Guidelines in mental health – National and international perspectives

High-quality clinical guidelines remain the most elaborate and well-developed expression of evidence-based medicine in clinical care, combining a systematic approach to reviewing the evidence with multi-professional and service user involvement, along with robust systems for consultation and review. This approach to clinical guideline development is a response to previous criticisms of clinical guidelines, for example, the limitations of the uni-professional guideline (Grilli et al., 2000) and the need for more robust methods for ensuring and evaluating quality in guideline production (Cluzeau et al., 1999). However, concerns about the quality and value of clinical guidelines continue to be expressed (Grol, 2010; Shaneyfelt & Centor, 2009). These concerns include the cost and speed of development, over-reliance on expert consensus, the misuse of guidelines as performance measures, the promotion or suppression of professional interests, and the considerable challenges of effective implementation.

These concerns are important. Equally important, and in many ways more pressing, is the place that guidelines have in health and social care, and the purpose of these products. Essentially a distillation of what we can claim to know, guidelines are subject to all the imperfections that any knowledge brings: is it reliable, is it useable, is it focused on what is really important? Much of the work described in this special edition, which draws heavily on the work of the National Collaborating Centre for Mental Health (NCCMH), attempts to address these concerns.

The first three papers by Clark (pp. 318–327), Barnes & Paton (pp. 328–335) and Worrall (pp. 336–341) address the very important issue of implementation of clinical guidelines. Clark describes the development of the Improving Access to Psychological Therapies (IAPT) programme, an initiative of the Department of Health in England to train over 7,000 additional psychological therapists over a 6-year period. This programme of implementation takes as its starting point the recommendations provided by the National Institute for Health and Clinical Excellence (NICE) guidelines on the treatment of depression and anxiety disorders. It is the largest single implementation programme for any NICE guideline. By completion of the implementation phase of the IAPT programme (in 2015), it is expected that an additional 900,000 people will be in receipt of NICE recommended evidence-based psychological treatments.

In his paper, Clark describes the main components of the IAPT programme including the range of interventions provided, the training programmes specifically designed to develop a new workforce based predominantly in primary care and skilled in the evidence-based psychological interventions needed to implement the NICE guidelines in common mental health problems, and the systems for service delivery and organization, including a stepped-care approach and routine outcome monitoring. The paper also reports some data from the early pilot programmes with recovery rates (with an aim of matching those achieved in RCTs) of over 50%, and an average of 42% for the first year of the national roll out programme. Clark also highlights important remaining challenges for IAPT, including the provision of evidence-based interventions apart from CBT (the primary focus of the initial programme), and the integration of the IAPT services with other psychological treatment services.

A second paper on the implementation of NICE guidelines by Barnes & Paton (pp. 328–335) describes the work of the Prescribing Observatory in Mental Health in the United Kingdom (POMH-UK). This highly innovative and successful initiative has employed a quality improvement approach focused on the quality of prescribing in mental health. Having developed a subscription network of healthcare organizations throughout the UK, POMH-UK established important benchmarks for prescribing a number of different drugs including antipsychotics, cognitive enhancers and lithium based upon standards derived from NICE guidelines. Barnes & Paton also identify areas where organizational barriers obstruct improving practice, such as in lithium prescribing. This systematic national network has made significant changes in the prescribing of psychiatric drugs, including in highly topical areas such as the use of antipsychotics for people with dementia.

Adrian Worrall has a key role at the Royal College of Psychiatrists’ College Centre for Quality Improvement (CCQI), which now leads the field of quality improvement in mental health in the UK. In his paper (pp. 336–341), he summarizes key projects developed at the CCQI and adumbrates a number of themes explored by Barnes & Paton, but realized not just in national quality networks. The CCQI has, in recent years, aimed to improve standards of mental healthcare across a range of services and practices through accreditation schemes. Worrall describes how they have used a broad range of data to underpin accreditation to ensure the local structures are in place for continuous quality improvement, using a framework to support guideline implementation. Anticipating the now developing programme of NICE quality standards, Worrall outlines a model for the development of NICE guideline-derived quality standards and how these standards should link to clinical guideline implementation. In his very practical paper, Worrall describes how to establish such schemes, the crucial importance of linking with professional bodies in their developments, and how the CCQI have linked up with national registration bodies to develop and undertake national sentinel audits to encourage quality improvement across huge numbers of trusts.

In our next four papers, Kendall et al. (pp. 342–351), Harding et al. (pp. 352–357), Hopthrow et al. (pp. 358–364), and Gould (pp. 365–370) attempt to address a variety of methodological issues with the aim of improving guideline quality, relevance and implementation. Kendall et al. address two key issues that have important but contrasting effects on the quality and uptake of clinical guidelines. The first issue, under the broad heading of bias, addresses some factors and practices that may undermine the positive value of guidelines and highlight the growing importance of removing bias from guidelines and, more generally, healthcare and clinical decision-making. Key issues they identify include problems in the selection of interventions included in guidelines; the methods used in psychological trials and the problems of comparators and meta-analysing disparate evidence sources. Also highlighted are selective publishing and reporting by the pharmaceutical industry and its impact on clinical guidelines and decision-making, the sometimes contradictory methods and guidance offered by different organizations dealing with similar evidence, and the effects of mass marketing on the development of guidance. Kendall et al. also take up the issue of service user involvement in guideline production. They provide a contemporary historical approach and argue, from the perspective of guideline developers, that service user involvement has an important impact both on the way in which a guideline development group operates and the final content of the recommendations, and suggest that this may be an important, if not crucial turning point in the power relations that pervade mental health.

