Recovery, psychosis and psychiatry: research is better than rhetoric

Recovery has been defined as 'a personal process of overcoming the negative impact of diagnosed mental illness/distress despite its continued presence' (1). The approach has received many endorsements. In the UK, policy states that 'Services of the future will talk as much about recovery as they do about symptoms and illness' (2), and professional groups such as psychiatrists (3) and nurses (4) have embraced the rhetoric. Internationally, the approach has been formally adopted in New Zealand (5), the Republic of Ireland (6) and USA (7), among others. Guidance is being disseminated (1) and editorials written (8). Recovery is in vogue. What are the defining characteristics of a recovery-focussed mental health service? Recovery is understood to mean something different to either sustained remission or cure – it is a way of '…living a satisfying, hopeful, and contributing life even with the limitations caused by illness'. (9). Goals are user-defined and therefore individual and sometimes ethically challenging for staff to work with. Staff have an optimistic and hope-inducing view about the ability of people to find meaning in their experiences – to generate a story or narrative about themselves which leads to a future beyond (though possibly including) mental illness. Language is different, because staff recognise the power of language to shape belief, and do not want to limit change by imposing an explanatory illness model when other models may be more helpful to the individual. The policy and practice implications of recovery-focussed service have been explored elsewhere (10-13). At its simplest, this will mean clinicians basing decisions less on professionally-defined goals and more on listening to and acting on the service user's wishes. If recovery-focussed mental health services differ from traditional services, what has been the response to this call to change? Several responses can be identified. Some researchers have insisted on symptomatological and functional improvement (rather than patient-defined criteria) as the sole indicator of recovery (14, 15). Others commentators express the view that the term 'recovery' is vacuous –'a redefinition of the term 'recovery' in order to give hope is to build hope on illusion' (16). Services have appropriated the term without meaningfully changing their function, e.g. re-labelling rehabilitation services as 'Recovery and Rehabilitation Services'. These responses are all consistent with an aim of maintaining the status quo. They may be the most appropriate responses, but the paucity of evidence means that it is currently impossible to have an informed debate. Empirical research is needed. We therefore propose a research programme to identify the costs and benefits of developing recovery-focussed mental health services. Recovery research needs to combine methodological rigour with not disadvantaging recovery-focussed approaches by evaluating them using methods and outcome criteria which are biased towards traditional working practices. We identify three strands to this research programme: identification and prevalence; outcome measures and evaluation methodologies; and interventions and cost-effectiveness. First, the active ingredients of recovery-focussed mental health services need to be established. As well as working practices, the ingredients will include staff attitudes, values and beliefs. This represents a change from the current modernist approach to describing services (i.e. they are fully characterised by what they do), in recognising that mental health services involve people, both using them and working in them, so how staff work also matters (17). Once identified, the active ingredients will inform the development of fidelity scales, which assess adherence by the service to the active ingredients required for a mental health service to be recovery-focussed. This approach to ensuring treatment fidelity is a standard approach to identifying whether a service model is actually implemented (18), a necessary element of evaluation (19). These fidelity scales can then be used to establish national baseline prevalence estimates of the extent to which services are recovery-focussed. Second, there is a need to develop recovery outcome measures. Some measures have already been developed, mainly in USA (20, 21) and Australia (22). Cultural validity will be an important psychometric property to establish when they are used in other countries. This will allow methodological developments which incorporate utility (i.e. individual preferences) into evaluations. For example, instead of using the same predefined primary outcome measure for all participants in a clinical trial, each participant could choose their own primary outcome measure from a selection of standardised outcome measures. Similarly, recovery-focussed evaluations are likely to include both qualitative and quantitative methodologies (23), and the use of established user-led methods such as User-Focussed Monitoring (24) or Strategies for Living (25). The development of new approaches to evaluation will be particularly important when evaluating existing interventions involving coercion and compulsion. For example, they will allow the empirical investigation of whether assertive outreach services promote recovery, rather than [as currently investigated (26, 27)] whether they reduce symptoms and hospitalisation rates. Finally, there is a need to develop and evaluate interventions to promote recovery-focussed working. For example, increased visibility of recovered service users in teams may combat the clinician's illusion that no-one gets better (28). Training programmes may change the language used, which in turn shapes beliefs and actions (29). This will allow the development of demonstration sites, with evaluation using innovative methodologies which incorporate both traditional and recovery outcome measures. Two key questions for these demonstration sites will be whether recovery-focussed services actually promote recovery, and whether recovery-focussed services also lead to improvement in professionally-defined domains of outcome. These questions are important, because high-quality decision-making by service users, clinicians and commissioners about interventions and service models is only possible when the costs as well as the benefits are known. For example, if all that is known is that an intervention reduces symptoms, it can be recommended. If it is known that an intervention reduces symptoms and fosters dependency and loss of hope, a more informed debate is possible. There are three possible futures for recovery in psychosis. First, the term 'recovery' and associated values and working practices fail to impact on services, and slowly disappear. Secondly, recovery values become embedded in clinical practice, but the term becomes slowly forgotten – similar to the 'social role valorisation' movement in learning disability services. Thirdly, recovery becomes the dominant model and leads to fundamental scientifically well-documented changes in mental health practice, on a scale similar to the impact of re-labelling dementia as Alzheimer's Disease. What is best for people experiencing psychosis should decide the future.