Scientists Say Mining EHR Data Could Boost Medical Research

Although the primary goals of electronic health record systems are to improve efficiency and reduce costs, the technology also provides an opportunity to mine data for medical research, the New York Times reports.

About Mining EHR Data

Mining data from EHR systems could help researchers evaluate the outcomes and side effects of specific treatments.

Russ Altman — a physician and professor of bioengineering and genetics at Stanford University — said that medical discoveries always have been based on “hunches,” but mining data from EHR systems would make it possible “to see if a hunch has statistical merit.”

Some scientists say that large-scale monitoring and analysis of EHR data could make all patients participants in a vast, ongoing clinical trial.

According to Nicholas Tatonetti — assistant professor of biomedical informatics at Columbia University — mining EHR data is faster and less expensive than setting up traditional clinical trials.

Challenges to Mining EHR Data

Despite its potential benefits, there are several challenges that are hindering the mining of EHR data.

According to the Times, the biggest challenge to mining EHR data relates to privacy concerns. Each EHR used for research must be stripped of all information that potentially could be used to identify the patient.

However, it is important for researchers to know when they are examining EHRs that are from the same patient but stored in different EHR systems.

According to Tatonetti, another challenge impeding the mining of EHR data is the fact that information in an EHR might be incorrect and diagnostic codes might be unreliable because they are used for billing purposes.

In addition, Tatonetti noted that physicians generally refrain from ordering follow-up tests that have no clinical usefulness, even though such tests potentially could contribute valuable information to data mining-based research efforts (Jaret, New York Times, 1/14).