Presented by Tim Kottak, CTO, Global Services, GE Healthcare
GE Healthcare is an $18 billion business that spans many healthcare solutions, including diagnostic imaging, mobile diagnostics and monitoring, IT and digital solutions, and life sciences. In the past years we have seen a shift toward value-based healthcare, and that is changing the kind of solutions that our customers need.
Today hospitals are up against changes in regulation, cost pressures, competition, and higher standards for patient safety. Because of this, GE has been able to use data and machine learning to become a partner that provides not only medical equipment but also solutions that improve outcomes. Three specific examples of this are Operational Integration, FastWorks, and Change Acceleration.
Operational Integration
This solution addresses the issue of integrating the technology we create with the operations within the hospital. It starts with co-creating with the customer to make sure that the technology is great and it is also simple and well-integrated with the user. After that, we use software that connects to and monitors the equipment to give us data which can be turned into insights and predictions.
A specific example is the MR machine, which has a big magnet inside that takes pictures to see what is going on inside your body so you can be diagnosed and treated appropriately. On the machine is a cold head which, if it does not work, could cause the magnet to clench. This would then pose safety risks to the patient and staff, and unplanned downtime that may jeopardize other patients who may be critically ill and have to reschedule their scan. The whole time, the customer (the hospital) is in a panic because it is a million dollar machine and they have no backup.
Using Operational Integration, our software is monitoring the cold head (and all other parts of the machine) 24/7. By gathering the data, we can create algorithms that can predict when the cold head needs to be serviced. Through constant monitoring, we can predict this 2-3 months in advance, so maintenance is performed well before there are any problems.
This is one example, but data is gathered, analyzed and monitored for all aspects of the MR machine using this software. If a deviation is detected, an alert is immediately sent to a GE field engineer. The engineer can then decide if the issue should be monitored, if it can be fixed remotely, or if they need to urgently dispatch someone.
The Operational Integration system has allowed us to deliver valuable outcomes to customers, including 13 percent fewer disruptions, 21 percent less time to service, and 32 percent less downtime.
FastWorks
Our FastWorks system addresses the question of how to scale quickly. It can be used to help grow our business and also to help our customers problem solve and deliver better service. FastWorks allows us to go fast methodically early in the cycle and not just have a technology mindset of “I want a prototype and I’ll figure the rest out later.”
There are five steps:
1. Problem Statement: Define the customer problem and the long-term vision to solve it.
2. Leap of Faiths: Identify assumptions that need to be true in order to achieve the vision.
3. MVPs: Build a series of tests to validate the assumptions.
4. Learning Metrics: Identify and track leading indicators to validate learnings.
5. Pivot or Preserve: Adjust the strategy based on the validated learnings.
A case in point of our FastWorks system is something that we did called MR Performance Excellence. This came about because several customers had bought equipment from us but couldn’t keep up with demand. They were turning patients away in the emergency room but because of cost pressures they were also not able to afford another MR machine. We investigated by following the FastWorks process and doing several pilots in the United States.
In the pilots, we used data from the monitoring software and also sent out nurses to see what the customer was experiencing. We then began identifying issues: the second shift was not trained properly, the referring physicians were demanding changes in MR protocol settings that weren’t needed, the third shift was understaffed, and so on. Our findings were delivered to the customer in a list of things they could do to minimize downtime, standardize protocol, and see improved results. In fact, Houston Methodist, a not-for-profit health system, saw 10-30 percent shorter MR exam times and was able to take on 3,250 additional cases per year.
The FastWorks process, therefore, allowed us to use machine utilization data to analyze how to optimize throughput and advise our customers on how they could improve outcomes.
Change Management
Our Change Management system allows us to use big data to help our customers implement changes that are better for patients and doctors. More specifically, we have used it in conjunction with our Dose Excellence Program to help doctors make changes to CT scan dosing.
In a CT scan, you get a dose of radiation to your body, but doctors say it is a necessary risk because the small amount of radiation you get is a small risk compared to the risk of not detecting and treating the cancer. However, sometimes things get out of control and patients get too high a dose. Radiologists want a good picture but that’s at the expense of increasing the radiation, so there is a tradeoff.
The Dose Excellence Program is a software that works with our CT scanner and monitors the dose of radiation given to the patient. It can provide dosage data by machine, care area, physician, referring physician, part of the body, and more. From that data, you can get automated customized reports, alert notifications, and a patient information management system that gathers detailed dosage reports for each patient. This information can help hospitals make important changes regarding radiation dosages that will keep patients safe but also get doctors the high-quality images they need.
The problem is that nothing will change unless the customer is willing to change, which is why our Change Management program takes the data to the next level to help hospitals and doctors make these important changes. We take dose management data and turn it into insights for the customer. After insights are gathered, we can determine a methodology for dose management and quality improvement, and set up a process to sustain these improvements.
Insights are then turned into outcomes for the client and it is a win/win for all involved. Radiologists get good pictures, patients are handled in a standard way, and dosage is minimized for the patient’s safety.
Operational Integration, FastWorks, and Change Acceleration are three ways that GE has used big data and computer learning to gather insights and make changes that improve healthcare. By doing so, we are forming partnerships that make our customers profitable and their patients safer.
“Across all our businesses we believe that some of these elements are part of the winning formula.”
“We want to fail early, fail often, learn from it, and then scale.”
“We want to work with customers for outcomes. A bigger vision, not that I want to sell you the latest and greatest CT. I want to help you be profitable, your patients safe in your healthcare system. So it’s starting to evolve as more of a partnership and more of an ecosystem.”