The potential of using large data sets to come to possible treatment-related conclusions not otherwise evident is demonstrated by results of a recent study by researchers at Vanderbilt University Medical Center.
Specifically, diabetic patients with cancer taking metformin, a first-line treatment against type 2 diabetes, tended to do better in recovering from cancer than diabetic patients with cancer who weren't taking metformin, researchers found by analyzing electronic health records. Much better, in fact. Their data analysis showed metformin use was associated with a 22% decrease in overall mortality, according to findings recently published in the Journal of the American Informatics Association.
The hypothesis for drug companies is to investigate whether metformin has general anti-cancer properties or whether something else is at play.
The study combed through 15 years of EHR records from Vanderbilt, looking at approximately 14,000 cancer patients with diabetes. Researchers then re-checked their findings with records from the Mayo Clinic over a similar timeframe and found similar results.
It's this sort of finding that health big-data advocates have been hoping for. The belief is that as healthcare data becomes more readily accessible, and as computing power to manage that data becomes available, the healthcare system should be repeating these types of findings far more frequently.
Follow Darius Tahir on Twitter: @dariustahir