Precision Psychiatry: The Future Is Now

Randomized clinical trials (RCTs) and meta-analyses allowed us to make broad generalizations about specific populations and specific treatments. For individuals that fit the inclusion and exclusion criteria of original studies, this may not be a problem. However, this approach fails to detect the granularity related to single individuals. In this same vein, significant results may not represent a real benefit for some individuals. Indeed, subjects included in clinical trials do not consistently reflect patients from real-world clinical scenarios. Of note, the very idiosyncrasies that characterize most patients, such as multi-morbidity profiles, are often considered exclusion criteria in most clinical trials. Another elusive goal in modern psychiatry is the prediction of the propensity for developing mental disorders and potentially preventable poor outcomes. It’s important to consider whether the very way we currently think about causality in psychiatry is preventing us from achieving more accurate predictions. The linear association between risk factors and clinical outcomes is important to understand the course of chronic disorders. However, linear patterns do not accurately stratify what patient will have a specific disease or, if a patient already has it, what will be his or her prognosis. In this article, we describe how big data, machine learning techniques, sensors, and other devices started to play a role in unraveling the above-mentioned clinical dilemmas. These new methods will enable the creation of the new field of precision psychiatry and will allow patients to have an active role in their own care.