By Kevin Puscas, CTO, RosettaHealth
Twitter: @rosettahealth
With the incentives in recent years for creating digital healthcare data there are more opportunities than ever to share, exchange and use that data in innovative solutions. These solutions have a lot of promise including improving patient care, managing costs, and helping develop new treatments. However, even though the data may be digital, there is still a major barrier. Actually getting to and/or moving the data remain difficult. (Ok I know that the quality/computability of the data can also be an issue but that is a topic for a different time). The problem usually boils down to the fact the data lives in any number of different systems and managed by different organizations. And though there have been standards around healthcare IT for decades, the actual adoption and/or implementation of those standards had been anything but.
The usual suspects for this problem are the HIT vendors who are reluctant to make sharing data (via standards or any other means) easy. And while many are making efforts to address this, it isn’t necessarily the only problem. Another big factor is healthcare organizations themselves, who for all the right reasons have created their own Frankenstein monsters comprised of different solutions from different vendors. For example, I’ve seen single healthcare organizations using as many as 3 different EMR systems. Today most healthcare organizations rely on a set of heterogeneous systems to help deliver care, and each of these systems usually has it’s own particular take on healthcare data and what exchange of that data looks like.
Now this is usually the point where someone says ‚”But wait! Here is a simple/turnkey/proven/blah, blah solution guaranteed to solve this problem”. Unfortunately I’m not going to say that because, except for the simplest of cases, it just isn’t true. The reality is there is no magic box, or exciting new standard that’s going to solve the problem today. It is a “system of systems” problem and this means to exchange data you need to:
- Understand the different standards. Be it IHE, CommonWell, HL7, ONC-Direct. Most HIT systems now implement some kind of information exchange standard. Understanding the in’s and outs of these standards will help make smart decisions on how to best to work with each system.
- Understand what systems your working with and what they can do today. An EHR that normally relies on HL7v2 but also has the capability for IHE provides much greater options to finding a working solution. Note that sometimes these capabilities may not be obvious or well advertised.
- Be prepared to be flexible in how you integrate with different systems. It won’t be a one size fits all. Getting the same data from the same system to two different exchanges may require two entirely different approaches. Here’s where understanding the standards is critical.
- Remember that data security has to be persistent and pervasive across the whole process. Every potential solution has to be measured against its ability to be done with the right level of security.
So how do we tackle this problem? Since there are no singular solutions out there, this is a system of systems problem it will take a “solution of solutions” approach to solve. This approach will have to feature the following capabilities:
- Transport: Be able to support different transport protocols (SOAP, LLP, Direct, SFTP, …)
- Transformation: Be able to transform exchanges between different protocols. For example, taking a HL7 message coming in on a MLLP socket and transform that into a Direct message using SMTP
- Translation: Not only must a solution support transforming and bridging between different protocols, it may also need to translate the data to a form that is compatible with different protocols. This is critical as often data is in non-standard or proprietary forms within the host system.
- Routing: One of the key functions of a solution is to be able to not only provide a simple pipeline or channel for data to enter and exit, but to be able to change the flow of that data to other destinations.
- Rules: Underlying all of these features is the need to be able to define sets of rules that dictate how all of these functions take place. This is how we achieve the flexibility that is essential in supporting an evolving healthcare information landscape.
- Resiliency: Finally this “solution of solutions” has to be resilient. This means it has to not only be available 24/7 but needs to be able to intelligently handle cases where things aren’t as planned. A simple unexpected data format can’t bring the whole thing to a halt. Resiliency also means it can respond appropriately to increases in demand with out requiring often lengthy and costly “enhancement” efforts.
So what can do all of this? A lot of this sounds like traditional interface engine stuff and in many ways it is. Functions like transformation and translation is where such tools work well in most cases. Features like multi-transport protocols, rules based routing and messaging resiliency is usually found in products like enterprise service buses or message oriented middle ware.
The problem comes in trying to stitch together some combination of these types of tools, and the required server and networking infrastructure, together into a workable, reliable, cost-effective solution. So now, not only do you need a high degree of knowledge and experience in different HIT standards and implementations, but now a similar capability around a very specific set of tools and technologies. Plus, you have to be prepared for the long-term care and feeding of all of the different components (not to mention the people who support them). If this isn’t your core business is it worth the investment?
There is another potential path however. It is common knowledge nowadays the tremendous opportunities that cloud based solutions can provide to organizations. Entire solutions can now be found in the cloud for every thing from HR to Finance to Labs to complete EMR solutions. So what if you could get your solution of solutions to information exchange from the cloud as well. It would need to provide capability on-demand when needed, be flexible to incorporate new needs rapidly, and eliminate the need to obtain/retain expertise in all of the different disciplines required for HIT exchange.
This article was originally published on RosettaHealth and is republished here with permission.