Scientists from Microsoft, Stanford and Columbia Universities have discovered information on unreported side effects due to prescription drug interactions. Even the U.S. Food and Drug Administration’s (FDA’s) Adverse Event Reporting System did not indicate drug interactions between an antidepressant, paroxetine and a cholesterol-reducing drug, pravastatin. The data indicated a combination of the two drugs, according to a recent New York Times article, causes an increase in blood sugar, hyperglycemia, in some patients.
The FDA requests physicians report adverse drug reactions they are aware their patients have experienced using the Adverse Event Reporting System. Because the data compiled are reported by physicians—and more often the adverse reactions are noted by patients who don’t always inform their physicians—the Adverse Event Reporting System is not completely comprehensive. Instead, scientists used a data-mining technique to review FDA reports on “drug-drug” interactions which contained “anonymized” data gleaned from a software toolbar used by people who allowed their search histories to be collected.
The data consisted of 82 million “individual searches for drug, symptom and condition information,” according to Markoff. (NYTimes, 3/7/13) The scientists determined that people who had searched on side-effects from both drugs were more likely to have included “hyperglycemia” in their searches than not.
The scientists noted that their manner of data mining “would be a valuable tool” for the FDA to use to track adverse drug effects. Presently, FDA has at least one on-going project to evaluate the risks of drugs currently on the market and eventually plans to monitor U.S. drug use on a larger scale.