Harnessing Clinical Data Interoperability for Personalized Care Advancement

Explore the need for improved clinical health care data acquisition and interoperability using APIs and the FHIR data standard with Don Rucker, MD, Chief Strategy Officer at 1upHealth and former National Coordinator at Office of the National Coordinator for Health Information Technology (ONC).  

Please share your name, title, affiliation, and a bit about your background. 

My name is Don Rucker, and I am the Chief Strategy Officer at 1upHealth. We're a FHIR in the cloud data platform that leverages all kinds of interesting cloud technologies in the cloud to make the payer and provider ecosystem interactions more efficient. I was a former national coordinator at ONC [Office of the National Coordinator for Health Information Technology] in the Department of Health and Human Services. As a presidential appointee, I had the honor of implementing the 21st Century Cures Act that was passed almost unanimously in 2016 by Congress. I also practice internal and emergency medicine. 

What are the current main challenges for consolidating and optimizing clinical data that supports a range of workflows including prior authorization, quality reporting, risk adjustment, and care management? 

The big-picture challenge we’ve had is to utterly separate data streams in health care. One stream is financial data like claims data. You go to the doctor, and they assign a code to the services and send a bill to a clearing house. It then goes to a payer in a historic X12 data format.   

Totally separate from that is the clinical care data that's sitting in the electronic medical records, typically in proprietary formats. Up until very recently, there's been a lack of standardization for these records. These two data streams should have been connected before but haven’t been.  

We've been trying to introduce value-based care for 20 years in the US. It somewhat goes without saying that if you can't connect what you got with what you paid for it, it's pretty much impossible to ascertain value. I mean, would you ever buy anything in the store without knowing the price? However, in health care, we have been doing that forever and that is finally changing.  

Why is there an increased need for payers to move to more effective clinical data acquisition workflows?  

We all know American health care is a poor value deal. Everybody who's ever been a patient or bought insurance knows something is deeply wrong with our health care system on many levels. This problem goes back to 1942 in WWII when health care was made a pretax benefit, making our employers the purchasers of health care and not patients. The control has been stripped from patients for over 80 years now. Health care is a big political issue, and Congress is always trying to figure out how to change the name of the game. They’re trying to connect the moving pieces of prior authorization, quality measures, network design, and more to efficient purchasing with efficient provision of care. Increasingly, payers and providers are forcibly intertwined.  

I also think we're seeing policy mandates coming forcing more payer and provider interaction. Policy folks such as CMS [Centers for Medicare & Medicaid Services] are doubling down by making rules to require APIs to share. The CMS interoperability rule being finalized is going to require APIs between providers and payers and in payer-to-payer exchange. This way when you switch from one plan to another, all your information such as hard-won coverage decisions for medication under prior authorization isn't lost. All of this is so that patients can control their data instead of portals. Requiring prior authorization history is a part of patient empowerment. The changes we are discussing are all predicated on shared data using the FHIR data standard.  

To understand the FHIR data standard, we must acknowledge the long history of trying to connect computers and health care. At scale, this goes back to the 1980s when we invented machines to automate medical testing, such as blood tests, which became digitally reported and led to the question of how to best insert data into the patients’ charts and hospital workflows. We've never had a standard on the core, computable clinical data in medical records before that would allow clinicians to examine blood pressure data to prevent a stroke or look at cholesterol levels and compute to prevent a heart attack.  

Around 2010 FHIR was introduced, which stands for Fast Healthcare Interoperability Resources. What the hell is that? It was the creation of an Australian named Graham Grieve who asked why don't we use the same JavaScript object notation that every app on their smartphone uses to communicate back to the server? As you're sitting here, your smartphone is sending thousands of JavaScript communications back and forth between the supporting servers for the apps running on your phone. FHIR is the health care–specific version of that JavaScript notation.  

Can you provide any examples or case studies showcasing lessons learned in the process of implementing clinical data acquisition workflows? 

The first step is to think about the point of the workflows and the tools needed to curate them. What are the communication protocols and standards? In health care it is also always important to consider security.  

A benefit of the modern world is that we've already solved many of these problems. We have electronic health records [EHRs], and the data is, at least digitally, under the Cures Act. We've now required since December of last year that API data must be standardized and accessible. This is not all the data in the EHR, just a set called the US core data for interoperability such as medication lists, allergy lists, demographics, etc. We finally have the tech in place even if the economic incentives are still in progress. Congress outlawed information blocking, and Health and Human Services are slowly enforcing our access as patients to the data we paid for that is about our bodies.  

How can clinical data acquisition be changed without sacrificing scalability to the payer market? 

I think we have the opportunity now with FHIR data formats for payers to access clinical data. Historically, they only received claims data through a clearing house in an X12 transaction. Payers under HIPAA treatment payment operations have a legal right to access the clinical data on the claims they're paying for. As patients, we generally want the payers to have some oversight of what's being done on our behalf, which we indirectly are paying for by paying the payer. We'd like them to be maximally empowered as our agents. Historically, hospitals don't want to share that data because it may be used against them in price negotiation. Increasingly though, this data is being shared whether it's prior authorization, accountable care organizations, or at-risk contracts.  

The bigger-picture problem that we’re starting to solve is how we use this data for not just quality measurement but [to] expand into overall value and care? We're making the quality-measurement world far more digital instead of manual. Our aspiration is to the same sets of tools and level of sophistication that retailers use to sell us stuff. For example, every time you open your web browser the ads you see are the product of sophisticated algorithms to personalize your next purchase. Wouldn't we like to have that same level of calculation to help prevent a stroke or a heart attack or maybe even lose some weight? In a world where your phone tracks your heart rate, why couldn’t that be used in my health care decision-making? But that data is literally unavailable to the health care workers I've paid to take care of me.  

What are your personal recommendations for future steps toward more efficient clinical data acquisition? 

I think one of the next steps is for payers to really insist on accessing and analyzing the clinical data for people in their networks. They should also receive this data in a computable format. For example, the VA [US Department of Veterans Affairs] must provide a lot of care outside the VA. They should insist on using the bulk FHIR API, which is required of every certified EHR in the country, to analyze data and then contract accordingly. Everybody can do that because it'll make care more efficient, lower cost, and reduce friction for patients such as prior authorization denials. 

I have spent hundreds of hours on getting prior authorization approvals for devices, for diseases with no plausible alternatives or cures other than the thing that I'm lobbying for. Much of this back and forth is highly expensive and wastes the patient's time and inundates providers with emails. Also, why isn’t appointment scheduling fully automated, or prescription renewals?  

Most Americans have to coordinate their health care themselves or go without. In contrast, as a consumer, when you order an item online you never have to arrange payment with a wholesaler, pick up the package at the Amazon warehouse, and figure out how to get it to your house or apartment. Yet that is the equivalent of what we're doing in health care when it could be totally digitalized. When we think about interoperability, it must be computable interoperability. We need to move away from these old policies that are largely there to prevent incumbents from competition and move to modern methods that are proconsumer.