Today, EHRs are in use by 95+% of eligible providers. With that massive adoption, EHRs, at least the meaningful use certified EHRs deemed by the government worthy of carrot and stick adoption incentives, play two foundational roles in healthcare today.
- EHRs are the hub of patient data. They collect and store both subjective information and objective data from devices. They store reports and other kinds of file attachments. They store discrete data.
- EHRs are the hub of clinical workflow. Though EHRs are designed for general workflows, they don’t provide the tailored fit clinicians actually need forcing them to have to adapt their workflows to EHRs and not the other way around as it should be. Orders, messages, referrals, you name it. If providers are not entering or looking for data in the EHR, they are using the EHR to accomplish a task.
These two roles are foundational to building solutions, technology, or technology-enabled processes on top of the EHR; there are exceptions to this rule but, increasingly, even these exceptions are being forced to integrate. This changes everything, especially for those innovators in healthcare that are trying to develop solutions for providers and patients. But, almost as importantly, it forces EHRs to now either a) build new solutions themselves, b) support new 3rd party solutions, or c) risk replacement by newer EHRs that do either a or b. This process isn’t happening overnight, but it is happening.
Healthcare technology tends to cycle every 15 years or so between best-of-breed solutions and all-in-one platforms. Today’s EHRs are at the tail end of a 15-year cycle that has seen their adoption grow massively as healthcare providers have implemented one of a few EHR options as platforms. Now, we are starting to cycle back to best-of-breed solutions as it has become abundantly clear that monolithic EHRs cannot develop solutions fast enough to meet the demands of the market. One platform cannot do everything. And EHRs, at least the major ones, are trying a mix of the a and b strategy above to hold their place as the central healthcare IT software.
EHRs still control access to patient data
EHRs still control access to data and clinician workflows and have used that control to slow down adoption of new digital health solutions. EHRs have developed proprietary programs, or app stores/orchards/ programs, in attempts to support, whether that support is real or perceived, 3rd party solutions — Epic has App Orchard, Cerner has Cerner Cares, Allscripts has the Allscripts Developer Program, and Athena has More Disruption Please (there are more examples, but these are the big four). While programs from Athena and Allscripts are mature at this point, the emerging programs from Epic and Cerner have been slow to take off at best or failing at achieving any meaningful adoption at worst. The reasons for their slow take off or failure are partially technology-related (the app stores are proprietary) and partially business-model related (Apple can take 30% of app store partner revenue because Apple is the primary driver of revenue whereas EHRs are trying to collect 30% of revenue as a tax merely to integrate with them.) The $15-20K EHR integration price tag we grew to loathe is now a 30% tax on partner revenue.
At the same time EHRs try to support 3rd party applications, they are trying to build their own solutions to shut out digital health products. Every EHR now has a data warehousing solution, most have a population health tool, and many have telemedicine offerings. Similarly to Microsoft’s strategy of using its Windows and Office customer base to sell Azure cloud services, EHRs are trying to sell their own new products. The problem is none of the major EHRs are as savvy, either technically or business-wise, as Microsoft.
The Slow Burn of FHIR
Concurrently, many major EHRs are supporting FHIR as a standard way to access EHR data. Some of that support may be vaporware, but most EHRs are at least announcing FHIR capabilities. FHIR has been in discussions for several years and, despite a lot of broad-based support, is moving painfully slowly towards adoption and actual implementation. Yes, the big four may support reading data from their EHRs via FHIR, but there is very little a real application can do by only reading data from the EHR. At times, the slow pace of FHIR adoption seems intentional by the many incumbents supporting it. Despite that, a recent survey found that health system CIOs consider FHIR to be the technology with the most promise in 2018, meaning FHIR is here to stay.
Is Replacing EHRs the Path Forward?
If EHRs are failing to support 3rd-party solutions and failing to develop and sell their own digital health products fast enough, does that leave option C — replacing EHRs — as the only path forward?
As both the hub of data and hub of clinical workflow, EHRs are extremely entrenched today. The vast majority of health system IT department resources are dedicated to supporting EHRs, so there are strange incentive conflicts. And, FHIR has been highly effective at buying EHRs more time.
As I said before, this process is not happening overnight, but it is happening. More industry leaders are speaking out against EHRs and the rampant blocking of innovation so needed in healthcare. EHRs can only buy so much time. What I think we are going to see over the next 3-5 years, a time that also coincides with when a lot of health systems will have written down the heavy upfront capital expense of EHRs, is EHRs becoming less and less a clinical workflow tool and more and more a clinical encounter documentation tool and one source, not the primary hub, of clinical data.
There are two parts to that prediction. The first just means that we’re going to see purpose-built tools for specific workflows. We already see that today with clinical communication tools like Voalte that are making the lives of nurses easier by enabling them to use a fast, mobile, purpose-built platform for their workflows and communications. The second is piggybacking on the inevitable march toward continuous and complete patient data. EHR data today is episodic. Clinical data tomorrow will include episodic EHR data but also remote monitoring data, social data, genetic data, and always-on sensor data. The EHR data will be one part of a much larger puzzle. I think the EHRs of today, the large ones I mention in this post, have too much technical debt to move fast enough to serve as the hub of clinical data tomorrow. They are saddled with their history.
6 Predictions about EHRs in 2018
Predictions are easy when you don’t have specific timelines. So what do I think we’ll see with EHRs and clinical data in 2018?
- More direct patient access to EHR data. There was already a major announcement of Apple enabling its customers to access Epic and Cerner data.
- More available production FHIR services. Most of these will be couched as pilots or beta, but progress is progress. FHIR will also be used for direct patient access to data.
- Social determinants of health matter…a lot. Everybody always cites studies that show WHO data of high US healthcare spending with resultant lousy health outcomes. This is true. And sad. But it’s not the whole story. In the US, we spend far less on social programs and services than most countries that beat our health outcomes. The result is a health system that doubles as a social safety net, with all costs incurred by the health system. Social data and care based on that data will be a major shift in 2018, and I think most of that will happen outside the EHR, at least regarding data sources and storage.
- Demand for data access by EHR customers. One of the challenges when integrating with an EHR is that the default answer is usually “no”. There is invariably friction to overcome and it is up to the health system to overcome this friction. I think 2018 we will see a lot less acceptance of “no” and more pushing for EHRs to open up access to data, whether that data is exposed via FHIR or HL7 v2.x or a proprietary API middle layer like that provided by Sansoro Health.
- Mega-mergers require new strategies around data sharing from unwilling players. Large-scale mergers like Dignity/CHI and Providence-St. Joseph’s/Ascension will show vendors willing to demonstrate a more willing ability to play nice so that they don’t lose market share as hundreds of hospital sites are brought under one IT umbrella for the sake of efficiency. Epic has already shown some cards around this with their Come Together/Happy Together/Working Together trio, but it’s hard to say if it’ll stave off mega losses in the usual hospital merger shakeup where two EHR vendors enter and only one leaves.
- More ambulatory EHR providers bite the dust. Now that the dust of Meaningful Use has settled, more companies are beginning to evaluate if they want to stay at the EHR table. With Allscripts acquiring both McKesson’s EHR suite and Practice Fusion (at a loss), anticipating more contraction in the EHR space as disinterested players try to divest assets while they are still holding water (GE Centricity, Greenway) seems like a safe bet.
If you enjoyed this post, you might want to take a look at my first 2018 trend deep dive into how cash-pay practices will grow.