Reforms of mental health care systems are fashionable. Google gives 12,300,000 hits on the term ‘mental health reforms’. Most of them refer to policy interventions for restructuring mental health care into an integrated care system. In the United States, for example, the President's New Freedom Commission on Mental Health came to the conclusion in 2003 that ‘the mental health delivery system is fragmented and in disarray ... lead[ing] to unnecessary and costly disability, homelessness, school failure and incarceration’. Only with a better integrated system of mental health care can these mental health problems be tackled [1, 2].

The conclusions of the New Freedom Commission are not isolated. In Canada, the Ontario Mental Health Care Reform has come to the same conclusions—integrated mental health care is a must in the care for patients with mental health problems (see http://www.ontario.cmha.ca/content/mental_health_system/health_care_reform.asp). In the UK, the Mental Health Legislative Reform, known as Brian's Law, is part of the government's plan to create an integrated, comprehensive, balanced and effective system of mental health services ( http://www.rcpsych.ac.uk/pressparliament.aspx). In Australia, mental health care is in a state of crisis, due to the fact that there are too many people providing care independently. According to Andrews (who describes mental health care as a chariot with too many horsemen) integrating the various care components can improve services [3]. It all seems so evident—good care for the (mental) health of the population needs integrated mental health care services.

The literature is less abundant on the realised effects of integration [4–8]. Integration of (mental) health services seems indeed to lead to better health care, but the effects at the patient level are quite modest. Different arguments have been put forward to explain the modest effects, such as staff morale [9] and lack of coherence among implementation strategies [10]. Kodner and Spreeuwenberg suggest that integration becomes more effective the more dimensions of integration are implied. According to them integration should be divided into at least five dimensions; funding, administrative, organisational, service delivery and clinical [11].

Are the effects of integration indeed related to the levels of integration implied? The Dutch mental health care sector provides an excellent case for testing this hypothesis. In comparison to other mental health care systems in Europe and the U.S., integration already has a long history and has been developing over the years. Moreover, due to the integration of mental health care institutions there is no fundamental distinction anymore between the care for patients with severe mental disorders and those with other mental health problems. All of them can be treated within the integrated mental health institutions [4, 12–15]. Despite these efforts, there are serious doubts about the quality of mental health care in the Netherlands. In the last few years, the mental health care institutions have become prone to a public discussion on their functioning [14, 16]. The recent National Commission for Mental Health [16] has even proposed to dismantle the integrated mental health care institutions and to accommodate the various functions of the mental health care services into the general health care sector. What has happened in Dutch mental health care and how is the critique on its functioning related to the way integration has been worked out?

In this paper we will summarise the Dutch process of integration and relate it to the dimensions of Kodner and Spreeuwenberg [11]. We will show that the way the integration is realised fits very well with these dimensions. However, the problems raised in mental health care cannot be explained by an unfinished process of integration. We will argue that the Dutch case is an example of Leutz's third law of integration: ‘Your integration is my fragmentation’ [17]. The Dutch case shows the risks for possible side-effects of the process of integration itself.

During the last three decades, much efforts have been put into the organisational development of the institutions for mental health care in the Netherlands (second level mental health care). In the last years of the 20th century this process has led to the emergence of large so-called ‘integrated institutes for mental care’. Primary care has stayed outside this process of integration.

The Dutch ‘integration movement’ has a long history. Already in the 1960s, mental health care professionals and policymakers warned about the lack of coherence in care supply [18]. Psychiatric institutions and the psychiatric departments of general hospitals provided inpatient care. Small outpatient institutions worked independently from these inpatient facilities for various groups of patients. There was no structural collaboration and on the management level, contacts were defined by the religious denominations; Catholics dealt with Catholics and Protestants with Protestants, and so on [18].

