maintenance' and improving standards of living for chronic, treatmentresistant patients who present a continued risk to society, in what would otherwise be a highly restrictive environment.  The expert in England acknowledged that different services are available to different patient groups, with 'low stimulus', 'homely environments' for treatment-resistant populations and a 'recovery-focused' pathway for low-secure, complex-diagnosis populations (with the latter identified as being effective in terms of discharge rates).

Quality of life
The importance of addressing quality of life in service provision and care was generally recognized by all participants: When you finally say, okay, listen we don't know how to get you out of the service [or how] to significantly reduce your risk of reoffending, so you have to stay here… what can we do to improve your quality of life? (Germany) In the countries that offer specialized long-term forensic mental health services, a common theme of quality of life was difficult to measure amongst patients. Difficulty in ascertaining patient-rated quality of life has led to uncertainty amongst practitioners as to how to achieve an improved standard for long-stay populations. …[W]hat was much more useful, in a structured way, was to assess what we deemed their needs as being. And if we deemed what their needs were through the Camberwell assessment of needs… then we have to provide an environment where those needs can be met. (England) Experts from The Netherlands detailed a study regarding comparisons of self-reported quality of life and proxy assessments in The Netherlands (Schel, Bouman, & Bulten, 2015). This research compared quality of life ratings of long-stay forensic inpatients with the ratings of psychiatric nurses, who predicted patient' responses. It was found that there was poor agreement between the patient scores and the nurse's proxy scores, indicating maybe that more staff training on quality of life issues will be beneficial in supporting and optimizing patient's quality of life experiences.
In countries without long-stay forensic services, quality of life was not regarded as high within prison or other mental health inpatient services for longterm populations, with lack of financing and uncertainty in meeting patient' needs as potential barriers to improvement. The experts in Serbia and Slovenia detailed lengthy travelling distance to centralized inpatient services as having a potential impact on a patient's quality of life, in terms of family visits and its impact on treatment. This highlights how it is not only the material conditions within long-stay services that may influence the complex issue of quality of life, but also external factors, including physical distance from family members. Quality of life factors were considered by all experts and remain a conscious focus in improving the quality of care for longstay populations, where there is current uncertainty as to how to improve care and living conditions if treatment progression is not made.

Challenges in the development of long-stay services
Long-stay services in Hessen (Germany) appear to have developed gradually over the past twenty years with little organizational or legal resistance. Some obstacles, however, were noted in the development of these designated long-stay services; difficult to manage and treat patient populations were simply secured in long-stay facilities (including those with personality disorders), which led to patient challenges regarding their right to treatment due to little prospect of recovery or release. Experts in The Netherlands described difficulties in a lack of prescribed criteria regarding admission and discharge of patients when establishing long-stay services, an issue which is still being clarified.
In the countries without separate, designated longstay facilities, according to experts interviewed (England, France, Ireland and Portugal), the term 'long-stay' is not a widely used concept amongst practitioners, nor is it always considered a helpful categorization. Opinions regarding the further development of specific long-stay services were mixed, with ideological and cost-related factors impeding further development.
I think if I was [a] commissioner I'd be a bit worried about [developing long-stay services] because, you know, obviously commissioners want as short a stay as possible in secure care because the cost is so high. (England) [T]he general consensus is that the psychiatrists are unhappy with the long-stay proposal… the right wing is strongly in favor, the left wing is strongly against… but if it's regularly checked [then] I think we need one. (France) …[T]he idea that anybody with a mental illness has a long-term need isn't acknowledged. This isn't just a forensic problem but it's an ideological non-scientific view that nobody with a mental illness… will not recover to complete autonomy in complete independence. (Ireland) …[C]urrently we are trying to deal with lack of funding to provide very basic health services in prisons. (Portugal) In the aim of reducing costs, one expert from The Netherlands contested whether long-stay facilities in fact reduced costs in the long-term.
… [T]he fact is, because these people are older they actually need more care, they need different care than the… regular TBS patients, they need more somatic care, they need more nursing… so I'm not sure that these facilities are really cheaper. (The Netherlands, Expert A) Countries with no long-term services For the remaining countries in which long-stay services for forensic inpatients have not been implemented, three offer a definition of 'long-stay' (Table 4). These definitions, however, apply to all patients receiving healthcare and as such are not limited to forensic mental health services.
Most country's experts stated that there has been an increase in focus on LoS in recent years and most experts observed a typical 'long-stay' of between four to ten years, with care needs not necessarily being met for these typically chronic, treatment-resistant, violent populations. Experts from Poland and Lithuania highlighted that increased LoS is not a typical problem within their forensic mental health care systems, with both country's experts indicating that efforts are currently underway to create a database for the monitoring of patient characteristics and LoS.
The five countries that expressed a need for long-stay service provisions include Belgium, Latvia, Serbia, Slovenia and Switzerland; only the expert from Slovenia was able to confirm that there are current plans to develop long-stay forensic services. Many country's experts commented that more investment and focus is needed on improving and developing regular inpatient forensic care as well as outpatient care before discussions regarding long-stay services can be considered (Finland, Latvia, Lithuania). Anticipated barriers to setting-up potential future long-stay forensic services included institutional barriers, lack of financing and public attitude towards MDOs.
Money, money, money and attitude towards offenders in society. (Belgium) [H]ealthcare professionals are not pushing the issue forward and again this isolation keeps them on the level that they have been working ten years ago. (Serbia)

