The full promise of electronic health records (EHRs) to support clinical workflows and the measurement of cancer care quality remains elusive due to the great variation in the way EHRs are used by clinicians, as well as by structural shortcomings of the EHRs themselves. This means that much of the information within EHR records remains isolated and unable to be easily extracted to complete external quality measurement reports, further burdening oncology practices.
A large study of EHRs published yesterday in JCO Clinical Cancer Informatics further highlights the need to minimize the administrative burden of quality reporting and to strengthen the use of data standards to improve the real-world quality of care for patients with cancer.
The authors used ASCO’s CancerLinQ® database to examine whether necessary data elements were made available in the EHR and actually used to record care, to make the data compliant with various external quality programs. CancerLinQ is a real-world oncology data platform developed by ASCO that collects and aggregates longitudinal EHR data from oncology practices throughout the United States. By using data from the currently more than six million patients in near real time, CancerLinQ can identify trends and associations between myriad variables, thereby enabling physicians to generate new hypotheses and apply findings to improve care in real-world settings.
In the study, "Chasm Between Cancer Quality Measures and Electronic Health Record Data Quality,” the authors evaluated 19 clinical quality measures (CQMs) in the Centers for Medicare and Medicaid Services’ (CMS) Merit-Based Incentive Payment System (MIPS) during the study period. These CQMs are required by law to be reported for MIPS by eligible physicians per the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) in order for these clinicians to be appropriately reimbursed for the care they deliver. The government’s goal in requiring physicians to report on quality measures is to reward high-value, low-cost care—i.e., pay for quality vs. quantity of care. The authors were looking at whether the structured data in EHRs was sufficient for automated reporting of the CQMs for the MIPS program.
“The author’s research shows that the administrative burden on practitioners to report their results on quality measures remains very high, and that the lack of useful EHR data standards, particularly in a complex field like oncology, has significant real-world consequences, including time spent on duplicate data entry,” said CancerLinQ LLC Chief Executive Officer Sean Khozin, MD, MPH.
The researchers examined a CancerLinQ data set pulled from the 63 practices actively using the CancerLinQ platform at the time the study was conducted, representing more than 1.6 million unique patients with cancer, checking to see whether the necessary data were available to fulfill CQMs against the total number of records that contained patient data to determine the percentage with sufficient data elements.
They found that only 35% of data elements were populated for at least one patient record at 95% of the practices, and for nearly a quarter of the data elements, not a single practice showed any qualifying EHR data. Because MIPS CQMs can’t be computed if there’s any unpopulated data elements, this meant that only two of the 19 CQMs, both related to tobacco (a required data element in many quality programs), were able to be calculated for more than 1% of patients.
“Automated CQM calculations for MIPS using EHR data are simply not practical unless and until we can get common data elements into widespread use,” said lead author Anna E. Schorer, MD. “While no single initiative is likely to resolve all of the deficiencies, in many cases, the current practice of transmitting data in text documents could and should be supplemented by electronic transfer of key elements from the originating source to data fields available to clinicians in the EHR. With commitment from practices or health care systems, laboratories, and software vendors, such interoperability is attainable and would expedite patient care and enable tabulation of quality metrics.”
As the authors note, “Our results highlight the challenges preventing meaningful exchange of oncology data, despite recent enactment of the “information blocking” 21st Century Cures Act legislation. One challenge is the complexity and frequent changes of oncology data elements, which do not lend themselves easily to standardization and maintenance of interoperable, computer-readable, and up-to-date data dictionaries. Another challenge is the lack of a mandate to implement data capture standards within EHRs. Data in structured EHR fields varies widely among implementations because data capture standards have not been widely adopted by EHR systems, and also because practices do not routinely share data capture templates.”
The authors propose several possible solutions for addressing the problems with the MIPS CQM data extraction in particular, as well as the larger challenge of data collection from EHRs:
- Develop policy to withdraw CQMs that couldn’t be automatically extracted
- Incentivize the development of national standards for the structured capture of data elements
- Create specialized standards for capturing and transmitting data, such as the Minimal Common Oncology Data Elements (mCODE®) project or the College of American Pathologists’ Cancer Protocols (CCPs).
Read the full JCO Clinical Cancer Informatics original report.
- JCO Clinical Cancer Informatics article: Improving Cancer Data Interoperability: The Promise of the Minimal Common Oncology Data Elements (mCODE) Initiative
- ASCO Connection article: New Operating Standard: mCODE® Collaboration Is Bringing About Uniformity
- ASCO in Action blog post: New Study Shows Difficulty of Identifying Transgender Individuals Using EHR Data