In the early 1970s, the plea for collaboration became strongly empowered by a broad western ‘anti-psychiatric movement’. It resulted in a new vision on good care [19, 20] and led to the emergence of three different mental health care circuits; one for inpatients, one for outpatients, and one for day care. In the outpatient care sector, where the ideology of psycho-hygiene, prevention, and public mental health dominated [19], optimistic care and treatment concepts evolved. The most important exponent of this treatment optimism was the formation of the Regional Institutes for Outpatient Mental Health Care (here called the outpatient care centres) in 1982 as a result of mergers between several small institutions. This merger was strongly supported by the government, which promised legislation to finance these outpatient care centres by introducing the Exceptional Medical Expenses Act (AWBZ). Henceforward, all the care offered by these outpatient care centres, including psychotherapy, was totally remunerated by the government.

In the inpatient care sector, modernisation and humanisation of the institutions was realised by a programme of rebuilding aimed at constructional as well as functional adjustments. This led to the birth of the psychiatric hospitals. The ideology became that of clinical psychiatry. The classical concept of total care—as in the psychiatric asylum—lost its power, whereby the psychiatric hospital as a home for chronic psychiatric patients was no longer the only and obvious solution. Sheltered housing schemes were set up, that would gradually develop into the independent semi-mural section of the mental health care sector. The psychiatric hospitals as well as the semi-mural section were also included in the Exceptional Medical Expenses Act (AWBZ); all their activities were reimbursed by the government, which made the national health insurance system, Medicaid, responsible for the distribution of the money.

In the 1980s a new group of patients emerged as a result of the above-mentioned reforms. These were patients who were initially admitted to the psychiatric hospital, and later had returned home and fell under the care of the outpatient care centres. Instead of becoming the winners of the reforms, they became its victims. On the one hand, the psychiatric hospital lost its responsibility for this ‘new’ group of outpatients, on the other hand, the outpatient care centres were not very much inclined to care for these ‘untreatable’ patients. After serious criticisms from society, policymakers obliged both types of institutions to organise continuity of care for these patients. In practice, this resulted first in new collaborations and next to the search for an integrated, undivided mental health care sector based on a geographical (regional) structure [13]. In the 1990s a large number of mergers between institutions were realised. By the year 2000, nearly all of the psychiatric hospitals and the majority of the outpatient care centres had been merged [4, 15]. Sometimes also the psychiatric departments of the general hospitals were involved, and sometimes the institutions for sheltered housing.

In 1990, the situation of mental health care was as follows: 56 regions had their own outpatient care centres. The 39 psychiatric hospitals too had a regional function, but originally they belonged to different denominations (40% were Roman Catholic, 40% Protestant, and 20% neutral or ‘other’). Beside these two types of institutions, there were 76 psychiatric departments of general hospitals and more than 40 institutions for sheltered housing.

Fifteen years later, in 2005, mental health care was provided by integrated institutions in almost all parts of the Netherlands. The outpatient care centres and the psychiatric hospitals were always made part of these new integrated institutions. The majority also contained one or more psychiatric departments of general hospitals, whereas only a few included institutions for sheltered housing. Finally, in a few cases, institutes for drug addiction therapy or institutions for psychiatric detention orders participated in a merger.

At the time of writing (2007) more than 80% of the care supply is provided by 39 integrated institutions, whereas 10% is given by a limited number of independent institutions [21]. The other 10% is provided by a small but flourishing number of private practices. In total, mental health care is estimated to involve more than 65,000 jobs, of which 1800 positions are for psychiatrists and 25,000 for nurses. Together, these institutions are responsible for a budget of €3.5 billion [22]. Up until 1 January 2008, all the mental health care institutions will be funded by the government. For this purpose, a special insurance is in place, the Exceptional Medical Expenses Act (AWBZ). Second level mental health care is free except for those patients who receive psychotherapy and are expected to make a small contribution.

Each year approximately one million people in the Netherlands make use of these services. This amounts to 4.9% of the population. Taking a yearly prevalence for mental disorders of 24.4 per 100 inhabitants [23] we can establish that for every 100 patients with a mental disorder, 20% make use of mental health care services. Incidentally, the differences per disorder are large. Out of every hundred patients with a mood disorder, for example, 36.1% are treated in one of these institutions. For those clients with a drug abuse problem or addiction, this amount is <10% [24].

