Medical data sharing: Are we there yet?

The move toward ubiquitous electronic health record sharing across the US is accelerating, but obstacles remain for providers, payers, patients, and other stakeholders.

1 2 Page 2
Page 2 of 2

From documents to discrete data

One challenge for doctors in accessing data from current HIEs is that patient records are stored as documents, rather than as discrete data elements that can be queried to return exactly the information the doctor most needs to know when they need to know it.

“Exchanging data using standardized documents is not the best solution for all types of exchanges,” says Health Gorilla’s Lane, so “ONC is pushing forward to support TEFCA exchange using Fast Healthcare Interoperability Resources (FHIR), an open-industry standard created by [electronic health information standards group] Health Level Seven for data sharing using modern application programming resources (APIs).” FHIR “provides for the more flexible exchange of data elements for both public health and payer exchanges,” Lane says.

“The FHIR and TEFCA standards are improving interoperability, creating standards for data exchange, and pushing the responsibility from developers to the providers who use those systems,” Stump says.

“By using FHIR, we’ll be able to support many additional use cases, and it will be more efficient in many cases,” says Lane. However, the QHINs and EHR system vendors will need to update their systems to support the new FHIR standard. “We’ve had slow uptake of FHIR on the existing exchanges,” Lane says. But, he adds, “Everything is moving forward in a coordinated manner. We’re going to see a big shift in interoperability over the next five years or so.”

Another challenge is that EHRs are built on legacy technologies, says CommonWell’s Wilder. “They still use old-school SOAP and XML on the edge, while phones are using RESTful services. It’s kind of archaic, but it works.”

The gaps

At this point not everyone is connected, including providers using smaller EHR system vendors, state exchanges, and 30% of hospitals. Bingman from eHealth Exchange is painfully aware of the consequences of being a patient at one of these hospitals. Her records are locked up in a hospital that doesn’t yet share its data. “They have an old, antiquated system, they don’t want to join our exchange, and I can’t get to any of my information,” she says.

The launch of the QHINs won’t connect all these organizations instantly, says Tripathi. “Those are the gaps that TEFCA will fill. We will identify the hospitals that aren’t connected and help them get into these systems.”

So far, 51% of hospitals are aware of TEFCA and will participate in the QHIN exchanges, according to an ONC data brief. If that sounds low, just 1% said they would not participate, and Tripathi expects most hospitals to join as the QHINs get closer to launching. “None of the vendors are charging additional fees for connecting to these networks at present,” he says, so “budgets will — hopefully — not be an obstacle.”

Also, says Bingman, “lab providers aren’t connecting with eHealth Exchange.” Labs such as Quest Diagnostics could be part of an exchange, but the lab data already goes back to the ordering hospital or a physician’s office.

While labs share data with the physician offices in most cases, the fact that they’re not participating in the exchanges still represents a gap, she says. “For example, lab results ordered by hospice or specialty practices that don’t use the larger EHRs will not be included.”

Medical imaging is another big gap. “If you have a CT scan today, you have to pick up the image and bring it to your doctor on a disc,” she says. Also, paramedics and emergency medicine would benefit from having basic medical information on patients, including medications.

There’s also a limit to the types of data provided. For example, the major EHRs receive information from pharmacies as to what medications have been prescribed, but not whether they were picked up or consumed.

Some technical aspects of integrating provider EHRs with exchanges also need working out. “Today there’s still no universal data architecture,” Stump says. Yale New Haven health chose the OMOP Common Data model and must map data to it.

“It takes work to make that happen seamlessly,” she says. “Every time a payer changes a field name, I have to change my interface to have it map that data field to my model. Data standards and governance need to be more universally taught and used.”

Wilder at CommonWell agrees. “If you ingest data, you have to have a data model,” he says. “So when a vendor’s data model changes, you have to redo it or the downstream consequences can be very severe.”

Another issue: The QHINs will only allow data sharing within the US and its territories, so those traveling abroad will still need to take their records with them — if they have full access to those records. “Global interoperability is something we need to work on,” Tripathi says. But patients do have the ability to download the USCDI summary of their records or access them through portals or apps, he adds.

One idea to make this even easier is to give patients a QR code that either provides an overseas hospital with basic health information, or that allows the hospital to access the patient’s record securely. “That’s not in place today, but we’re exploring that with the [SMART Health IT] API,” he says.

Interoperating with healthcare systems abroad would carry its own unique challenges. In the European Union, for example, member countries form a single trading block, but have different models for healthcare, says Tripathi.

“In the UK, the National Health Service comprises a single payer, and all providers are under the same organizational umbrella, so the data is, in effect, combined, [while] in Israel, three or four HMOs cover the entire country, so nationwide data-sharing is easier to figure out. The point is that interoperability is easier within those countries that have more integrated or consolidated healthcare delivery systems, [but] that doesn’t necessarily mean that interoperability between those countries is that much easier,” he says.

Finally, there’s some data that may never be included in provider-centric EHR systems and data exchanges. “EHRs are heavily indexed around clinical data,” says Batra at Deloitte. They’re essentially billing systems that capture data that’s useful to clinicians. “But there’s a whole slew of health information that goes beyond that, including environmental, behavioral, emotional, cultural dynamics. When integrated, that can change how you interpret the clinical data.”

That, he says, probably won’t come from the incumbent EHR vendors but from patient-centric “disruptive entrants” that are focused on consumer access. In that race, there are “literally thousands of organizations trying to figure this out,” but no standards or winners yet.

The five-year journey ahead

Once the QHINs are up and running, the next step is to expand the purpose of use beyond treatment and individual access, says Bingman. Payments and operations will come up next. “Right now, payers aren’t participating in the exchanges,” she says, so claims and clinical data remain separate. Data exchange for public health use, such as by government agencies, is also a major focus.

The push to get to 100% participation from all types of healthcare providers will also need to continue. Incentives may help. “TEFCA isn’t going to be a cure-all,” says Bingman. “It depends completely on how this rolls out and what kind of teeth CMS [the Medicare and Medicaid payer] put into it. If CMS makes a rule that they’ll pay more if providers exchange data electronically, or pay less if they don’t, that’s when providers will say ‘we have to do it.’ CMS can use the carrot and stick pretty well.”

“We’re at least a year away from any big movement, and it will probably be longer than that for using exchanges for additional purposes,” such as life insurance, prescription drug monitoring, syndromic surveillance notifications, and research, she says.

All of this is going to take time, Wilder says. “You’ll see movement in 2024, but the tail end of the adoption curve won’t happen until 2025 or 2026. Five years out is when we finally get the remaining 30% connected.”

Copyright © 2023 IDG Communications, Inc.

1 2 Page 2
Page 2 of 2
It’s time to break the ChatGPT habit
Shop Tech Products at Amazon