Recognising the language of evidence

Welcome to the bright new year of 2009. It is indeed bright midsummer in Australia, but for us in the Northern hemisphere we need something to cheer us through the gloom of winter. When in darkness we have time to think, and there is plenty to think about in this issue. Much of the thinking is related to evidence-based psychiatry, a set of empirical data that we now take notice of more than any other guide, even if we do it passively such as following the guidelines of the National Institute for Health and Clinical Excellence, the American Psychiatric Association or other official bodies. In interpreting the evidence we classically rely on the randomised controlled trial, often called the gold standard. In fact, it is more like the original penny black, the first official postage stamp in 1840, and since the inventor of the controlled trial, Austin Bradford Hill, was a direct descendent of Rowland Hill, the instigator of universal postage, the penny black might be a better metaphor. The randomised controlled trial, like the penny black, is recognised for its universal properties, its acceptability in many different places, and its freedom from bias or favour. But there are now many more stamps and postage labels, and, just as we no longer bow in obeisance to the penny black, we should not worship the randomised trial. Gordon Parker (pp. ) certainly takes it down a peg or two in discussing the implications of the highly publicised findings of Kirsch and his colleagues ^{Reference Kirsch, Deacon, Huedo-Medina, Scoboria, Moore and Johnson1} suggesting that antidepressants are only marginally effective in depression, but like the brandy of the same name, they should only be taken in small quantities for the best effect. Randomised trials, Bradford Hill reminded us, only answer ‘precisely framed questions’ and if the questions are wrong or not well framed, the answers are useless. Antidepressants are of little value in the wrong populations, and we are only just beginning to find out which are the right ones; ^{Reference Fournier, DeRubeis, Shelton, Gallop, Amsterdam and Hollon2} and when Parker claims our diagnostic systems are wrong he is almost certainly right, and we might even do better with the Parker system of classification by chocolate. ^{Reference Parker and Crawford3} We would certainly not want you to abandon evidence-based psychiatry, but to realise that although some evidence is now highly standardised (e.g. Geddes et al, pp. ; Young et al, pp. ), there are often other reasons why good evidence is either being ignored by practitioners, ^{Reference Young and Hammond4} not replicated reliably (Salib & Cortina-Borja, pp. ), or not maintained (Adeponle et al, pp. ), and this helps to refine the best research questions for subsequent trials. We must also not decry the place of good independent diagnostic tests (O'Brien et al, pp. ) as excellent in separating Parker's apples and oranges from Kirsch's cherries, and recognise the value of good qualitative studies (Bisson et al, pp. ), including the views of carers and patients ^{Reference Lawrence, Murray, Samsi and Banerjee5,Reference Priebe, McCabe, Bullenkamp, Hansson, Lauber and Martinez-Leal6} in ensuring the right questions are asked and the applications of evidence followed through.

So we all have to recognise the full language of evidence to interpret it properly. I have recently come back from a conference in Holland, where the background Dutch was politely replaced by familiar English whenever I entered into earshot. So I understood a quarter of what was going on, but my polylingual Dutch colleagues understood it all. Native English speakers tend to be just a little arrogant about their language, assuming that nothing else is worth knowing until translated into English; similarly, it is not enough to say the words ‘randomised controlled trial’ in the context of evidence-based psychiatry and expect everything else to fall neatly into place.

Cinderella's cat

This is the season of pantomimes and one of the hardy favourites, if not the absolute winner, is Cinderella. I have always been intrigued by psychiatry being described, often by its practitioners, as a Cinderella specialty in medicine. This metaphor, if extended properly, means that in time Miss Mental Heath will indeed go blushing to the medical ball and be embraced as an equal by a handsome president of one of the original medical Royal Colleges, destined then to live forever in bliss with proper doctors in attendance, fulsomely apologising for how badly they have behaved towards her and us over the years. This may or may not be true, but what of that other Cinderella, the professionals who work with those with intellectual disability, often described as the ultimate Cinderella when compared with all the psychiatric disciplines. They are like Cinderella's cat, also exploited and reduced to poking in the cinders, and with no fairy godmother to transform them. But the feline advance is also beginning, and I am pleased to announce that the highest cited paper in the Journal in 2007, the cat's whiskers, derives from intellectual disability research, written by Sally-Ann Cooper and her colleagues from Glasgow. ^{Reference Cooper, Smiley, Morrison, Williamson and Allan7} So come into the coach, Sally-Ann, and remember it is not a converted pumpkin, but a leather-liveried limousine (not to be scratched now), that is going to take you to the ball.