The described process of integration can be interpreted as a hierarchical or ‘top down’ process driven by more generalised organisational exigencies for optimisation [11]. Due to all kinds of practical obstacles, lack of mutual understanding, competition, different funding streams, institutional and professional ultures, which are typical for the architecture of most health care systems, the separate organisations did not work well together, which interfered with efficiency and quality goals. Therefore, the fulfilment of system aims necessitated cooperation and collaboration between them. Integration was the ‘glue’ that bonded the entity together, and enabled the achievement of common goals and optimal results.

In this organisational view of integration three strategies should be fulfilled [11]. The first strategy aims at the realisation of common funding, this is because the division, structure and flow of funds for health care affect virtually all aspects of integrated care. The second strategy concerns the fine tuning of administrative systems. The manner in which government regulatory and administrative functions are structured can help to eliminate programme complexities, streamline eligibility and access, and better manage system resources. The third strategy concerns organisational measures. Networking and collaborations are major methods to improve how organisations work together.

The described integration of Dutch mental health care includes all the three strategies. The mergers have led to a common flow of funds, one administrative system, as well as one straightforward organisation. >From this perspective, the process of integration has been very successful. Dutch mental health care has gained a very complete degree of comprehensiveness and formality in integrated care. However, integration is not complete with this organisational process of mergers [11]. What is also needed is a second, more patient-centred and ‘bottom-up’ process, in which the characteristics and needs of specific patient groups determine the content of integration.

Patient-centred integration implies a bottom-up perspective on integration: the process of care should follow the logic of the patients and their course through the health care system. Concepts such as continuity of care and disease management are exemplary for this perspective.

The process of patient-centred integration relies mainly on integration strategies concerning service delivery and clinical aspects [11]. Service delivery strategies are dealing with delivery and management, i.e. how staffs perform their responsibilities and tasks, work together, and relate to patients. These strategies include service access, continuity and co-ordination of care. The clinical strategies concern shared understanding of patient needs, common professional language and criteria, the use of standards, communication and feedback. As will be shown below, all of these strategies have been adopted within Dutch mental health care and have ensured that the disorders of patients have become the guiding principle in the organisation of care.

One of the major problems in the new merged institutions was the lack of clear lines of responsibility with, as a consequence, a lack of unambiguous framework for interpreting the needs of patients and translating them into distinctive treatment programmes.

Within the outpatient care centres the multidisciplinary approach was especially important [25]. In the diagnostic phase, patients were seen by various disciplines, and decisions were made by the multidisciplinary team. The treatment was firmly directed at structural changes, showing that even within the environment of social psychiatry, the psychotherapeutic perspective dominated [26].

In the psychiatric hospitals, the ideology was defined by social and clinical psychiatry [27]. The psychiatrist was considered the primus inter paribus of the interdisciplinary team. Other professionals made important contributions but it was clearly the psychiatrist who took ultimate responsibility and who decided what kind of psychiatric treatment was needed.

On the psychiatric wards of the general hospital [28], the treatment programme was set up around the medical specialist—the psychiatrist. In accordance with the somatic specialists, it was the psychiatrist who, on the basis of the medical diagnosis, determined what kind of treatment and care the patient would receive. His working method was characterised by short consultations, largely relying on the prescription of medication as well as the large number of patients that he saw on a daily basis.

How did the new merged institutions deal with these fundamental differences? Remarkably, there was hardly any discussion on this matter; psychiatrists simply were put in strategic positions in the institutions. The Health Care Inspectorate strengthened this movement. In addition, the increase in pressure seen in the number of registrations meant that the multidisciplinary treatment ideology quickly lost ground. Finally, also the Netherlands Psychiatric Association in its turn, set as a ‘rule’ that psychiatrists had to be responsible for the diagnosis.

The repositioning of the psychiatrist evidently had consequences for the other disciplines. A significant change concerned the psychotherapist. They quickly lost their power, with the result that many psychotherapists left the merged organisations and set up in private practice [26]. A comparable development took place with the vocational professions (social worker, social pedagogical worker). Their place was taken over by (social) psychiatric nurses.

