From the Editor's desk

Conspiracy, causality and contexts

A satirical piece ¹ in the New Yorker by Andy Borowitz reports that the Earth is endangered by a new strain of fact-resistant humans, whose defences outmanoeuvre a mountain of facts and scientific evidence. The BJPsych aims to provide the best evidence to the widest possible audience, and to certainly attack the immunity of fact-resistant humans through new strains of publishing journals, books, and offering a plethora of continuing professional development and learning resources. New research and facts were abundant at the Royal College of Psychiatrists' annual International Congress, and I hope will contribute to combating the defences found in those fact-resistant humans. The themes included new research on genomics and schizophrenia, translational research that proposes new pharmacological and psychological interventions for suicide in psychosis, exciting and hopeful research in preventing dementia making use of animal models, risk assessment and safe practice, prevention of terrorism and violence, and critical commentaries from arts and the medical humanities on the place and representations of psychiatry in the public imagination. The keynotes were noteworthy for political, media and scientific narratives of the importance and future of psychiatry and mental healthcare. The newly appointed Minister for Community and Social Care at the Department of Health, Alistair Burt, made a spirited announcement of intentions to increase spend on children's services and urged zero-tolerance approaches to suicide. The Chief Executive of NHS England, Simon Stevens, encouraged professional confidence and stability but no new public funding. Change is expected in order to improve effective treatments without any new monies. What is the evidence base for driving such reform, and how can reform be justified in the absence of evidence?

Most health providers and clinicians are fully aware that evidence is rarely necessary for powerful new policy directions, ² and that there is no room for complacency around the production of better evidence. ³ An alternative perspective is that scientists are not well placed to produce policy, or policy-relevant research, and actually a new type of science is needed that disrupts impediments to translational research and to the implementation of research findings in a more timely way. ⁴ Enter the new science of strategic studies to tackle different strategies for implementing research, including better communication and two-way dialogue, a process that I hope the BJPsych can strengthen. We need to better understand the process by which researchers, policy makers, practitioners and managers come to believe the unbelievable with no evidence, and at the same time, how they do not believe or take note of good evidence.

An excellent keynote by David Aaronovitch provided a first-hand account of an ICU psychosis following keyhole surgery; the account is well described in the media (www.thetimes.co.uk/tto/magazine/article3220179.ece) but Aaronovitch's presentation sought answers about how any human being, with a psychosis or not, could believe in the unbelievable and construct conspiracy theories in the face of evidence to the contrary. ⁵ Conspiracy theories can cause conflict, war and perhaps drive political violence and they can harm public health if effective interventions are not taken up or the evidence ignored. Several conference themes illustrate this. For example, the place of addiction services in local government rather than in health agencies, and the place of zero tolerance in suicide prevention were raised as controversial but relevant questions. In the UK, the relocation of addiction services and funding to public health and local government are predicated on addiction being not only a medical disorder but having social and criminal justice consequences and causes. This should foster more integrated care, yet the emerging decisions seemed to reflect the notion that addictions are only a criminal justice or social problem. Although integrated care is needed, the total move to local authority commissioning risks a failure of integration because of a whole recalibration of priority, and what resources will be allocated to those with severe mental illness and complex social, criminal justice and health problems. Such judgements rely on a good grasp of the evolving evidence, and better evidence to inform new configurations of care. Delivering a new medication would not be lightly undertaken, but delivering a whole new service configuration that risks failing some patients, perhaps even resulting in premature death, is adopted as if addictions are entirely social phenomena and can be safely configured in an un-evidenced policy and political environment. I have previously referred to the remarkable work of Henry Rollin, and his seminal paper suggesting that psychiatry is such a complex area of practice that it lends itself to extreme empiricists as well as to magic and spells. ⁶ This reflection may help explain the vastly different perspectives taken of mental illnesses when they span from common and familiar experiences found in public health and primary care, to the more severe and disabling and less common illnesses that bring people into contact with specialist mental health, forensic and prison services. Turning away from the complex, the uncertain, the nuanced, is easier if the evidence base, for what works and for whom, is ignored, as if by restructuring services or re-estimating the priority of mental illnesses, we are actually reconstructing them to be what we wish them to be. And the true cost is that the lives of people whose care is removed, minimised or radically altered, are disrupted again for many years to come, when they may only have just established themselves on a comprehensive caring context and nexus of therapeutic relationships.

A further example, is that risk assessment is not predictive of violence or suicide, yet zero tolerance as a phrase is creeping into preventive health policy. A paper in this issue (King et al, pp. ) identified those in community justice pathways to be at risk of suicide, yet individual risk prediction is imperfect and any solution has to involve a number of agencies all working to the same model of suicide prevention and understandings about the role of mental illness. Why does the public believe that all depressive illnesses, or psychoses or suicides are preventable in a way that they would not countenance for heart attacks, diabetes or cancer? Hope and optimism are essential and must be nurtured, and there is no scope for complacency in discovering new treatments and better applying existing ones. Yet, we need to balance this drive to minimise incident disease and chronicity, and find simpler solutions through redefining mental illness, with the evidence to prevent and treat effectively.

Several popular research presentations proffered the importance of Big Data, that is, observational studies of large patient cohorts rather than trial designs including linkages across data-sets and across countries. Clearly, Big Data are appealing, as the observed benefits or harms are demonstrated for real patients, and the large numbers across centres, regions, and countries offer substantial power to ask subtle questions about relatively uncommon risk factors and outcomes. Yet non-trial designs continue to raise significant questions, not only about selection bias and contamination effects, but also of measured and unknown confounders.

Papers in this months' BJPsych include research findings from cohorts in real clinical settings, for example, outcomes from psychotherapy and in a trial design and meta-analyses, outcomes may not be dependent on the duration of psychotherapy if the duration is itself selected to suit patients' need by experienced clinicians when real-world observational data are inspected (see Stiles et al, pp. )! Interventions tested in trials seem to rely on specific context and unnoticed packaging around the intervention; for example, see Bee et al (pp. ) on how service user involvement must not be judged or driven through a service definition and actions, but through patients; and Bhui et al (pp. ) show that empowerment and ethnographic engagement are important antecedents to effective therapeutic relationships to improve existing interventions. Practitioners are becoming more expert and knowledgeable about intervention components for individual patients as well as phenotypes of psychopathology that we are targeting (e.g. Valmmagia et al (pp. ), Green et al (pp. ), Bölte et al (pp. ), Verdejo-Garcia et al (pp. )). Some studies show an underlying mechanism for clinical observations, permitting more refined targeting and better understanding of coping strategies. For example, Reitz et al (pp. ) confirm that people with borderline illnesses self-cut to calm and regain control; and Fischer-Kern et al (pp. ) show that transference-oriented interpretations correlate with improvements in reflective function, but could reflective functioning have enabled or facilitated better use of and recording of transference interpretations?

We are becoming more familiar with understanding complex packages of care, for example the impact on carers of our community treatment services (see Lavis et al, pp. ), but we are still in our infancy regarding the influence of contexts and cultures of care (buildings, deprivation, political and social attitudes including stigma, and financial). And a new context for us to grapple with is recession driving a new priority for mental illnesses that risks the enactment of stigma from the past entering our present experience. Ah, is that a conspiracy theory of context or simply the repetition compulsion at work?