Intel's medical ambitions

Who isn't trying to get into health care these days?

IBM has touted several supercomputing wins with drug discovery firms and created a group focusing on health care technology. Similarly, Intel is working with a research institute to see if semiconductor equipment can detect the onset of cancer. It also has devised baby monitor systems.

Anyone whose eyes have popped out reviewing a medical bill knows why. Health care costs have skyrocketed, and medicine has remained somewhat resistant to giving up pens and paper for computers--think of those paper charts in hospital rooms. The worldwide graying of the population, combined with longer life spans, means a system will have to be created that lets patients send in vital statistics remotely.

Louis Burns, who once ran Intel's desktop group, manages the company's Digital Health Group, formed earlier this year. Intel hasn't had much success to date in branching out beyond PCs and servers. In a recent interview, though, Burns laid out the case why, conceivably, you could see "Intel Inside" on a chilly medical instrument coming at you one day.

Q: Why are the costs so outrageous in health care? Is it an inefficient system? Weren't HMOs supposed to straighten this out?
Burns: There are a lot of reasons for that, and it is all over the map. What we've had to do is look at places where they have done a pretty good job of it. Kaiser (Permanente) has got a pretty interesting integrated system. Kaiser is a provider and a buyer. They have very clear motivations for improving the overall quality of care, of improving safety for patients, and in lowering the cost. Where the payer and provider are integrated, like at the VA or an NHS (National Health System), it is much easier to go about lowering the costs.

Now I'm not suggesting that the U.S. adopt a nationalized or socialized concept, but in those places (where such systems exist) they made pretty good progress in some cases.

The technology industry spent more money and time improving first-person shooter games than the outcome in the emergency room.

If you look in the U.S. at most of the non-teaching hospitals, they have a percent and a half margin. They run a pretty tight budget. There are examples above and below that, but they don't have huge amounts of money to spend. And, quite frankly, probably a more visible thing to do is buy a new MRI machine or a new ward rather than make the hospital wireless or improve the nurses' efficiency by 20 percent. So there are a bunch of issues.

Now if you look at our industry, the technology industry spent more money and time improving first-person shooter games than the outcome in the emergency room. Think about that for a second. What we're really trying to do is get our industry focused on health care issues. We think there is a very big growth opportunity for our industry. We're not doing it just to feel good about it. You have to be patient and listen to understand the issues, but we think the opportunity is huge.

The numbers associated with health care are pretty staggering. Some estimate that 15 percent of the world's Gross Domestic Product goes to health care and could reach 25 percent by 2015.
Burns: The market size is undeniable. You can't deny that millions and millions of people are living longer. That wave is coming at us. We were looking at some data on Japan. Today for every three people working, there's one retired, and it's heading toward less than two.

In the U.S., it's five people working and it's going down to three, over a longer period. Twenty-five percent of the GDP? That's not doable.

IT companies have tried to break into the health care market before, and it didn't take off like a wildfire. Is there anything that you've seen in the past that explains what went wrong?
Burns: It really comes back to the ability to drive standards: to get agreement and then let the economics work. When you get standards you get massive innovation around those standards, and then you get massive economic benefit. The attempts before have been spotty and not organized.

Now, it's not going to be easy. I've been in pre-ops and post-ops in the same building, and all the (computer) systems in pre-op are completely different than they are in post-op. I've been in post-op where there are three separate systems with three different interfaces to take the same vital signs.

Why can't the human interface be consistent? Why can't the connectivity be consistent? There are a lot of simple examples where there is unnecessary differentiation.

The Sensitron system is a great system. (Sensitron has created a PC on a medical cart that gathers and stores patient data.) They looked at the legacy system and said, "How can we do that with Bluetooth and Linux?"

They took how doctors and nurses work in the world today and then automated it, so they wouldn't have to write it down and write it down and write it down.

