WASHINGTON — Here’s a health disparity laid bare by COVID-19 that you might not have expected: patient matching — identification and linking of one patient’s data within and across health systems — is worse for minorities and underserved groups.
A recent survey by OCHIN (formerly the Oregon Community Health Information Network), a national non-profit health information technology provider for 500 care delivery sites, found that:
- Black patients make up 13% of OCHIN’s population and 21% of duplicate records, almost twice the expected rate
- Hispanic/Latino patients make up 21% of the population that OCHIN serves, yet they make up 35% of the duplications, for almost twice the expected rate
- 3% of OCHIN patients have experienced homelessness, but they account for 12% of mismatches and duplicates, or almost three times the expected rate
In addition, Black patients were 2.5 times more likely than non-Hispanic white patients to have a COVID-19 diagnosis in their electronic health record (EHR), while Hispanic patients were twice as likely as white patients to have a COVID-19 diagnosis in the EHR, according to the survey, which was the subject of an article in the Journal of AHIMA (the American Health Information Management Association).
“This is just one healthcare organization; others are starting to learn … what the real impact is on health disparities,” said Thomas Leary, vice president of government relations at HIMSS, an organization for health IT firms, at a virtual meeting on patient identity and matching sponsored by the Office of the National Coordinator for Health Information Technology (ONC).
These types of mismatches would be a problem under any circumstance, but it’s even worse when getting a vaccine — such as one for COVID-19 — is involved, especially if the vaccine requires more than one dose. “Preparing for a vaccine, particularly vaccines that require multiple doses,” will make matching especially important, said Leary. “If the COVID-19 vaccine requires more than one dose, making sure we know who’s had the disease, who’s been vaccinated, what are the patient outcomes and where the stockpile is for distribution — all of that needs to be taken into consideration.”
Leary’s group is part of a coalition called Patient ID NOW, which seeks to have the government develop a unique patient identifier (UPI) for every patient. The group is “solution-agnostic” in terms of what type of identifier is developed, as long as there is one, explained Meryl Bloomrosen, senior director of federal affairs at Premier Healthcare Alliance, another coalition member.
The Health Insurance Portability and Accountability Act (HIPAA), a federal law passed in 1996 to regulate health information, included a provision allowing the government to begin developing a UPI. However, in 1999, then-Rep. Ron Paul, MD (R-Texas), a libertarian, inserted a ban on any government money being spent on a UPI into an appropriations bill.
“There was some concern that the government was going to identify each one of us and take our information and watch us or control us,” Joel White told MedPage Today in an interview this past February. White is executive director of the Health Innovation Alliance, a group of 75 different organizations focused on using technology and data to make healthcare work better.
Since then, Paul’s son, Sen. Rand Paul, MD (R-Ky.), has taken up the cause, and has argued against the idea. The House passed a measure last year to get rid of the ban as part of an appropriations bill, but the provision didn’t make it in the Senate, White said.
The appropriations bill provision against developing a patient identifier, along with other barriers, “have led to a failure to institute a nationwide patient identification strategy,” said Leary. “We’re not going to get to interoperable healthcare unless we have a national strategy for patient identification. We can’t wait any longer.”
Noting that his son was 2-years-old when people started looking at this issue and just went off to college last week, Leary added, “We need HHS [the Department of Health and Human Services] as an equal partner at the table with the healthcare community. We need a national vision for interoperability and the ability to share information across jurisdictions and across communities.”
But not everyone is on board with getting to a UPI. “There are a lot of challenges that have to be solved before we talk about using a UPI,” said Norman Adams, PhD, chief data scientist at Surescripts, a health information technology company that provides support for e-prescribing. “You’re putting a lot of trust in the person producing those identifiers. If there are issues — if they’ve created false positives and false negatives in their database — you’re essentially importing those. So there has to be some transparency for anybody who wants to propose a UPI, in how they produce that identifier, and there has to be some way to evaluate [it]. A lot of people are looking at UPI as a magic bullet, and it’s not.”
And “once a UPI is out there and all those groups are using it to enhance their matching — what are they going to do when discover error? How does that get distributed to everyone using the identifier, and how do they fix it?” Adams said. “So I think that must be considered.” Any group that adopts a UPI should never rely solely on it; they should always use their own technologies to verify those matches, he added.
Adams also put in a pitch for another element to improve patient matching. “If we could get access to birth records, that would help us tremendously; we could build up a complete picture of our users,” he said.
Last Updated September 01, 2020