Using EHRs to Find Hidden Drug Benefits

Researchers at Vanderbilt University Medical Center have used electronic health record analysis to uncover hidden drug benefits, according to a study appearing in the Journal of the American Medical Informatics Association.

Researchers at Vanderbilt University Medical Center have used electronic health record (EHR) analysis to uncover hidden drug benefits, according to a study appearing in the Journal of the American Medical Informatics Association.

Scans of healthcare billing data and computer simulations can be used to find drugs that are good candidates for repurposing trials. Doing so could mean putting additional indications on the label of already established drugs, which could save pharmaceutical companies big money in research and development of new drugs.

The Vanderbilt study, led by Josh Denny, MD, MS, associate professor of biomedical informatics and medicine, and Hua Xu, PhD, MPhil, MS, adjust associate professor of biomedical informatics, in particular studied diabetes drug metformin for uses in some cancers.

“Our EHR allowed us to delve into details of treatment and response—cancer staging, control of cancer, the various timelines involved and cancer subtypes,” Denny said in a statement.

The researchers scanned records of 32,000 cancer patients seen at Vanderbilt since the mid-1990s and focused on the 5-year survival rate with and without exposure to metformin. The Vanderbilt team replicated these findings in 79,000 cancer patients seen at the Mayo Clinic.

Use of the drug as a therapy for type 2 diabetes decreased all-cause mortality by 23% compared with metformin-free non-diabetic patients. Compared with diabetic patients receiving insulin, metformin decreased mortality by 39%.

Overall, the cancers that showed a decrease in mortality with the use of metformin included breast, colorectal, lung, and prostate cancer.

The goal is to adapt this strategy to larger, more geographically dispersed cohorts as EHR networks take shape.

“We’re now building on this study, pursuing opportunities for using our EHR to look at all drug exposures across a given disease—starting with cancer,” Denny said. “We’re trying to find other signals that may look like metformin in terms of affecting patient outcomes.”