The use of electronic health data for quality measurement and improvement in healthcare has not yet realized its full potential, a problem that must be addressed if the industry is to successfully transition from fee-for-service to value-based payment.
That’s the consensus among industry stakeholders who participated in a roundtable discussion hosted last week by the Office of the National Coordinator for Health Information Technology.
“We are at a critical point in how we think about quality measurements and how we think about quality improvement in general,” Robert Anthony, senior policy advisor in ONC’s Office of Clinical Quality and Safety, told the gathered group of subject matter experts.
Quality measures are vital as the industry moves from fee-for-service to value-based payment. According to consulting firm Discern Health, as much as 80 percent of U.S. healthcare spending will be linked to quality measures or value-based payment models by 2020.
Also See: CMS proposes rule for collecting, submitting quality data
“We’re in a world where it’s not just about reporting any result, it’s about reporting your optimal result—and, payment is tied to it,” said David Kendrick, MD, chair of the Department of Medical Informatics at the University of Oklahoma’s School of Community Medicine. “We’re seeing the level of emotion about measurement go through the roof among providers, and that means they’re starting to really care about all those terminology issues and the issues in their EHRs.”
Despite the widespread adoption of EHRs, the mere fact that providers are using the systems does not automatically translate into improved quality of care. In addition, although health IT has the potential to greatly improve the quality of care, the evidence that HIT improves health outcomes is still relatively limited.
To help address the problem, Kendrick argued that measures must be standardized, replicable, validated, timely, as well as actionable.
“Medicare and most payers are betting the farm on measurement right now, and I’m very worried about our infrastructure to support that,” he added. “Claims data is a mile wide but only an inch deep. So, it has the unique benefit of encompassing most of what’s happened with the patient, but it doesn’t have the deep clinical variables in it (that) we would like to have, and it’s certainly not timely enough for most measurement that we would want to do for actual quality improvement.”
On the other hand, Kendrick said electronic health record data is “a mile deep, because it has all of that clinical information in it, but it’s only an inch wide.”
While EHRs “may be considered the core for data for eMeasurement, it’s really important—and I would even say essential—to integrate with other types of data in order to truly capture what’s happening at the patient level and being able to capture what happens across multiple providers and systems,” said Sarah Sampsel, vice president at Discern Health. “To improve quality of care, not just the quality of life for that person, we have to be able to capture more than what’s in an EHR.”
Speaking figuratively, Micky Tripathi, president and CEO of the Massachusetts eHealth Collaborative, said “it’s criminal that providers don’t have access to all their claims data—they only get it for the risk patients, if they happen to have a risk contract, but they don’t get it for everyone else.”
Another challenge cited by Kendrick is that “real patients’ care is scattered across a number of locations,” and “to do measurement appropriately, we really need to have all of this (data) so that we have the most complete picture on each patient that we’re measuring.”
Likewise, Tripathi lamented the fact that healthcare is fragmented and, as a result, so is the electronic health data.
“If you look at the distribution of patient visits across the country, 68 percent of patient visits happen in small physician practices,” said Tripathi. “When we think about quality measurement, we tend to focus on the Mayo Clinic and Intermountain Healthcare. But most of the action and therefore most of the data is in smaller settings.”