Ethical issues
Amongst the countries, attitudes to long-stay were mixed and experts raised some pertinent ethical issues regarding treatment philosophies and lengths of stay. The expert in Belgium specifically emphasized that ethical issues regarding coercion and mandatory treatment, as well as the right to have a sexual life, were important factors, particularly when patients are detained for prolonged periods of time. The expert in Serbia identified general institutional ethical considerations, including 'professional isolation' of forensic psychiatrists. Not all country experts identified specific ethical issues concerning the development of long-stay services and anticipated populations.
… I think you have to have a good structure for arguing why you need to continually detain somebody and as long as you've got that clear structure… I don't think the ethical issues arise, to be frank. (England) Experts in Poland and Portugal acknowledged the conflict between the personal freedom and autonomy of patients as well as safety of the public, particularly when patients lack insight into their mental disorder. Providing efficient, effective treatment, beneficial for improving mental health as well as risk reduction and achieving a higher quality of life for patients, is challenging in a forensic environment, in which restrictions are placed upon patients (Buchanan et al., 2011;Mason, 1999). It has been suggested that addressing quality of life for patients within a restrictive forensic psychiatric setting may have an important part to play in improving treatment outcomes and lowering recidivism risks in the long-term (Nieuwenhuizen & Nijman, 2009;. These competing concerns were, unsurprisingly, recognized by all country experts.

Main findings
The purpose of this study was to explore existing longstay forensic mental health service provisions, as well as identify characteristics in potential long-stay populations and services offered in the included 18 European countries. Each country offers some form of mental health care for MDOs either in prison, general psychiatric practice or in forensic settings. Representative experts from eight countries stated that specific services are available for long-stay forensic inpatients, either in a separate hospital or specific treatment wards.
It is clear from the information we gathered that what constitutes a 'long-stay' varies widely between countries, as do treatment philosophies, service provisions and attitudes towards potential long-stay services. What is not clear from the information gathered is whether longstays are due to an inappropriateness of treatment interventions in various jurisdictions, or whether the interaction with service provision promotes recovery or rehabilitation of patients. In order to answer these questions, all countries would need to make seismic efforts to contribute towards establishing an evidence-base for appropriate treatment and outcome measures for particular patient populations, including personality disorder and sex offenders (which at present is limitedsee Khalifa et al., 2010;Khan et al., 2015), as well as improve record-keeping and progress of individual patients. The Netherlands and Germany are highlighted as providing the most well established specific long-stay services. These two countries are identified as having progressed the furthest in the development of long-stay services, but are still in the process of justifying their need in terms of demonstrating that it is possible to discharge patientsnot all are lost to long-stay.
The remaining countries that currently provide longstay services (or 'slow-stream' or similar terms used in England and Ireland) do so in special wards or treatment units, with the aim of improving quality of life and the promotion of wellbeing forming the fundamental treatment philosophy when attempts to engage in traditional or standard models of treatment have failed. Of the countries that do not currently offer specific long-stay services, five expressed a need to initiate the development of such services for their longer-stay populations. Importantly, the label 'long-term' and any potential specific long-stay services carry a political and ideological concern, as identified by some of the countries' experts.
For countries that are in the process of developing long-term forensic mental health services, international dialogue can serve to be invaluable by learning from other countries, particularly those with which we share a common bond in unity under the EU or through our understanding and promotion of international human rights. It is clear that long-stay patients are a reality in many of this study's included countries, demonstrating that care is happening either formally (for countries that currently provide long-stay services) or informally (for those who do not). The perceived importance of developing designated forensic long-stay services rests in recognizing the proportion of patients who do not necessarily respond well to standard treatment and who are still deemed to present a risk to society. The balance seemingly to be had is differentiating between the 'longstayers' and those with shorter stays who are more 'able' to move through services. Indeed, not all countries expressed a need for separate services for 'long-stayers' and 'non-long-stayers'. The Netherlands was the only country included that provided a clear process of assessing patients suitable for transfer to a long-stay facility (TBS), while other countries markedly differed. Other countries also offering some type of long-stay services were not necessarily guided by specific national laws or policies, meaning that the design of such services is left to individual units, resulting in inconsistencies. This is a potential conflict with the CRPD, notably in terms of 'reasonable accommodation', in which persons who are deprived of their liberty through any process should be able to exercise, on an equal basis with others, human rights and fundamental freedoms. With services so varied in definition and delivery between countries (all signatories to the convention) this leads to lack of clarity as to what would constitute 'reasonable accommodation' for the purposes of the CRPD.
As individual country experts indicated, common characteristics associated with long-stay included treatment non-response, chronic mental disorder, and dangerous or violent behavior. Long-stay services emerged in part as a response to this chronicity and treatmentresistance to focus less on risk reduction and more towards improving quality of life, where standard services are not perceived to suitably cater for the needs of long-stay patients, nor provide positive treatment response. An anticipated benefit of long-stay services, particularly in The Netherlands, was also to reduce costs of lengthy stays in low-volume, resource intensive inpatient forensic services; however, country experts cast doubt onto these expectations, with an aging population requiring much higher, costly levels of care and support.
With a move towards longer periods of care, it is understandable that concerns are raised regarding potential (re)institutionalization of patients. In conflict with this concern is the political and societal ideology of risk-based containment integral to modern day European society (Priebe et al., 2005). The responses of experts involved in this study largely confirm this conflict, with no current answers on how to overcome the complexities of balancing patient's rights whilst ensuring public safety.

