Commentary A European perspective on the Canadian guidelines for bipolar disorder Within the past 10 years, a number of guidelines and expert consensus have been developed to increase awareness of the misdiagnosis of bipolar disorder and the growing number of potential treatments, mainly pharmacological, but also, more recently, psychological (1–10). The guideline of the Canadian Network for Mood and Anxiety Disorders (11) represents the most recent such document. It is a comprehensive text of what we currently know and how that may have an impact on clinical decision making. Not only Canadian psychiatrists, but also anybody involved in the care of bipolar patients across the world, including patients themselves and their relatives, will be happy to see such a balanced, and evidence-based document come to press. Methodology of the Canadian guideline All guidelines have similar objectives, but they often reach different conclusions. This will usually be due to the methodology employed to appraise the evidence. The approach used by the Canadian group is well described. The process included rating the strength of evidence, the inclusion of issues not only of efficacy, but also tolerability, clinical experience, sample case reports, and the expertise of the professionals involved in this project. The eclectic approach reflects the variety of evidence that bears on clinical decision making. It also includes sections on difficult issues such as pregnancy, and the treatment of bipolar II sub- types. The guideline is relatively specific and moves to a finer grained level of recommendations than some that have preceded it. This is simultaneously an apparent strength and a potential weakness. Clearly we would like to have a complete and believable summary of preferred actions for all clinical circumstances, even individual cases. Tra- ditionally, experts are supposed to know what to do. The weakness in any guideline is where to draw the line between statements that are based on a consensus derived from convincing evidence and those that are really statements of opinion, how- ever widely the opinion may be shared. To be more concrete means usually to move away from the most reliable evidence, because reliable evidence is usually a statement of average effects in average patient populations. To depend more on current opinion carries the further risk of becoming outdated before other similar documents, as more evidence about the management of bipolar disor- der accumulates. In a sense, there can be no objection to very detailed appraisals, if the operational rules are explicit. However, they also need to be valid. We have a specific concern that the approach to levels of evidence has encouraged an over-gener- ous ranking of some treatment options. From our point of view, level 2 is too inclusive. It gives the same rank to a medicine with one positive adequately powered placebo-controlled trial (12) plus a positive randomized trial against an active comparator, as to a medicine with a single underpowered, ÔequalÕ efficacy trial against any Vieta E, Nolen WA, Grunze H, Licht RW, Goodwin G. A European perspective on the Canadian guidelines for bipolar disorder. Bipolar Disord 2005: 7 (Suppl. 3): 73–76. ª Blackwell Munksgaard, 2005 Eduard Vieta a , Willem A Nolen b , Heinz Grunze c , Rasmus W Licht d and Guy Goodwin e a Bipolar Disorders Program, Hospital Clı ´nic, University of Barcelona, IDIBAPS, Barcelona, Spain, b Department of Psychiatry, University of Groningen Medical Centre, Groningen, The Netherlands, c Department of Psychiatry, Ludwig- Maximilians-University, Munich, Germany, d Mood Disorders Research Unit, Aarhus University Psychiatric Hospital, Risskov, Denmark, e University Department, Warneford Hospital, Oxford, UK Bipolar Disorders 2005: 7 (Suppl. 3): 73–76 Copyright ª Blackwell Munksgaard 2005 BIPOLAR DISORDERS 73