In the first years after the mergers, the supply of care was offered in the traditional way; outpatients remained patients of the former outpatient care centres, and inpatients as if their clinic was not merged. Gradually changes were introduced. First, new regional centres were created. Here, the professionals of the former institutions had to work together in providing care for the inpatients as well as the outpatients. A second development was the redesigning of the process of care around groups of patients with the same problems and disorders. Care was provided in so-called care programmes [29]. Care programmes in mental health care are multidisciplinary in nature and describe the various diagnostic steps as well as the special treatment interventions involved in the total care process. Most care programmes are based on diagnostic categories, for example, programmes for clients with an anxiety or a psychotic disorder. The most sophisticated care programmes also describe the organisation of care chains and the subsequent steps in the care process. With programmes a better knowledge of the different care arrangements is ensured as well as more coherence in the care supply. They also help to realise more harmonisation between the various disciplines and bring more clarity in the roles of different disciplines.

The introduction of care programmes gave an impulse to the further elaboration of quality systems. In the early 1990s these systems were still protocol driven; the process of care was described, the work of different care providers and departments were attuned, procedures were set down and agreements were formalised. Later the quality systems became more oriented on output measures and performance indicators were introduced [30].

The repositioning of the psychiatrist and the implementation of new care programmes exemplified a much broader transformation within mental health care: the introduction of a common and unifying vocabulary. The effects of these changes in language had not only direct consequences for the organisation of care, but as will be argued below, had a profound impact on the way professional expertise became conceptualised.

The new unifying language was the language of the specialised medical-psychiatric perspective. The needs of patients became translated into the unambiguous language of classification of psychiatric disorders (the International Classification of Disorders (ICD) or the Diagnostic and Statistical Manual of Mental Disorders (DSM)), whereas the professional answers to that need became defined in terms of cure (treatment) and care (support, nursing, etc.). What evolved was a standardisation of both diagnostics as well as interventions, in such a way that these interventions became understood as the central and causal agent for the results of care. The better the intervention, the more effect could be realised. In this view, professionals were no longer seen as responsible for the therapeutic process, but became executors of standardised interventions. Good care became thus the supply of the appropriate intervention given the disorder of the patient.

This development is nicely illustrated by the Dutch multidisciplinary guidelines for diagnosis and treatment of mental disorders [31]. These guidelines are diagnosis-specific and put much emphasis on the importance of good medical diagnostics. Subsequently, the evidence-based character of the guidelines ensures that the recommendations almost exclusively concern only those interventions that are aimed at symptom reduction. Whilst the guidelines are referred to as multidisciplinary, there is hardly any attention paid to the individual position of the caregiver [31].

So what does the patient experience from this development? Firstly, more standardised diagnostics, which result in a DSM classification. Given this DSM classification, it is clear what kind of evidence-based intervention should be offered. According to the figures of the integrated institutions, this way of care supply leads to an augmentation of treatment contacts as well as an increase in the average treatment duration. Furthermore, the patient may observe a number of more subtle changes. Better procedures, better teamwork, and higher chance that a psychiatrist is involved, as well as higher chance that pharmacotherapy is made part of the treatment.

The integration of Dutch mental health care contains organisational as well as patient-centred elements. From the perspective of integrated care, one could say that the Dutch approach has been quite successful and may be a good example of how policymakers, managers, professionals and even patients have worked together on the redesign of mental health care. However, is the result as appealing as it looks?

Table 1 shows the registration and admission data of mental health care institutions in the period 1980–2005. During these years, the various types of institutions published their own national yearly reports on care consumption. After 1997 the registration system changed, making comparisons more difficult, whereas some other figures were no longer presented. Therefore, the figures for 2005 are just approximate estimates.

The yearly incident statistics show the number of new admissions or registrations per year per 1000 of the population [23]. The yearly prevalence concerns the number of new registrations and the number of clients still under care on 1 January of that year. In 2005 more than one million registrations for mental health care were counted, out of a total population in the Netherlands of 16 million.

