No one solution to patient matching challenge, GAO says
- Efforts to accurately match patient records will continue to challenge providers, payers and others, with no one-size-fits-all approach to ensuring that information in different health records refers to the same patient, the U.S. Government Accountability Office says in a Tuesday report to Congress.
- GAO interviewed 37 stakeholders to see how they match medical records, the challenges they face and what can be done to improve patient matching. The approaches ranged from fully automated to a combination of automated and manual verification.
- Stakeholders blamed incorrect, incomplete or inconsistently formatted demographic data in patient records for many of the problems they encounter in correctly matching patients.
Patient matching is a huge problem in healthcare and a major barrier in achieving full interoperability. Mismatching patients and records can lead to missed diagnoses or treatments, putting patient safety at risk. Over a 30-month period in 2013 to 2015, ECRI’s Patient Safety Organization counted 7,613 wrong-patient events voluntarily reported by 181 healthcare organizations.
Patient matching is also time-consuming. A 2015 survey by the American Health Information Management Association found most organizations were spending time weekly doing patient matching cleanup.
The 21st Century Cures Act requires GAO to study patient record matching and take steps to reduce mismatches, and it directs HHS and the Office of the National Coordinator for Health IT to support nationwide exchange of patient information.
In the GAO report, stakeholders described an array of approaches to patient record matching. Of seven providers interviewed, all said they use manual matching as one way to verify records refer to the same patient, while six also use digital tools to automatically identify and match records stored in EHRs.
Health information exchange organizations rely on software based on varying algorithms, but all include name, sex, date of birth and address to match records, according to the report. Six of the seven providers reported sometimes using HIEs to exchange and match records, but none relied on them as their chief means of matching records and data exchange.
"According to the stakeholders we interviewed, it is difficult to determine the accuracy of the health IT tools used to match patients' medical records automatically," according to the report. "While the algorithms typically match records belonging to a patient and identify potential matches that need to be manually reviewed, users of these algorithms do not know how many matches the algorithm may have failed to make."
Among efforts to improve matching was a 2017 initiative by 23 Texas providers that implemented standards for how staff record patients' demographic data. ONC also launched a $75,000 challenge to improve patient matching algorithms.
To move the ball forward, stakeholders suggested employing common standards for recording demographic data, sharing best practices and other resources and encouraging public-private collaboration to improve matching.
"Next step would be for @ONC_HealthIT advance data standardization," tweeted Ben Moscovitch, project director for health information technology at the Pew Charitable Trusts. "ONC could do this through the Common Clinical Data Set/US Core Data for Interoperability. Specify standard for address & what additional data to use for matching (eg mailing address)."
GAO's findings echo those of a recent Pew report, which concluded no one solution currently exists to achieve highly reliable matches for all patients across all EHR systems. Pew recommended clarifying government funding restrictions for unique patient identifiers and agreeing on standardized demographics among near-term steps that could be taken. Longer-term approaches included creating a national oversight body and exploring the use of biometrics to securely match patient records.
In a report published last summer, RAND suggested using mobile phones and smartphones as a patient-centric way to boost patient matching — such as verifying a patient's phone number with the provider or having patients check in for appointments using an app that shares updated identity details.