How far has health IT come in meeting Meaningful Use standards?
“An ounce of performance is worth pounds of promises” — Mae West
Nearly 10 years ago, I wrote about the potential of using data from electronic health records (EHRs) for quality measurement and population health. Back then, a series of impediments to interoperability were identified. Some came from EHR limitations and some stemmed from the intricacy of any standard that organizes complex medical data. Despite those concerns, I and many others in the field remained optimistic about the potential of clinical data standards.
As we enter 2019, I’ve taken a moment to reflect on just how far the industry has come. We’ve been through three stages of Meaningful Use — the federal program for EHR adoption — each of which has used progressively better standards to summarize and communicate clinical data. First there was HITSP C-32, then there was C-CDA 1.1 and today there is C-CDA 2.1 for clinical documents.
These standards have integrated years of experience and hard work from the community, but like all standards, they have flaws and inadequacies. To examine the most recent version, C-CDA 2.1, I had the chance to lead a large research team that included colleagues from the U.S. Department of Veterans Affairs, the Office of the National Coordinator for Health IT and Intersystems. We tested more than 50 certified technologies to see how they did at achieving interoperability.
The results show promise. We saw progress on many issues we identified in previous research, like this 2014 study on Meaningful Use-certified EHRs. But we still found some issues that could be improved through better implementation of the standard and tools that check compliance and integrity. In particular, we found that schematron validation, which checks XML released as part of the C-CDA standard, was inadequate to identify some critical issues affecting clinical data interoperability.
Our complete study results were presented and recognized as a “Most Distinguished Paper” at the American Medical Informatics Association (AMIA) 2018 Annual Symposium, which is available for download.
One of the most intriguing things revealed in the research is the massive variation among different EHRs displaying the same medical data. The C-CDA standard focuses primarily on machine-readable content, leaving the rendering of human-readable tables and lists up to individual EHR developers.
As one example, 38 EHRs display the same three medications in radically different ways; here are few samples of the variation for the medication directions (similar variation exists for the medication name and end date):
Looking ahead, the industry will continue to close the remaining gaps on semantic interoperability. However, this improvement will not occur because of the introduction of a new standard, such as the Fast Healthcare Interoperability Resources (FHIR). Standards, both old and new, will present the heterogeneity demonstrated in our research because of the way medical care is delivered and the varied technologies that are used.
Healthcare needs a second-tier ecosystem to normalize medical data. This is like Google, which organizes the heterogenous ways content exists on the internet to make it searchable. In fact, Eric Schmidt said as much during his HIMSS 2018 keynote.
This middle tier will deliver semantic interoperability and unlock digital uses we cannot even fathom today. Looking back 10 years, I see our progress toward the promise of interoperability. Looking forward, I am excited at what the next several years will hold.
John D’Amore is president and chief strategy officer at Diameter Health, an enabler of clinical insight through the cleansing, deduplication and enrichment of clinical data across the care continuum.
Get the best insights in healthcare analytics directly to your inbox.