John Flannery

Discussing health IT with GE Healthcare CEO, John Flannery

February 22, 2017
by Gus Iversen, Editor in Chief
The health care industry has converged on Orlando, Florida, for the annual HIMSS conference. It is an event that brings together hospital executives, tech startups, medical manufacturers and physicians with a common goal: discovering the latest and greatest in health IT.

In an exclusive interview, John Flannery, CEO of GE Healthcare, gave HCB News a wide-angle perspective on what HIMSS means for his company, and how he views GE's role in the rapidly-evolving world of health IT.

HCB News: What kind of message is GE Healthcare conveying at HIMSS17?

John Flannery: One of the impressions you’ll walk away with from our HIMSS discussions — and frankly our RSNA discussions as well — is that radiology and informatics are merging together.

I certainly see a path down the road where these aren’t two separate ideas and two separate shows, and it’s really the interplay of the equipment and the data and the analytics and how those combine to produce better clinical and economic outcomes. Our equipment business is aligning digitally with our IT business, so these things are becoming one and the same.

Second, General Electric has developed an industrial IoT (Internet of Things) platform called Predix, this is a multi-billion dollar investment that we’re leveraging across the verticals inside of GE. Our particular aspect of that is GE Health Cloud.

At HIMSS we’re going to be showing about 20 different workflow and workstation products across ambulatory care, financial management, clinical operations, etc. — we have 13 apps that we’re going to show, five of which are commercially available and the rest are coming out. These apps are all about leveraging analytics for better clinical or economic outcomes.

HCB News: So these apps will exist on the GE Health Cloud?

JF: Yes, think of GE Healthcare’s Predix platform as similar to an iTunes, that’s the easiest analogy. We will be putting up a number of our own applications which we're developing now alongside several clinical and academic partners, but we envision the vast majority of these apps will come from third-party developers using the platform and bringing new analytics to our customers.

At GE Healthcare, we’re spending about $500 million over the next few years on digital. We’ve got 5,000 software engineers today and that number is growing — it’s about 10 percent of our total workforce.

With regard to our discussions at HIMSS, on Monday we announced an outcome of one of our more robust applications, DenialsIQ, which is a tool that unearthed $2 billion worth of savings for our customers.

If you visit our booth at HIMSS, you'll see we’re constantly trying to drive this to an outcome-based solution — we’re not a gadget company or a box company or a narrow software company — we’re focusing on outcomes for our customers.

The health care industry has been behind other industries in applying data and analytics to the field. What we're witnessing now is the dawn of a new era in health care, driven largely by digital-based data and analytics.

HCB News: How do the challenges of software development compare with the challenges of device development?

JF: Historically, hardware development was a fundamentally longer cycle. A lot of major changes in hardware would be two- or four-year engineering and development cycles.

The challenges were technical in nature, physics in nature; trying to solve an engineering problem with a belief that if they could produce that technical solution then the market would want that product.

Software is very much the opposite. It starts with defining the problem. What is the customer trying to solve? What information are they trying to ingest, analyze and derive conclusions from? You start off already understanding the problem, understanding the solution, understanding the value proposition.

Also, the development cycles are faster. While it takes significant investment to get the platform up and running, once you have that, it’s a rapid process of saying, "Hey I’m going to drop a denials application in here, drop a revenue cycle application in here, drop a patient contact application in here."

HCB News: In health care, would you say we're in an era where software analytics are trying to catch up with devices? Or are devices and analytics developing in equal partnership?

JF: It’s fair to say that we’ve been in a catch-up cycle on the software side as an industry, where the hardware was generating significant amounts of information that was fundamentally going unused. In that regard we’re harnessing the data and analytics capabilities that the hardware has already enabled for a number of years.

For example, with UCSF we are co-developing an application to look at lung screening images, and very quickly — through analytics and pattern recognition — the algorithm sorts through the images and identifies what's normal, so that the radiologist can focus on what's abnormal.

So you start to automate with the hardware as it is today, but ultimately these things will start to move forward together.

A lot of our software engineering research is around making the scanner image acquisition itself more automated, using AI to, maybe, position the patient differently. So there’s a lot of software development on the actual device.

I think you have to start looking at these things as a broader ecosystem, which is, could I have sophisticated scanners on a very distributed basis moving images up to the cloud? Can I process them on the cloud? Where do the reports go? You have to start looking at this whole environment together — how the equipment, the cloud, the customer, the end user and the patient will all connect.

HCB News: That ties in with the notion of these industry events sort of merging into something more singular.

JF: I think we’ll look back in a short period of time and say, “I really can’t believe that’s how we did this.” Especially on the clinical side, where people are still looking at images in a dark room and trying to determine a diagnosis.

I think we’re really at a massive inflection point and I’ve seen that in the two years that I’ve been in the business, how fast it is changing.

HCB News: What are the areas where software still has the biggest unmet potential?

JF:We are just scratching the surface of what it can mean on clinical diagnosis. Automation of reading, improving the accuracy of reading, speed, cost, the ability to provide remote access to radiology. We have lots of activities in emerging markets where they just don’t have the infrastructure for radiology.

When it comes to mobility and collaboration, I’d say we’re a little more advanced and gaining traction. That’s another area that we see evolving quickly – the ability to move images around and read them on your laptop, at home on your iPad, the ability to convene a tumor board virtually, versus having all the clinicians in the same physical place, the ability to take a scan in Seattle and shoot it to your friend in Boston to get input...

A third thing I'll mention is that there’s been big development of software on the operational side of the business (EHR, EMR) but the general consensus has been, "Great, we needed to digitize and not have paper, we needed the ability to consume information from different sources, but we haven’t turned it into a productivity tool" — that’s the next generation that we're seeing now with solutions like DenialsIQ. We're taking the inert digital input and turning it into productivity, improving departmental workflow, improving billing, and so on.

HCB News: Looking ahead to the next few months, can you clue us in on anything we can we expect to hear from GE Healthcare?

JF: On one level you’re going to hear more of the same: more clinical partnerships, we developed a digital command center for Johns Hopkins University that has tremendous traction. We know exactly what we want to do, we’re putting a huge amount of resources on it and we’re riding on the back of GE, so we’re not going to veer left or right, we’re going to be stepping on the gas even harder.

We will also continue to partner with other leading academic institutions and health systems, like UCSF and BCH, targeting key disease states, inefficient processes, and care areas, to build a library of deep learning algorithms.

Looking further ahead, by 2020 we will have hundreds of apps in the GE Health Cloud, enabling insights that will truly transform health care.