Electronic health record data can be harnessed to identify and hypothesize the survival rates of patients with cancer mutations, a major step forward in "precision oncology" according to a recent study published online in the Journal of American Medical Informatics Association (JAMIA).
According to the researchers, "the future of cost-effective and high quality cancer care depends on rapid learning systems that optimize the utility of routine observational data gathered from the clinic, including outcomes." They developed a "proof-of-principle" tool, called the Continuously Updating System for Tracking Outcome by Mutation, to Support Evidence Based Querying (CUSTOM-SEQ) to automatically calculate and display mutation-specific survival statistics from cancer patients using EHR data.
The researchers evaluated all genotyped cancer patients at Vanderbilt University Medical Center; by August 2015 they had sufficient data from 4,310 patients to conduct survival calculations.
The tool worked; it successfully generated tumor-mutation specific survival curves in near real time at the institutional level. For instance, as expected, epidural growth factor reception mutations in lung cancer were associated with superior overall survival, validating the approach of using this algorithm.
The researchers speculated that this type of tool will be a "promising new addition" to the precision oncology environment.
"Retrospective studies are often limited by study of specific time periods and can lead to incomplete conclusions," they said. "Because data is continuously updated in CUSTOM-SEQ, the evidence base is constantly growing. Future work will allow users to interactively explore populations by demographics and treatment exposure, in order to further investigate significant mutation-specific signals.
Prevision medicine is an important new step in patient care, and EHRs are expected to be a major part of such efforts. They are already being used to identify genomic data. President Obama unveiled a new Precision Medicine Initiative in 2015; the pilot program hopes to enroll 79,000 patients in data collection for the initiative by the end of 2016 and more than one million patients overall.
To learn more:
- read the abstract