A new JCO Clinical Cancer Informatics (JCO CCI) article highlights the challenges in and need for obtaining real-world evidence for assessing the effectiveness of precision medicine therapies on areas such as survival, quality of life, costs, and health disparities. The article, “Clinical Molecular Marker Testing Data Capture to Promote Precision Medicine Research Within the Cancer Research Network,” was published in JCO CCI on September 5.
Real-world evidence can be difficult to obtain primarily due to the differing or incomplete ways in which data is entered into or captured by electronic health records (EHRs). The authors had particular difficulty in finding results of germline tests and tissue-based tumor markers as they searched various Cancer Research Network (CRN) health systems. With germline tests, the results “were stored as scanned image reports in the EHR,” the authors note. “There were no separate electronic data feeds from testing laboratories for test results. Some germline tests were noted as text in patients’ clinic visit notes, but the results were not consistently or standardly documented.” Additionally, some health systems used custom codes generated by each respective system to identify particular genetic tests.
As with the germline tests, the tissue-based tumor markers tests results were stored in each health system’s EHR as scanned image reports. “Results of tumor marker tests were inconsistently reported in the text of pathology reports, and these lacked standard documentation formats,” the authors state. “Often, the pathology report referenced the scanned image results as an addendum, without describing the type of test performed or the results of the test.”
Overall, using the best data available—those from the Health Care Systems Research Network Virtual Data Warehouse—the authors only were able to capture population-level data from six out of 14 tumor-based markers and none of the nine germline markers. “This provides enormous challenges for evaluating the use of precision medicine approaches in real world clinical settings,” the authors state.
ASCO’s mCODE Initiative
To help mitigate the problem of disparate information entered into and captured by EHRs, ASCO, in collaboration with the MITRE Corporation and other stakeholders, has established the mCODE™ initiative. Short for Minimal Common Oncology Data Elements, mCODE has assembled a core set of structured data elements for oncology EHRs to form the basic data that would populate all EHRs for patients with cancer. The data elements are intended to be sufficiently precise so that a query related to a particular patient would return a cohort of matched, de-identified patients from a larger database, such as CancerLinQ®. This would facilitate cancer data interoperability and improve overall cancer data quality for patient care and research.
The first version of the core mCODE data specifications has been completed and was released June 1, 2019. The data dictionary is available free of charge for download at www.mCODEinitiative.org. The mCODE data specification is currently undergoing implementation pilots at two health systems across the U.S., with results expected later in 2019. Further details about mCODE and information for organizations wishing to participate in the initiative are available at the same web address.
JCO CCI is an online-only, peer-reviewed, interdisciplinary journal that original research, reviews, and editorials on biomedical informatics studies relevant to cancer treatment, care delivery, outcomes, and prevention.