Are there cultural hurdles you'll have to overcome with the medical community. Doctors seem to go into these trials almost jaundiced, making comments like, "What, you want me to use a Pocket PC to conduct rounds?"
Burns: Yeah, there is a cultural aspect of this. They're highly

motivated people for the most part that are in that business because, one could argue, they feel strong about what they are doing. Oftentimes when new things are thrown at them and they don't see the benefit, they'll say, "I don't want to deal with that. I am trying to take care of patients. I am not trying to learn how to use these new systems."

Does Intel have to work with an entirely new set of equipment manufacturers, or is this the kind of market where you can team up with Dell or Hewlett-Packard?
Burns: It depends on the business. There are a lot of hospitals trying to figure out how to automate. They will buy servers and clients and stuff, and they buy those from the traditional customers. There are some new players we are not ready to talk about yet.

Some of the prototypes Intel has shown off are modifications of typical PC products like servers or tablets. Do you also think you might move toward providing chips for classic medical equipment like MRI systems?
Burns: We already do that in some systems. If they can get more standards there, it's going to be more feasible. Standardization would allow more equipment to be deployed. If you go to a community hospital, they should have access to the same things a teaching hospital does. We're talking to the bigger players in that space: GE, Philips, Siemens.

If you put the RFID in patient bracelets--now they have bar codes--you don't have to move the patient to get near it.

How do you evangelize yourself in the medical community? Are you teaming up with key hospitals or the pharmaceutical community?
Burns: We're doing all that. We've been nine months quiet. St. Louis (KC) hospital is an example of that. They have RFID (radio frequency identification) tracking, and why is it important? I don't know if you've ever been in the hospital and tried to find somebody. They should be able to tell you at the front desk, "He's down in MRI and will be back in 30 minutes, so why don't you go to the cafeteria and we will call you when he's back in his room?"

There's a guy in our office who recently kept setting off the alarms in a maternity ward by standing too close to the elevator with his new baby. The baby had an RFID tag.
Burns: Right, and that's a simple integration of technology. We run massive amounts of RFID in our factories today. If you put the RFID in patient bracelets--now they have bar codes--you don't have to move the patient to get near it.

You've shown off prototype tablets for conducting rounds. What sort of feedback are you getting from doctors and nurses with those devices?
Burns: Those are wooden models.

So they aren't real?
Burns: They are really wooden. What we did is, we went and talked with doctors and nurses at hospitals and conducted focus groups. Then we came up with something and had them look at it...Is it too big? Is the bar code scanner in the right place? Then they'll ask questions like, "What if I lose it?"

We have a new version that has an integrated keyboard that has nine or 12 symbols that medical professionals understand. We're tweaking form factors, doing things like that. The people who are going to use it are going to be an active part of designing it.

Once we get it figured out pretty good, we will build a reference design and let the industry go at it.

How far away before the product hits the market--two, three years?
Burns: It won't be that far out.

The prototypes you are showing look pretty small. Are they based around a regular Intel chip or an XScale?
Burns: You could do something with a Yonah or a Merom (two upcoming notebook chips). You want that horsepower and that capability. People want something small. But they have done some experiments with PDAs (personal digital assistants), and PDAs are too small. You don't want to just input data. It's not like UPS forms.

They also want things like voice recognition. Let's say you're a doctor and you've written an order, and a nurse captures it (on a tablet) and then says, "Dr. Kanellos, can you approve that order?" Or let's say I walk into your room, it picks up your RFID bracelet, and I get a good understanding of what is going on.

We all know the health care system in America can drive people nuts. How is it in the rest of the world?
Burns: We were at a Pacific Health Summit in Seattle that Lee Hartwell (Nobel laureate and president of the Fred Hutchinson Institute) organized. The phrase that came out of that was that talking about the U.S. health care system is like talking about my (hypothetical) brother. "I know he's a jerk, and I know I have to deal with him, but when you talk bad about him I am going to defend him." That is what you see in the U.S. health care system.

Is one better than the other? I don't know. The (British) NHS is a very a cool system in some respects. But for some reason, there is a second system that is developing in the U.K. where you can take out private insurance and pay to get to see the doctor sooner.