Big data may be able to help clinicians make connections between symptoms and outcomes or between drug combinations and side effects, the New York Times Magazine's Veronique Greenwood writes.
For instance, Columbia University's Nicholas Tatonetti sifted through the US Food and Drug Administrations' adverse drug reaction records to uncover a link between taking both paroxetine, an antidepressant, and pravastatin, a cholesterol-lowering drug, and high blood glucose levels. Individually the drugs didn't affect glucose levels, but together they did.
But, the Times points out that making connections like these in patient data brings up privacy and consent concerns — is such data mining part of medical treatment or is it medical research? Additionally, medical records include data from people who sought treatment and there is a lack of controls for such studies.
Still, Greenwoods says that "[w]ith time, and with some crucial refinements, this kind of medicine may eventually become mainstream."