Predicting treatment outcome in depression: so far, so good.

Personalised or precision medicine is nowadays a reality in some fields of medicine; for example, research has revealed many of the molecular lesions that drive cancers, showing that each cancer has its own genomic signature1 and the specific production of recombinant factors for patients with hemophilia depends on the precise diagnosis of the type of hemophilia for each individual patient.2 In clinical psychiatry, however, treatments are not targeted to individual patients on the basis of genetic biomarker or other phenotypic or psychosocial characteristics.3 Recent studies have shown the benefit of genotyped-guided treatment for patients with depression4 or the reduction in costs by using pharmacogenetic testing for CYP2D6 and CYP2C19 in patients with schizophrenia,5 but unfortunately, such studies are still too few. However, there is general consensus that the development of new statistical models that use clinical and demographic data of patients from existing trials may help identify patients who are likely to respond to a particular intervention.