The term "integration" is getting tossed around quite a bit in healthcare today. Providers want integration between medical devices and EHRs, patients want integration with various apps and IT professionals need systems to share information. Ilan Sehayek, CTO of Jitterbit, has been working toward a solution to the integration problem for nearly 30 years.
"It was obvious from the beginning of my career in 1988 that, in terms of the offerings that were out there, this was a void in the marketplace," Mr. Sehayek says. " A lot of what Jitterbit was doing at the beginning was making common and modern integrations much easier for users to consume and adopt."
Prior to founding Jitterbit in 2005, Mr. Sehayek spent time at companies that tried to tackle integration early on in their respective industries, including Amazon, which needed to design a way for thousands of merchants to rapidly integrate into its platform.
"That's where the vision around Jitterbit started," Mr. Sehayek says. "There was a need for a more agile approach. I was seeing a clear trend in the market where more cloud apps would be adopted, so there was also a need not only for integration to be quick and to solve specific solutions, but a need to also be able to integrate on-premise and cloud applications."
Mr. Sehayek spoke to Becker's Hospital Review about the various hardships involving integration in healthcare, from devices, to EHRs, to building databases for insights into cancer care.
Editor's note: Responses have been lightly edited for clarity and style.
Q: Did you always envision Jitterbit's platform would be applicable to healthcare?
Healthcare is something I got involved with once I was injured a bit. I was aware of the need for integration in the industry, but it was only about two years ago that I became aware of many of the gaps in providing quality care to patients that stem from integration issues. There are a lot of gaps in interoperability and the type of information being shared between various parties in healthcare. Also, the lack of being able to expose certain data made it really difficult to build these new, modern cloud applications in healthcare. Building a cloud application without access to a patient's medical records, without being able to intelligently know what you can or cannot recommend to a patient, is ineffective.
It became very apparent existing integration solutions wouldn't solve the problem for all the various different types of data you need to amalgamate across different EHRs, practice management systems, customer relationship management systems, etc. We now know device manufacturers, life science companies, pharmaceuticals, as well as the payers, all need the ability to interface in a more meaningful way. So with that in mind we created a vertical within our company just to focus on healthcare, and we've grown very good relationships and quite a few production pieces the new types of integration and data needed in healthcare.
Q: Was that lack of effective solutions for integration in healthcare a surprise to you?
It was a surprise to see all of the traditional systems that we come across in healthcare really revolved around making sure that a provider got paid. Many of the EHRsand what you recorded in them were very transactional — they basically were there to record all the services performed so you got paid for those services. If you look at those systems, we don't see a lot of workflows and things in place that provide guidance to what a procedure for a given patient should be; that's all in the head of the provider, and that's all shared in various different forums, but not in the systems themselves. That was a surprise because it just seemed very transactional and it doesn't seem like anybody from a system and an application perspective sat down and thought about patient's care holistically, versus giving somebody a way to record the work they did so they could get paid. We saw opportunities there.
The whole regulatory aspect of healthcare was not a surprise, and we were ready for that. That was the first thing we started to invest in. I was certainly aware of what was going on with HIPAA and protected health information. I had come across those factors just as an individual going to my own providers with all the forms I had to sign. There were some details, though, like how you can match patient records across different providers or things you are or are not allowed to use to do that joining, [I was unfamiliar with]. There are various rules around how you can or can't use Social Security numbers, for example. Normally when you deal with systems, as a technologist, you look for any data that you can use to match records, but there are some limitations in healthcare based on regulations as to what you can do, which creates some challenges.
Q: How do you envision Jitterbit's role in healthcare in the future?
One example of what we hope to do is our recent engagement with the American Society for Clinical Oncology. They have undertaken a big initiative to work toward providing much higher quality cancer care. All of the 1,500 ASCO member practices basically came together as an organization and said, 'We want to look at the quality of our data and start analyzing it internally to know what kind of patient observations and medication and treatments and outcomes we've had. Also, as a community we want to de-identify any of the protected health information included in that data and look collectively to even derive recommendations for small practices or just look at the trends in cancer care.'
To facilitate that, Jitterbit brings a very extensive set of clinical cancer patient data into CancerLinQ — ASCO's database — from various EHR systems.
We've built knowledge around how to ontologize that data, so we know how to use all the industry codes and apply it to that data even if it doesn't come with industry codes. We can get all that data into a common semantic, and once we do that it can be used in many other places beyond CancerLinQ. So, where I see us going:we're building more and more of that knowledge, we're building more and more of those solutions, which will be needed for anyone who's going to develop a modern application in healthcare.
Secondly, healthcare providers have a number of applications that must interoperate. You can go into a provider network or an integrated delivery network and they've got multiple EHRsthat all have to be able to take patient data and consolidate it. Device manufacturers require integration too. A lot of the treatments that are performed through devices aren't integrated directly into an EHR.
Clinicians have to manually open an EHR and go through to look up patient information, asking the patient along the way for answers they couldn't find to determine whether they can perform a procedure — do they have allergies? When is the last time they had a white blood cell count test? We see a way for us to make using devices much more efficient and less prone to oversight or to errors while also updating EHRs in real time.