Is it possible to achieve Oracle’s vision for a national EHR database?

Bold declarations from big tech companies about “fixing healthcare” are nothing new.  Google, Microsoft, Amazon, Apple, IBM, and a slew of other notable names have all tried their hand at sifting through healthcare’s quagmire of digital data woes and competing business incentives – but the results have been mixed, at best. 

Now it’s Oracle’s turn.  Immediately after closing its $28 billion acquisition of Cerner Corporation, the information management giant is planning to shake up the status quo with a national EHR database that sits on top of the health system’s existing infrastructure, unifying all patient records in a single, entirely interoperable, fully private structure. 

Oracle co-founder Larry Ellison laid out his ideas for the database during a webinar celebrating the Cerner acquisition: 

“Each hospital system has their own patient electronic health record database. There are thousands of them throughout the United States. This…fragmentation causes tremendous problems.  We’re going to solve this problem by putting a unified national health records database on top of all of these thousands of separate hospital databases.” 

“We continuously upload electronic health records from the hospital databases into this national database. So that data is up to the second accurate.  Note that to give doctors access to your health records, there’s only one person in the world who can [do that]. And that’s you.  You hold the key to all of your health records in the national EHR database. All the health records are stored in the national EHR database…are all anonymized. Nobody knows who they belong to until you supply the key. So there’s absolute data privacy.” 

There’s a lot to unpack here.  First, there’s no question that the concept is intriguing.  Fragmentation is indeed healthcare’s Achilles’ heel.  Clearly, using a unified, top-down approach to connect thousands of disparate systems would be a game-changer, particularly for public health, clinical research, and emergency care. 

But when we start to examine what it would really mean to build this database, harsh reality starts to set in.  There are (at least) five major challenges that would require a significant rewrite of the financial, technical and clinical landscape, each of which may take decades to achieve.

Convincing everyone to contribute their EHR data

Rivalries among the top EHR solution providers are fierce as companies continue to compete for a share of a relatively finite market.  Cerner, Epic Systems, Allscripts, MEDITECH, and hundreds of other companies have a long history of throwing elbows, and business incentives remain misaligned despite recent industry-wide commitments to greater interoperability.

Health systems, too, have financial reasons to jealously guard their patient populations and their data assets, especially since fee-for-service is still the dominant form of payment in many settings.

Neither the lure of value-based reimbursements nor the intervention of regulators have been able to solve these problems yet (the new information blocking rules are clear proof of that).  Oracle may be in for a steep uphill battle – and possibly a hefty incentive program of their own – to entice EHR vendors and healthcare providers to play ball.

Getting all the data into one place

Even if everyone agrees to contribute, gathering data from thousands of walled gardens isn’t going to be easy.  Many EHR systems are still struggling to undue a decade of coding designed to keep data inside their own systems.

In good news, we now have APIs and extremely viable standards, like FHIR, to help along the process.  And they are working.  Nonetheless, just grinding through the queue of hospitals, primary care providers, specialist offices, urgent care centers, long term care centers, behavioral health facilities, and more to aggregate data in one location would be a massively expensive and lengthy undertaking for a single company.

Preventing a detrimental data monopoly

Speaking of which, Oracle is a for-profit business entity.  While its healthcare activities will be governed by a number of stringent regulations, at the end of the day, it is a publicly traded company.

Consumer tech monopolies, including a few in health IT, are already under scrutiny for anti-competitive behavior and using their size to block innovation, mine data for questionable purposes, and violate consumer protections.  Giving control of all health data to one large technology company could be a disastrous move for patient privacy, and patient-provider experiences.

Solving the patient matching problem 

Ellison’s emphasis on “anonymized” data strikes a particularly curious chord.  Unlike large databases that are primarily designed to deliver randomized, representative sample data for research purposes, such as Epic’s Cosmos, the Oracle database likely cannot function as well as intended if it is as fully blinded as Ellison seemed to imply it would be.

Data that is uploaded and aggregated into a longitudinal record for each individual must be carefully and clearly tied to its owner in some way.  If that data is going to be shuttled to and from disparate providers with different infrastructure, the company will have a major patient matching problem on its hands – especially if it commits to full anonymization of all the data it holds in its cloud.

Even in small health systems, patient matching error rates reach as high as 20 percent.  With millions of individuals sharing key demographic elements, avoiding inappropriately merged or duplicate records without full insight into individual patient identities will be astoundingly difficult.

Creating a workable system of patient consent 

Last but not least is the issue of getting patients to participate.  Patient consent is a delicate and complicated topic, governed by varying levels of health literacy, trust, and education alongside a patchwork of opt-in and opt-out state laws.   

Attitudes toward data ownership are constantly shifting, but patients remain wary of giving out their personal information to big tech. Giving every person in the country a unique key to their health data sounds promising, but ensuring that they can access it, know how to use it, and are comfortable doing so, could be quite challenging and may present health equity issues among vulnerable groups. 

Despite all the potential negatives, it’s still important to dream big in healthcare.  Even if Oracle’s plan doesn’t come to fruition as originally conceived, the company can still do many good things for the interoperability landscape.   

Before acquisition, Cerner was a vocal proponent of information sharing, and it appears that the new owners will continue the quest to streamline and simplify the data exchange environment.  It will be fascinating to see how Oracle works with its new partners across the industry to change the way patients, providers, health plans, and EHR companies interact with digital data.

Jennifer Bresnick is a journalist and freelance content creator with a decade of experience in the health IT industry. Her work has focused on leveraging innovative technology tools to create value, improve health equity, and achieve the promises of the learning health system.