© 2004 The Royal College of Psychiatrists
© 2004 The Royal College of Psychiatrists
Editorial |
Psychiatry in the future
The future of mental health services research
Head of Psychological Medicine, Department of Psychological Medicine, Imperial College, London W6 8RP
It might be thought now that mental health services research is a thriving discipline. Researchers are in on the ground floor of every new intervention and should be well placed to determine its value through the exacting requirements of evidence-based medicine. Once their work is done, their conclusions are fed back to the planners, who adjust mental health policy as a consequence of the new evidence they have received.
Would it be that this logical new dawn is a blissful reality? No, it is not. There are three problems requiring solutions.
Improving methodology
Although ‘evidence-based practice’ is now a mantra repeated incessantly by all in the health service, it is rarely understood at the level its originators intended. For many, evidence is literally the collection of data. Although this has undoubted value, the interpretation of these data is often inadequate. Repeatedly we all come across studies in research and development settings that, sadly, will never see the light of published day, but which are honestly presented as studies of efficacy. These include the mere presentation of numbers of presentations to a service, pre–post comparisons following the introduction of something new (the positive effects of novelty are almost always ignored) and audit data with 20–30% completion rates. This would be less of a problem if the top end of research output was vastly better, but it is not. There is abundant evidence that control groups in service trials are badly selected (Burns & Priebe, 1996), that many research outcomes are poorly chosen or of little perceived value in ordinary practice (Priebe et al, 1995; Gilbody et al, 2002) and that the results of studies in one continent cannot be transposed directly to another (Fiander et al, 2003). We cannot argue that health service evaluation is an essential element of activity until we have identified exactly what we want to measure and how. At present, if we just record whether people are satisfied with what is delivered we may have as valid a measure as any other (Shipley et al, 2000), and at least this is easy to record.
Local is limiting; general is sublime
There is a general tendency in health services research to examine a subject locally and then to extrapolate results to a larger stage. Although this might be appropriate for a new drug or a simple intervention, it is not suitable for a complex intervention in which several factors interact (Campbell et al, 2000). The setting in health services research is a neglected variable; it cannot be ignored, and only when a complex intervention is carried out in many centres and the effect of setting is taken into consideration can we have confidence in the results.
The bedrock of health service policy should be evidence
It is a sad reflection of priorities that policy in mental health is based more on political than scientific imperatives. Almost all of the major reforms of mental health in the last few years have followed national scandals rather than research-based evidence, and although some changes, such as deinstitutionalisation following the Ely Hospital inquiry into the abuse of those who had intellectual disability, accorded with the research evidence already demonstrated by Albert Kushlick (Evagorou, 1970), others – such as the setting up of assertive outreach teams – have been introduced in spite of the evidence against their value in the UK (Burns et al, 1999). We must do much more to establish effective links between research and policy to prevent these embarrassing gaffes.
All positive futures depend on a vision that is broad enough to cope with all considerations, ranging from known to barely perceived, and in health services research the motto ‘good practice is evidence-derived practice’ is one worth holding to. However, we need to do more to make the evidence robust, widely disseminated and understood, and ultimately, inspiring.
References
- BURNS, T. & PRIEBE, S. (1996) Mental health care systems and their characteristics: a proposal. Acta Psychiatrica Scandinavica, 94, 381 -385.[Medline]
- BURNS, T., CREED, F., FAHY, T., et al (1999) Intensive versus standard case management for severe psychotic illness: a randomised trial. Lancet, 353, 2185 -2189.[CrossRef][Medline]
- CAMPBELL, M., FITZPATRICK, R., HAINES, A., et al
(2000) A framework for the design and evaluation of complex
interventions to improve health. BMJ,
321, 694
-696.
[Free Full Text] - EVAGOROU, D. (1970) Kushlick: the wizard of Wessex. Nursing Times, 66, 1530 -1532.[Medline]
- FIANDER, M., BURNS, T., McHUGO, G. J., et al
(2003) Assertive community treatment across the Atlantic:
comparison of model fidelity in the UK and USA. British Journal of
Psychiatry, 182, 248
-254.
[Abstract/Free Full Text] - GILBODY, S. M., HOUSE, A. O. & SHELDON, T. A.
(2002) Outcomes research in mental health: systematic review.
British Journal of Psychiatry,
181, 8-16.
[Abstract/Free Full Text] - PRIEBE, S., GRUYTERS, T., HEINZE, M., et al (1995) Subjective criteria for evaluation of psychiatric care – methods for assessment in research and routine care. Psychiatrische Praxis, 22, 140 -144.[Medline]
- SHIPLEY, K., HILBORN, B., HANSELL, A., et al (2000) Patient satisfaction: a valid measure of quality of care in a psychiatric service. Acta Psychiatrica Scandinavica, 101, 330 -333.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||