Experts Discuss Ways to Improve Accuracy of Electronic Health Record Matching

When people need medical care, correct and complete electronic health records can help doctors and nurses promptly determine the best treatment, forgo duplicative tests, and avoid harmful errors. Unfortunately, in many cases—such as emergencies—clinicians do not have an accurate and thorough patient history at their fingertips because of the difficulties in matching digital records for the same person within and across health care facilities.

The Pew Charitable Trusts brought together a panel of experts in health information technology on Oct. 2 to discuss approaches that could help address challenges in record matching and interoperability (the accurate exchange of patients’ health data among the places they seek care). The event coincided with the release of a report examining four potential ways to improve record-matching rates and detailing a series of recommendations. The panelists agreed that no one solution can solve this problem, but a common set of strategies across the nation and standards for electronic health data are key ingredients.

Micky Tripathi, president and CEO of the Massachusetts eHealth Collaborative, said his team interviewed executives at hospitals, medical practices, and organizations that facilitate the exchange of patient records between health facilities. “Increasingly, there is growing demand—and where we’re seeing a real gap [due to] matching—is with respect to interoperability,” he explained. Many interviewees noted that their organizations are getting more records and requests for information from other health care facilities and systems. In those circumstances, according to Tripathi, “people are saying, well, our matching rate there is maybe 50 percent, 60 percent.”

To address this problem, panelists underscored two key strategies: data collection standards and biometrics. For the former, Shaun Grannis, M.D., director of the Center for Biomedical Informatics at the Regenstrief Institute, conducted an analysis using health data in Indiana to study the effects of standardizing patient information captured in electronic health records (EHRs). That analysis, which has yet to be published, found that recording addresses and last names in a consistent format would help improve match rates. “Standardization of data elements … can improve matching—in the right combinations—by double-digit decimal points,” he said.

Second, speakers discussed both the potential and challenges associated with the use of biometrics—such as an iris or fingerprint scan—for matching. That approach could further boost record matching rates, but would likely take more time and cooperation among developers and users of EHRs to implement fully. Joe Trelin, senior vice president at CLEAR—a company that uses biometrics across various fields, including aviation—explained that developers would need to agree on how to capture, process, and use the information, all while ensuring patient privacy remains a priority.

Panelists also discussed giving patients a greater role in matching their records. As Robert Rudin, an information scientist at the Rand Corp., explained, patients know their information and have a personal stake in the problem. “After all, if their records don’t get matched, it’s their health that gets adversely affected,” he said. To that end, Rand conducted an analysis—supported by Pew—that found smartphones could improve matching, either by letting patients verify their identity through a text message or other alert, or through the use of applications to share data.

Speakers also underscored how coordination on a larger scale could relieve some of the patient identification and record-matching burdens on individual hospitals and medical practices. “Every health care organization today has … to act as the identity service provider,” said Catherine Schulten, vice president for product management at LifeMed ID. If health systems could get those services from a shared external entity, she suggested, “we can start lifting some of that heavy load.”

Mark LaRow, the CEO of Verato— a company that uses data from sources such as the U.S. Postal Service to help clients match health records—proposed taking a public utility-style approach to EHR interoperability. Rather than relying on every health care provider to invest in and maintain compatible matching technologies and accurate patient data, LaRow recommended a system akin to an electric grid, in which each hospital and medical practice could use a “referential matching-based utility service” that matches records for the same person at different facilities.

For matching improvements in the short term, the Office of the National Coordinator for Health Information Technology—the federal agency that oversees electronic health records—should identify nationwide standards for demographic data, such as addresses. In the longer term, technological innovations like biometrics have promise, but success will require greater collaboration across the health care industry to determine how to collect, use, and share data securely while respecting patient privacy.

As Josh Rising, Pew’s director of health care programs, explained, “Whether that’s standards for using and protecting biometric data or sharing information among providers, standards are the linchpin to success.” That emphasis should remain as the effort to improve patient matching moves forward.

Ben Moscovitch directs The Pew Charitable Trusts’ health information technology initiative.