Patient matching and biometrics: A match made in heaven?
Electronic health records are now all but ubiquitous in hospitals and doctors’ offices in the U.S., and enable better care coordination for patients across different settings. But challenges with patient matching – the ability for organizations to know the person in both records is the same individual – continue to be a major impediment to realizing the potential of EHR interoperability.
Emerging biometric technologies offer a big opportunity to solve some of those patient-matching challenges, one expert showed in a recent HIMSS20 Digital session.
Three patient matching use cases
“There are often three use cases when people talk about patient identification and matching,” said Ben Moscovitch, project director for health information technology at the Pew Charitable Trusts, a co-presenter in the session Healthcare Lessons from Biometric Technology Use Worldwide.
“The first is when a patient walks into a healthcare organization and registration staff attempt to pull up the right record. It’s that use case when biometrics are most often used in healthcare today,” said Moscovitch.
“The second use case is patients remotely demonstrating their identity,” he said. “Increasingly, given new regulations, patients are able to obtain their health records on their personal devices, like a smartphone. Establishing the person requesting the records is actually the person whose record it is another one of those use cases.”
The third use case is knowing that those two healthcare organizations sharing data on a patient to coordinate care are referring to the same individual – two databases communicating with one another, how they are matching that record for that patient, he added.
“Matching today is a problem,” Moscovitch noted. “Research has shown that up to half of patient records may not be matched when transferred from one healthcare facility to another. This can harm patients. For example, if one record indicates that a patient has a particular allergy that is not documented in the other record, and the information is shared, and the patient is given a drug to which they are allergic.”
A survey has shown that one in five CIOs has said that at least one patient has been harmed due to poor matching, he added.
The costs challenge
“It also drives up costs,” he said. “Some research has shown that having duplicate records in the system can cost $96 to fix each duplicate record. It can cost up to $1,000 to remove data of two patients that have been comingled.”
Matching today is difficult. Demographic data is used. Typos, similarities, missing data and other factors lead to poor match rates.
“What Pew did over the course of a couple of years is to try and identify a solution to this patient matching problem,” Moscovitch explained. “One of the things we did was focus groups in four states with patients to understand their perspectives on patient matching and some of the solutions they would find most amenable.”
Many patients expressed an interest in using biometrics for matching because, from their perspective, they are using biometrics every day to unlock their phones, board an airplane or go to Disney World. So why can’t they also use biometrics to match their records and solve this problem that can lead to patient harm, he said.
Talking with healthcare execs
“We also conducted interviews with healthcare executives across the country and found there was interest in biometrics, but there were some concerns around cost and privacy,” said Moscovitch.
“Biometrics are being used in healthcare today, but that is typically for getting a patient to pull up the right record. Cross-provider patient-matching between two organizations, that use case for biometrics is not widely occurring.”
There are six key unanswered questions to that last use case, Moscovitch said:
- How do you address the tension between privacy and the interoperability of the biometrics?
- What are patients willing to use, and how should informed consent be handled?
- Is healthcare different, as in infection control?
- How is biometrics deployed in tens of thousands of sites?
- Should everyone use the same biometric modality, and if not, how do you compare different modalities?
- How can you maintain accuracy with different patient subpopulations and large databases?
Daniel Bachenheimer, principal director at Accenture Security, also co-presented in the session. He went on to offer a detailed “Biometrics 101” and discussed three prominent aspects of biometrics technologies: enrollment, verification and identification.
Bachenheimer said in Accenture Security’s work with biometrics technologies, there have been three major lessons learned that healthcare organizations can apply:
- Standards-based biometrics image exchange is needed for interoperability among heterogeneous systems.
- Privacy considerations are key for compliance and customer adoption, including transparency, consent and data protection.
- Accuracy in matching patient records will be enhanced where biometrics can be used in conjunction with demographic information.