Siemens Medical's Miller on data, the W word

FRAMINGHAM (03/18/2004) - In the second part of our interview with Siemens Medical Solutions Health Services president Thomas J. Miller, he continues with our discussion of the Soarian health information system. We pick up with Miller describing the third level of Soarian's capabilities (the first two levels were discussed in part 1, in our March 16 edition). Health-IT World News editor John Dodge interviewed Miller at the annual HIMSS conference recently.

Q: Describe the third level of Soarian.

Miller: The worst people thing you can do (to patients) is something inefficient and ineffective. In healthcare, there are hundreds of examples of procedures that are wasted every day. Say your PSA goes up. You'll have the great pleasure of experiencing prostate biopsy and getting poked where it is not pleasant multiple times, often in the search for cancer thatís not there at all.

You have a stenosis in a coronary artery. You get angioplasty, but you may be one of those patients programmed to re-stenos. Or we think you are pre-programmed to re-stenos, so we give you a medicated stent. We know that medications have nonresponder rates of 20 percent to 30 percent. Think of it: 20 percent of all medications being given do nothing.

Two bad things happen. Number one, you're wasting time when medications should be effective, and two, it's a total waste. The more interesting thing is: How do you attack this problem?

You better have a system that captures information across the enterprise in a holistic way and analyzes whether seemingly disparate pieces of data are actually related (so) I can tell what you are doing is correct or incorrect to the patient.

One of my favorite examples is Alzheimer's disease. We know you can predict like clockwork the steep increase of Alzheimer's over the next 20 years. It's a nasty disease because if you don't catch it early, you have dementia for a long period of time. There are huge amounts of cost, huge amounts of morbidity. It's just terrible.

There's no genetic test for Alzheimer's except to explain 1 percent of it. Your hypothalamus shrinks as you get older. It shrinks differently if you're a woman or man. How do I tease all these pieces of information together to tell you if you're forgetting your keys in the morning because you're just forgetting your keys or you have something you should worry about and medicate early?

This to me is the real interesting piece of IT. We have gotten to the point where the information flood in healthcare cannot be handled by the brain. The human genome has been encoded. Genes are responsible for 50,000 proteins. You can't just understand each because they interact with each other. It's a 50,000 factorial problem.

Q: Is Soarian a shipping product?

Miller: The product being shipped attacks the bottom rung of the hierarchy.

Let's be honest. This is a hard thing to do. It's going to take many years. By the way, healthcare-IT is not simply a product. It's a partnership. On average, people keep their IT system over 10 years, and the system they have at the end of 10 years is not the system they bought in year one.

If you don't have a vision where you want to go over a 10-year period, you are pre-programmed to buy something that's obsolete and something you'll have to amortize over a short period of time. We are ready to begin shipping the first elements of the workflow engine in the next couple months. So it's not like we're years away. We are months away.

Q: What's shipping now?

Miller: Shipping now is the enterprise IT system that automates existing processes. It's all level-one stuff, which is what the majority of everyone on this (HIMSS show) floor is doing. Workflow engine is stage two. The knowledge engine is stage three.

Q: Does Soarian mean I have to do one of those megamillion-dollar deals with Siemens, or can I start small?

Miller: It would bring you to your knees if we shipped it all at once.

Q: What kind of investment does it require?

Miller: It depends on size and on how deeply and how widely you want to deploy it. We will have a version of an enterprise system that starts at $200,000 and scales up to a couple hundred million.

Q: So it won't be for small medical practices?

Miller: We have ambulatory products as well, but that's not what I just talked about (with Soarian). But you have to connect ambulatory with tertiary care setting. 80 percent of healthcare is done in the ambulatory setting.

Q: How much work is left to do on Soarian?

Miller: As I said, it is the largest, most expensive thing we've done in our history. I will tell you I imagine we will be developing Soarian 10 years from now.

Q: What does it compete with?

Miller: How dare you even ask that question? It's almost an insult. The question should be rephrased, 'What is attempting to compete with us?' Frankly, since we've started to talk about workflow, you'll find the W word mentioned on the booths of everyone here. If you go back a couple of years, we were the only ones mentioning that word. (Workflow) is a hard thing to do well.

Q: What's the hardest part of Soarian development?

Miller: You need to design the system upfront to do this. If you try to bolt it on to an old architecture, it will fail. It has to touch every piece of the software and be deeply embedded. If you have a 10-year-old architecture and are now going to do workflow, you don't have a chance. That's what weíve been spending time on since (Soarian development started in) 2000.

Q: How do all the devices from Siemens such as MRI machines roll up into Soarian? And does Soarian include all the subspecialties in medicine?

Miller: It is a single architecture covering all of that stuff. Almost no institution doesn't have legacy data. You'd better be able to interface with legacy data. It cannot remain an island. To get full power out of Soarian, you deploy it everywhere to do everything.

Q: How many customers do you have for it so far?

Miller: If you define customers as people who have placed orders, over 100. They span under 100 beds to thousands.

Join the newsletter!

Error: Please check your email address.

More about IslandPSASiemens

Show Comments
[]