Study limitations
To our knowledge, this is the first exploratory study that has compared the availability of long-stay services from 18 countries within Europe. We were able to interview individual experts with extensive experience, training and knowledge who could offer authoritative observations regarding service use, implementation, current provision and future prospects of long-term forensic care in their countries. This study also builds upon the existing (now relatively dated) comparative work (Salize & Dressing 2005;Salize, Dressing, & Kief 2007;Salize, Dressing, & Peitz, 2002). This study contributes towards this developing area of long-stay service provision, with both questionnaires and follow-up interviews adopted in methods. Further, the addition of two former Soviet Socialist Republics (Lithuania and Latvia) and four former Yugoslavia Republics (Croatia, Macedonia, Serbia and Slovenia) provides for more inclusive discussion, and sheds light on service provisions in countries with still a relatively young history under modern governments. In acknowledging these strengths, weaknesses must also be considered.
Firstly, all experts interviewed were able to provide insight into their local practices, which may not necessarily be representative of their entire country. In the same vein, interview data rely largely on the observations and impressions of individual experts, rather than empirical data on LoS, recidivism and risk, and differences in treatment approaches. Further research should be directed towards these areas, as well as gauging the use of evidence-based treatment practices.
Secondly, each country operates under differing governments with widely varying populations and markedly different histories (particularly in the provision and practice of psychiatry and the availability of experts in the sub-specialty of forensic psychiatry) making comparison between countries difficult, with heterogeneity of concepts, legislation and practice greatly expected. This heterogeneity, however, serves as a reminder of the pluralism and diversity within Europe.
Thirdly, all questionnaires and interviews were conducted in the English language, which presents a challenge in fairly representing and interpreting participant's responses for non-native English speakers (Van Nes, Abma, Jonsson, & Deeg 2010). Contextbounded concepts may also pose an obstacle to 'effective and meaningful international comparison' (Hantrais, 2009). However, we sought to understand the interpreted experience of the participants and represent their opinions and responses in a meaningful way within the appropriate contexts. We also sent participants their transcript to read and amend as necessary to ensure we captured the correct meaning.
Also, not all European countries were included in this study; therefore results are not generalizable or representative across Europe.
Finally, all participants interviewed were psychiatrists, academics and other highly trained specialists in the field of psychiatry or forensic psychiatry; however, patient voices are missing from this discussion. In particular, patient perspectives on quality of life from those currently within long-stay services would be invaluable in order to paint a more holistic picture.

Conclusions
Our findings have important implications for policy and service developments. Efforts are needed to identify a definition of what constitutes 'long-stay'. Without a clear definition, whether based on actual years of detention or a measure relative to the average length of stay, it is not possible to develop specific policies for this patient group. Given the specific needs to long-stay patients such specific policies are necessary, whether or not they include separate service provision.
Further research should encapsulate the patient perspective of long-stay service provisions, as well as performing longitudinal outcome and economic analyses of existing long-stay forensic service provisions. Such observational research will inform us about how longstay services are currently performing, their associated costs, and the attitudes of their patient populations and allow the development of best practice recommendations for this group.