Practising evidence-based mental health By John Geddes, Andre Tomlin and Jonathan Price Edited by Sharon E Straus Radcliffe Medical Press 264 pages £30

Medicine is in the throes of unpicking its knowledge base. At the same time, practitioners are being bombarded with an unprecedented amount of information in print and on the Internet.

This is a challenge for all specialties, particularly psychiatry, which is infamous for its different, even opposing, valuations of types of knowledge. Will an evidence based approach find a neutral and effective way through?

Practising Evidence-Based Mental Health sets out to do two things: first, to provide a seven-session course for clinicians with an interest in psychiatry on how to practise evidence-based medicine; second, to help membership candidates prepare for the critical review paper in the Royal College of Psychiatrists' examination.

It is obviously a workbook for a particular course, which could be very valuable. But I imagined myself attempting to put on a similar course armed only with this material, and it is clear that unless you are quite practised at the techniques this would be difficult.

On the one hand some of the subjects are quite complex - such as 'confidence intervals' and 'numbers needed to treat'- and on the other, much of the advice about where to look on the Internet for answers - PubMed, the Cochrane Collaboration and the Evidence Based Mental Health journal website itself - is obvious. The topics chosen for particular examination are interesting and include preventing recurrent homelessness among mentally ill men, screening for eating disorders in a student population, screening for dementia, prognosis in schizophrenia, the use of St John's wort for depression and medication for dysthymia.

I sought the views of one trainee psychiatrist, who said he found the book difficult to read and not clearly relevant to the exam. To get a good grasp of the topic one needed an interactive course, he said.

I'm not criticising the message of evidence-based mental health, but it is not accessible through this book. The field, however, is moving ahead all the time, and I am looking forward to easy access to websites with names such as 'Missing answers: topics for researchers' and 'Best evidence for the mental health service user', not to mention interactive courses on 'Maximising your power as an evidence-based placebo'.