The irresistible rise of the selective serotonin reuptake inhibitor is a fact of contemporary life; antidepressant prescription volumes and costs have more than trebled since their introduction in the early 1990s.

GPs’ views are important because we do most of the prescribing of antidepressants. We also have some influence on whom is referred for psychotherapy; the government has made a commitment to increasing availability of psychological interventions in primary care, but psychotherapy is still a carefully rationed resource. There is something slightly disconcerting about reading Macdonaldetal’s summary of their interviews with 63 GPs. Many of the explanations offered by GPs for the increase in prescribing are thoughtful and insightful. The usual suspects are lined up: The Defeat Depression Campaign, the pharmaceutical industry, the consumerist attitude of the patient with high expectations of happiness, and a pill for social ills. The causes of the increase are largely seen to be external; as the authors comment ‘GPs did not see themselves as drivers of change’. There is some soul searching on the medicalisation of misery, but the prevailing view seems to be that this is an almost inevitable consequence of our limited resources and the pressures we face to solve social problems. Stavrouetal found their inner London GPs to be altogether more confident in deciding whom to refer and not to refer for psychotherapy. Fourteen GPs discussed pairs of patients: one referred and one not referred. The plausible picture that emerges is largely one of rational decision makers, acting in concert with their patients’ wishes. As the authors point out, this may reflect a bias, in that the GPs selected the patients they wished to discuss. Areas of potential uncertainty or difficulty do not appear to have been explored by the participants. For example, ‘non-psychologically minded’ patients with ‘multiple minor physical health complaints’ who were ‘not necessarily improving with the GP’s treatment’ were still considered unsuitable for psychological therapy; whatever happened to the reattribution of minor somatic symptoms? These papers address two important and not unrelated questions: how did we come to be prescribing so many antidepressants; and how do we decide who will benefit from referral to psychotherapy? If we want to answer these questions we need to use other methodologies to look at patient outcomes. The missing piece of the puzzle in any discussion of GPs’ selection of patients for referral to psychotherapy is evidence for predictors of response to treatment. This is currently lacking. And even more surprising is the fact that we bewail the increase in antidepressant prescribing without knowing very much about its impact.