The prognosis for patients with major depression is bleak. Not all patients respond to treatment and relapse rates are high in those who do.¹ The challenge in general practice is to identify those at greatest risk of relapse and to reduce that risk.

This review is based on a literature search of studies published between 1960 and 2006. Inclusion criteria were not defined, and it is possible that inclusion bias and publication bias have led the authors to overestimate the prevalence of residual symptoms. The review does not indicate whether the studies included refer to patients in primary or secondary care settings.

The authors conclude that 'most patients report residual symptoms despite apparently successful treatment'. Residual symptoms - anxiety and irritability are the most common - are associated with a poor prognosis. Relapse is also more likely in patients with residual impairment of social functioning (for example interpersonal friction) and patients with persistent dysfunctional attitudes and attributions.

The authors suggest that the traditional concept of recovery, the maintenance of a score below the cut-off point on a rating scale for a specified period of time, is inadequate as it does not take into account residual symptoms, ongoing need for drug therapy and the risk of relapse. To be deemed to have recovered, patients should no longer require antidepressant treatment, have no residual symptoms that impair functioning and should satisfy a test of psychological wellbeing.

The review cites a number of studies that provide evidence that cognitive behaviour therapy (CBT) following successful pharmacotherapy reduces the risk of relapse. Given the paucity of CBT therapists, our patients are unlikely to receive CBT if we do not provide it ourselves.

Currently I offer selected patients a single hour-long appointment during the recovery phase of their illness. Sadly, I do not think that it is realistic to expect other GPs to commit even this limited amount of time to their depressed patients, and basic training for GPs in brief CBT does not appear to improve patient outcomes.²