Abstract
Guidelines are readily available for the treatment of depression, and more recent ones are explicitly evidence-based. Their core messages vary little but they tend to minimise uncertainties and gloss over difficult areas. This article examines three areas of uncertainty: the thresholds of severity and, for milder depression, the duration of illness for which antidepressants are more effective than placebo; the next step in drug treatment when a patient has failed to respond adequately to a first antidepressant; and how long continuing on antidepressants should be recommended in relation to individual patients' needs. It is concluded that the uncertainties in relation to treating individual patients are a combination of lack of evidence and individual patient factors but there is also an intrinsic uncertainty that will continue to require good clinical judgement.