Treatment options for real-life patients.

Our current models of depression and clinical trials data provide inadequate and relatively meaningless clinical information. They do not take into account the multiple phenotypes of depression and often do not include patients with "real-life" clinical depression. Psychiatry is called an evidence-based specialty, and this is supported by the wealth of evidence provided by the antidepressant drug trials. However, if the evidence demonstrated by these trials is not accurate, then psychiatry may fail to provide satisfactory treatment for patients with depression. Trial data commonly show the effect of antidepressant treatment to be indistinguishable from placebo. On first examination it could be said that antidepressant drugs act mainly as placebos, or, even worse, that antidepressant drugs are no better than placebos in effect. However, this view is at odds with the observations of clinicians in everyday practice, and the impact of this perception is worrying, with the pharmaceutical companies, patients and practitioners all being adversely affected. Randomised controlled trials (RCTs) provide limited data about the true effectiveness of an antidepressant. However, such RCTs are required by regulatory authorities for drug approval. Antidepressant effects in "real-life" depression need further investigation. When efficacy data for drug and non-drug treatments for major depression are compared, there are very few differences. This lack of differentiation across treatments allows every therapy to be perceived as efficacious, but also non-specific. This leads to patients being fitted to their therapist's preferred treatment, and not vice versa. This "all roads lead to rome" model is contrary to the rest of medicine, where differential treatment effects are to be expected. Why, therefore, is there confusion? A dimensional model for depression homogenises the multiple underlying subtypes of depression. This leads to treatments being tested as having universal application instead of targeting the specific depressive subtypes. This largely underpins the lack of specificity in RCT evidence. These trials involve patients who bear little resemblance to those who clinicians see in everyday practice. These trials also select and favour natural and rapid responders. Therefore the failure to differentiate between drug and placebo is unsurprising in the RCTs. Our spectrum model seeks to identify clinically meaningful expressions of depression, allowing drugs to be targeted to separate depressive conditions and their underlying cause. This allows a rational model for prevention and long-term management. For example, when treating depression associated with anxiety, selective serotonin reuptake inhibitors (SSRIs) produce a high response in patients with internalised anxious worrying or externalised irritability. Not only do they treat the depression but also the fundamental cause. In summary, efficacy data will continue to provide little meaningful clinical information while treatments are tested as "universal" in reference to non-specific conditions such as "major depression". Through use of the spectrum model, therapy can be better fitted to depression subtype, through identification of clinical phenotypes and their causes.