The Dutch specialised mental health care system seems to be in good shape: most institutions are integrated and offer both ambulatory, day-care and clinical care. More than 90% of patients receive treatment that is part of an integrated care chain. If we see the number of admissions as a sign of growing confidence in mental health care, then integration is a complete success. But unfortunately that is not the case.

In the late 1990s, the criticism of the integrated mental health care institutions became louder. Patients complained about the long waiting times before receiving help, whereas the integrated services reported expanding waiting lists. Additional funds, which were offered by the Minister of Health in the late 1990s, did not resolve these problems. On the contrary, despite all efforts, the waiting lists only grew [32].

Meanwhile, reports mentioned an increase of mental health patients with complicated disorders who were not receiving any health care at all [24]. The sector was blamed for the subsequent social problems that ensued, such as homeless or destitute persons and violence on the streets. The Council for Public Health and Health Care remarked that long-term care-dependent patients did not reintegrate sufficiently back to ‘normal’ life and remained too much under the control of the mental health care institutions. It gave out a warning on the negative effects that chronic psychiatric patients would cause within society [14].

At the same time, the sector was attacked because of a lack of clarity regarding its functions [33]. The Council found that the mental health care sector was not making enough choices, which resulted in a lack of clarity on its functions. Further, they postulated that if specialised mental health care continued along the same lines, the ultimate consequence would be discontented patients. This, indeed, is what happened. The patients who were treated within mental health care were not enthusiastic about the treatment they received. They reported less satisfaction and a decrease in the quality of care. The general satisfaction on the mental health care services significantly decreased during these years [34, 35]. Moreover, patients mentioned elaborate diagnostic sessions, having no influence on the treatment they received and also treatment periods that were too long and too intensive.

The new merged institutions tried to realise patient-centred integration by focussing on a medical psychiatric approach. By doing this, they made more differentiation between their own approach and that of the primary care suppliers (general mental health care). What were the effects of this process on primary care?

In the Netherlands, primary care for people with mental health problems includes that of the general practitioner (GP), social worker and psychologists in primary care. Table 2 presents the registration figures for the period 1980–2005. The largest increase in care on this level is that of the psychologist which started after 1990. Also the social workers provide more mental health care. But mental health care provided by the GP (by far the most important care provider in primary care) has decreased from 150 to 108 per 1000 inhabitants. This retirement of the GP from mental health care has even led to an overall decrease in primary mental health care in total [36]. GPs see the treatment of patients with severe psychosocial problems less as their responsibility. This retirement has not been compensated by other providers of general mental health care [37, 38].

The decrease of care within primary services and the increase in specialized mental health care are related. Care for people with mental health problems has become more and more the exclusive domain of specialized health care professionals. Although this development of leaving care to specialized professionals is probably not restricted to mental health care, the consequences here are quite dramatic.

In 2003, the Dutch Commission for Mental Health Care concluded that the trend of a decrease in care provided in general practice went much further than the GP, it also concerned other professionals not specialised in mental health care [16]. These professionals, such as teachers, community workers, police, and judicial authorities refer clients more quickly and more often to specialised mental health care professionals and institutions. The Commission called this process the monopolisation of mental health care.

According to the Commission, monopolisation has a spiral effect. Informal carers (working below the primary care level) refer more quickly to primary care services, which make way more quickly for professional care from the specialised level. For their part, these professionals in specialised care determine more quickly where the gaps are concerning the informal care and primary care levels. Subsequently, in order to prevent the clients from any suffering, they jump in and fill the gaps where necessary. However, their attempts to strengthen the informal and primary care levels have the opposite effect of weakening them.

The real effect of monopolisation is that mental health care has come to mean ‘care of few’ whereas, as the Commission's 2003 report suggests, it should be ‘care of many’. Another result is that the integrated mental health care institutions increasingly do not match the expectations of the many clients seeking help. At the same time, it explains why so many chronic psychiatric patients do not receive the care they need, which results in more homelessness and more social harassment.