Harding et al.'s paper has been developed from the perspective of service users and addresses some common issues. They too argue that service users have a crucial role to play in the development of clinical guidelines, and describe the problems and challenges encountered in attempting to do so. They locate the issue of service user involvement in guideline development in the wider context of service user participation in mental health services, and argue for a link between the development of clinical guidelines and the work of the recovery movement (NIMHE, 2005) and shared decision-making in mental health (Duncan et al., 2010).

Hopthrow & colleagues (pp. 358–364) address the important, but often neglected issue of decision-making in guideline development groups, drawing a contrast between the well-developed methods for the identification and synthesis of evidence with the much less well-developed methods for decision-making in groups. Their paper provides a very useful overview of some of the important social psychological influences on decision-making (this contrasts nicely with the work of the GRADE group (Atkins et al., 2004) on methods for better characterizing the quality of the evidence as an aid to decision-making); but they are clear in pointing to the need for further research in this important area.

The final paper in this group, by Nick Gould (pp. 365–370), looks at the collaboration between NICE and SCIE in the development of the dementia guideline and the challenges in addressing the concerns of social care professionals about the nature of evidence and the manner in which recommendations are constructed and presented. Gould's paper highlights not just methodological issues but also serves as an important reminder of the role that social factors play in all aspects of health, not least in mental health. As in other papers Gould has written on the subject, he makes the strong case that, especially for people with longer term and intractable conditions such as dementia, the divorce of health and social care is untenable and unsustainable. To achieve comprehensive guidance for professionals in this area, guidelines need stereoscopic or binocular vision that encompasses perspectives from health and social care. Not many guidelines have done this, but NICE is now embarking on a comprehensive programme of social care standards that are likely to take cognizance of Gould's analysis.

The final three papers in this edition, from Franx (pp. 371–378), Gaebel et al. (pp. 379–387) and Tyrer (pp. 388–394), adopt a broader international perspective. Franx describes an impressive programme of guideline development and implementation in the Netherlands between 1999 until 2009. There was wide service user, carer and professional organizational involvement. A total of thirteen guidelines have been produced, along with a range of other products, such as versions for service users, checklists and implementation materials. As part of a recent review of their programme, the Dutch are further developing their methodology focusing on a more transparent process with greater integration of service user and professional perspectives, along with a renewed focus on guideline implementation.

Gaebel and colleagues’ paper builds on previous work undertaken for the World Health Organization by the authors in assessing the quality of a number of national schizophrenia guidelines. Previously, Gaebel assessed different national guidelines on schizophrenia using the AGREE instrument to review the quality and relevance of some 25 clinical guidelines. They have now repeated this study for five of the same guidelines, all of which have been updated since the original study published in 2005. All are judged to be of good quality, with the NICE guideline best overall. They identify a considerable number of similarities across all guidelines (the use of lower dose antipsychotic medication for maintenance, recommendations for the use of CBT) but they also identify some important differences (three of the guidelines recommend second generation antipsychotics over first generation drugs but two do not). The area of greatest discrepancy between the different guidelines reviewed centred on the extent of service user involvement in guideline development. Given the many similarities between guidelines they argue that a trans-national approach to guideline development should be adopted with further work required to determine a robust process of adoption of guidelines to local circumstances.

In a comparative review of two national guidelines on the treatment of borderline personality disorder, Tyrer & Silk (pp. 388–394) examine those developed by NICE (2009) and by the American Psychiatric Association (APA, 2001), and raise a number of interesting methodological issues. The two guidelines were largely in agreement in their recommendations for the use of psychological treatment, but differed considerably in their advice on drug treatment, with the US guideline being much more pharmacologically liberal, suggesting that there is a place for selective serotonin reuptake inhibitors (SSRIs), mood stabilizers and antipsychotic drugs as treatments for borderline personality disorder. Tyrer & Silk speculate why these differences may have arisen. They suggest that the US guideline's conceptual separation of people with borderline personality disorder into those with ‘affective dysregulation’, ‘impulsive behaviour dyscontrol’ and ‘cognitive-perceptual’ symptoms, which the NICE guideline development group regarded as a post hoc classification and not supported by any other evidence, and a lower threshold of evidence for making recommendations in the case of the US guideline, could account for these differences. This points to the need to develop trans-national methods for guideline development.

The future

Clinical guidelines have traditionally synthesized and summarized the ever growing volume of evidence faced by professionals about the healthcare they wish to provide. In part aspirational, in part practical, the guideline has always distilled the hopes of committed practitioners to achieve better outcomes for service users. Guidelines are now beginning to suggest new solutions for ever more complex problems in the fields of mental health and social care, highlighted by the endlessly growing body of knowledge generated and by the practices we continuously revise and shape. More challenging than ever before, the fields of health and social care have become infused with a broad range of influences, bias and differing perspectives that now can only be united through a perpetual focus on the service user, what works for each one of them, and the experience of care that we, as professionals, can provide.

The quality of guideline production has, around the world, improved immeasurably. In Germany, the Netherlands and the UK, and many others, guidelines are dealing with these challenges admirably but not perfectly. There is still much to do, methodologically and practically. The need to deal with increasing costs of production means that an international collaborative approach to guideline development is needed. And none of this can or should happen without the means of encouraging and enhancing guideline implementation. This volume is a collaboration of service providers, academics and service users. Implicitly, the authors writing in this issue understand that this is not just about the knowledge we generate, analyse and synthesize about mental health and social care, the practices we recommend and the guidelines we develop to underpin this. It is most of all about the people who use mental health services. It is both exciting and heartening that, around the world, service providers and service users are collaborating ever more to realize the hopes and aspirations we all have to make mental health and social care effective, safe and as good an experience as we can make it.