The Dutch Commission postulated a direct relation between the effects of integration and the current problems in mental health care. The commission argued that the mental health institutions themselves were at least partly responsible for what had happened. For years they were so busy with integration that they did not focus on the effects of this process on mental health care providers outside their institutions. They were so eager in stimulating citizens with mental health problems to seek specialised mental health care, that they forgot to ask the vital question of what specific roles should be involved (client, informal carers and other professionals) for bringing relief.

Again, the distinction between organisational and patient-centred integration could be helpful in understanding what had happened. As an effect of organisational integration, frontiers between the former institutions were demolished, but new borders were erected. These borders aggravated the gap between primary care and the integrated institutions. The process of integration within the merged institutions led to a process of disintegration between the merged institutions and the other providers of mental health care. Integration caused as a side-effect disintegration. But also the process of patient-centred integration probably had a strong side-effect. The unification of language within the merged institutions could have led to larger communication problems with other health providers. Moreover, one could argue that the new specialised psychiatric vocabulary became so dominant, that other health professionals were more or less forced to use the same language. This was at least what happened with the introduction of multidisciplinary guidelines in mental health care. These guidelines were ‘DSM-proved’. Professionals in primary care were thus obliged to define their own expertise in the same vocabulary as their colleagues in specialised mental health care, with the paradoxical result that they had to give up their own unique position in the chain of care.

In many western countries, the mental health care sector has been faced with the problem that groups of patients receive either no or insufficient care for their needs. This is a result of professionals and institutions working at cross purposes. Especially the most vulnerable clients, who are unable to formulate their request for care in such a way that one specific provider can offer a solution, are the victims of this situation. The theoretical solution for this problem is clear: more harmonisation in the various care components by the integration of care. This creates a seamless pathway for the patient as they make their way through the various parts of the mental health care system. Integration should provide what Andrews calls ‘one horseman for the chariot of mental health care’ [3].

The Dutch mental health care services have apparently developed over recent decades towards one integrated care sector. Organisational as well as patient-centred integration have been successful. However, these two forms of integration have not made the problems disappear. On the contrary, although it is true that collaboration between mental health care professionals has ameliorated the situation and each region gets only one ‘horseman’, the paradox is that this has caused more and more criticisms. We think that the Dutch integrated system shows the truth of one of Leutz's laws on integrating services: ‘Your integration is my fragmentation’ [17]. Integration has led to the situation where mental health care has become the exclusive domain of small groups of specialised care professionals (monopolisation) who, with an eye towards improved integration and harmonisation of the care, work more and more from the perspective of medical specialist care. In this discussion we place the most important conclusions in a row and pose the question what the Dutch situation can teach us about the paradigm of the integration of care.

The integrated institutions cannot be held solely accountable for the monopolisation of mental health care. Parallel to the integration within the institutions for mental health care, a development has taken place in society in which citizens have withdrawn from providing care and increasingly leave these tasks to the professionals. At the same time, the general practitioner has withdrawn from giving care to people with mental health problems, partly as a result of his/her so-called reduced task perception [16]. This change in the position of the general practitioner is, in turn, part of the general crisis within general practice, reflected by complaints on the part of GPs of large workloads, demanding patients and long working hours that have resulted in an increasing shortage of GPs. The monopolisation of mental health care is, in other words, not only the result of mergers between institutions but also the effect of developments within the broader social context.

The Dutch situation clearly shows that the integration of care requires more than closer collaboration between the various providers of specialised mental health care. Integration requires an intrinsic and broadly shared vision on good care for the various target groups. This also applies to the other people involved including general health care and even social services. ‘Broadly shared’ is the term here that should be sharply underlined, because this is where it has gone wrong in the Netherlands. Within the institutions, the choice has been made for further specialisation, while there was hardly any shared vision with primary care on the question of harmonisation. Clear objectives can only be formulated if there is a collective and shared vision, which is monitored regularly to see if it is on course. This did not happen in the Netherlands. If however it had occurred, then it would have been clear earlier that integration leads to monopolisation and that with this the ideal of integration (improved harmonisation of various care components) is rather further away than nearer.

But even if that harmonisation had been achieved, we still have to ask the question whether integration would and could have led to the intended effects. Integration, we suspect, can only succeed if the principles of the differences of the supply of the various care partners are recognised. Within mental health care the framework for naming and evaluating these differences is lacking. This is connected to the second side-effect mentioned previously, that of a further standardisation and specialisation of care. This development goes much further than the mental health care sector and is to a large extent based on the emergence of the evidence-based ideology, which states that care should be based on scientific evidence as much as possible.

Within the tradition of evidence-based medicine and mental health, mental health problems are primarily seen as symptoms of a disorder and interventions are subsequently judged on the degree to which they contribute to a reduction in symptoms. This ideology has turned out to be unusually productive for improving the quality of care, mainly due to the fact that the focus is on the effects of the interventions. In this tradition, the placement of the intervention in the care process and the person who performs the intervention, in other words, the context, comes second place. If the focus is put on these contextual variables, then it is usually in the area of cost effectiveness. The result of this is that this ideology is less suitable for providing a good differentiation of care as provided by informal, generalised, and specialised care providers. It is more likely that the opposite is true: in this ideology it is nearly always the specialists in the area of the specific intervention who perform it the most effectively. In this manner, specialised care becomes the implicit normal standard to which other care providers aspire. The effects of this have become visible—the standardisation and homogenisation of care result in non-specialised care providers being almost automatically encouraged to withdraw from giving the care because it is a foregone conclusion that they are less good at it.

None of this would be a problem if there was one and only one common goal within the mental health care services. The actual daily practice shows that this is not at all the case. Very often numerous other objectives are aimed for, such as strengthening patient autonomy, a reduction in the influence of contextual risk factors, minimising the chances of iatrogenic damage, and changing adaptation strategies. According to the World Health Organisation, other more global aims such as social and civil integration, emancipation and citizenship should also be addressed in the (mental) health care sector. The number of actual aims of (mental) health care is so great that the World Health Organisation has even developed a separate classification system for them. This system is the International Classification of Functioning disability and Health (ICF) ( http://www.who.int/classifications/icf). The clarification of these aims demonstrates that mental health care can do and should do much more than specialised medicine can offer. It is at this point that the individual value and the strength of generalist care and even that of non-professional care should come into view.

This broadening of the perspective should in no way put a strain on the original approach of evidence-based medicine, such as proposed by Sackett et al. [39]. Evidence-based medicine begins not with a description and application of effective interventions, but with the formulation of questions on care practice as well as searching out the most appropriate intervention for a specific problem that needs a solution [40].

On the basis of these experiences within the Dutch mental health care sector, questions should be asked concerning the real nature of integration. If integration always leads to monopolisation and homogenisation of care, then we should surely question whether the solution is not worse than the problem. Are monopolisation and homogenisation not a specific sign of failed integration?

The integrated care paradigm puts the harmonisation and coherence of care at centre stage. The aim of this is clear, only through the harmonisation of care can the patient receive the help she or he needs both quickly and adequately. Harmonisation leads to less duplication of care and enables any gaps in the care to be discovered and solved quickly. Integration assumes that care providers know how to find each other on an equal basis [11]. This requires (as Andrews states) clear mutual agreements to be made on responsibilities and direction [3].

It is, however, questionable whether or not integration requires just one horseman. The Dutch situation has shown that the horseman (merged institutions) has monopolised the definition and the solution of the mental health care problem, with the result that other care providers have withdrawn. With this, the integration paradigm has become the playing field of the power of struggle going on between the institutions and disciplines that is described so beautifully in the literature on professionalisation. Our view is that integration can only be successful if the qualitative differences between the various care providers are recognised.

Helen Lester, Prof., National Primary Care Research and Development Centre (NPCRDC), University of Manchester, Manchester, UK.

Peter Verhaak, Ph.D., Research coordinator for mental health services research at the NIVEL, Netherlands institute for health services research, Utrecht, The Netherlands.