Announcer: It's Tuesday, August 7, 2007, and you're watching ARCast TV.

[music]

Ron Jacobs: Hey, welcome back to ARCast TV, this is your host Ron Jacobs. And today I am here at about 1500 meters up the slope of Mount Fuji in Japan.

I've come to Asia to learn about what's happening in architecture in perhaps one of the most important economies in the region.

Mount Fuji, of course, is perhaps one of the most sacred places in all of Japan, and it's a joy to be here with you.

So stick around, on today's episode we're going to meet Furuyama San, who is an architect MVP here in Japan. And he's been working on healthcare systems in the Japanese marketplace with some very interesting challenges. So, stick around.

Yes, I had a blast in Tokyo, and in fact I had some amazing user experience with their gadgets there. Check this out.

Today I'm at the Hilton Hotel in downtown Tokyo, and Japan has some very unique gadgets everywhere you go, and sometimes they're a very different take on certain things.

Now, I am always looking for ways of illustrating different ideas, and in user experience one of the particular challenges is always, how do you help people to wait patiently for something?

Now, think about an elevator. A lot of times when I'm in hotels in elevators, especially I like to go out for a run in the mornings, I'll have my earphones on, kind of listening to some music, and you push the button.

And usually there is like six or eight elevators in the hallway, and if you can't hear the beep that they make you find yourself kind of looking around which elevator it is going to be. So you push the button, and then you're kind of in this uncomfortable state just sort of waiting, waiting, waiting, which elevator?

The thing I noticed about the elevator here at the Tokyo Hilton is that the very moment you push the button the computer system has already decided which elevator it is going to send to your floor, and it immediately lights that light.

Now this takes away all of the question of which elevator it's going to be, you know which one, you can walk over to it, look right at it, and then when it gets there the light flashes and it beeps, which is very cool because now you know that it's here.

It's a subtle, it's a simple thing, it's probably something that every elevator in the world could do. Yet, as far as I recall this is the first elevator I've ever seen that does it.

Now, I'm sure a lot of people will say "Oh, I know other countries or places where this happens" If so, that's great. But isn't it interesting for user experience to think about the fact that telling someone what to expect, when to expect it, where to look, and where to go makes them more patient about waiting, think about that.

Hey, this is Ron Jacobs and welcome back to ARCast TV, where today I'm in Tokyo in this beautiful garden... what's the name of this?

Kazuo Furuyama: Hapoya.

Ron: And I'm joined today by Furuyama San. What's your full name?

Kazuo: My full name is Kazuo Furuyama.

Ron: Oh, OK, all right. And so today we're going to be talking a little bit with you. Now, you're an architect MVP.

Kazuo: Yes, yeah.

Ron: OK, and how long have you been doing the architecture?

Kazuo: Hard to say, maybe 25 years or more, because my age is not young.

Ron: [laughter] You look young. Wow, you've been in this business longer than I have, I'm surprised by that.

Kazuo: Yeah.

Ron: So, you know, here in Japan, this is my first time to be here, I've been very impressed with the people and businesses going on here, is there a lot of people doing.NET development in Japan?

Kazuo: Maybe, I wish. In the country all of the people go to Java.

Ron: Oh, OK.

Kazuo: But the amount of people is not small, maybe 50 percent are.NET people, I believe so.

Ron: Yeah, well that's pretty good. I think we find that in most places in the world, that it's at least half. So, tell me about the work you've been doing lately. What are you working on?

Kazuo: My main business is medical system. It's a medical total system including, how to say - community medias and including hospital total information system joined to all of our hospitals and including Japan-wide hospital system.

Ron: OK. So you are building a kind of a medical information system that can link the hospitals, kind of community-based clinics and all across Japan.

Kazuo: Yeah. It's a total system.

Ron: Yeah. Tell me a little bit about the health care system in Japan. Is it a government system that's here or there are private hospitals or both?

Kazuo: Both. Because Japanese medical rules are quite different from United States. I think including all the people join to insurance - medical insurance.

Ron: OK. So everybody gets the health care through the government.

Kazuo: Yeah. It's right.

Ron: OK.

Kazuo: About the hospital is not on government. It is a normal style, same as the United States.

Ron: OK. So private hospitals as well.

Kazuo: Yes, it's right. But all of the hospitals and the local community medias. It's controlled by only one law. I'm sorry, this kind of English is very difficult for me. I'm sorry.

Ron: [laughs] That's all right. So you mean that everything is controlled by one government law, one government organization.

Kazuo: Yes, yes.

Ron: OK. So now in United States, health care is one industry that has not been very good at taking advantage of information technology.

Kazuo: Yes, I know.

Ron: Is it the same in Japan?

Kazuo: Yes. In Japan it is more critical situation, because currently Japanese people ages are more older than in United States. Therefore every people have some problems.

Ron: Yes, all right.

Kazuo: Almost in hospital join to a lot of people now.

Ron: So a lot of people aging in the population, they need more health care.

Kazuo: Yeah.

Ron: OK. So in fact, I think that's a problem all across the world, now that in many areas the population is aging. The health care industry is under tremendous pressure. But this pressure has been good. Because it's forcing them to do something about their information technology.

Kazuo: It's right. But the most difficult point is - I believe the most difficult point is different history from medias and computer science. Because media's people need a long term information and a lot of complex information they have. For example the biggest hospitals have a lot of sections. Maybe large hospitals have 20 or 30 big sections.

Ron: They are using the different departments, yeah.

Kazuo: And different system inside. And it must to be including - integrated, this system, but it is just only one hospital.

Ron: Yeah.

Kazuo: And how to say -- one patient visit to any hospitals; therefore doctor needs to pick up to all of the information, all of the hospitals - very difficult situation.

Ron: You know, what I found when I go to see a different doctor, they don't know anything about me.

Kazuo: Yeah.

Ron: And so I have to tell them all the same thing every time I go to a different doctor. They can't see my information from the other doctor, or the other hospital. It's not integrated. The same kind of problems that is happening here.

Kazuo: Yeah. It's right. Can't we use old-style system - make solution by paper; moving to medical documents.

Ron: Yes.

Kazuo: That's... sorry. But in that case go through a paper now. It is a very big problem. Therefore, I designed another kind of solution by IT technology.

Ron: Now you were mentioning, not yesterday, but day before when we met, when we talked about a little about this, that medical records have an interesting property that they have to last a long time.

Kazuo: It's right.

Ron: If somebody's born a baby, they might live a 100 years. So you would like to have their medical records still working a 100 years from now. So you can't just put them in proprietary binary formats or things like that. What do you do?

Kazuo: We must choose a long time right for the medical document and with the result we opt for tools with very simple results. For example, take it to xml, to a PDF. We must have chosen a long time format, that's the only - this is the most important point. In this case, the system that needs a smaller the information, the information will run long time.

Ron: So the information is going to long outlive the systems that you're building. They know how to interpret it. So you could go with plain text. You said PDF or maybe XPS.

Kazuo: Of course, of course. And normal XML. But XML is not a document; it is a problem for reading.

Ron: Yeah. When you think about a 100 years from now - will that make sense?

Kazuo: In is the future, may be all of them or nothing. May be, it is not a problem. But we can at least 50 years ahead all the style file format. Very simple for IBM 360 in that you got a file. No problem is to record, no problem. In the future, it is the same situation, I believe. Therefore we chose the normal standard format, it is better for the reading for the future.

Ron: OK. As you are putting this system together then I am curious to see is it the government that has hired you to do this project or...

Kazuo: Yes. But the government to be making standards about this kind of a job is quite difficult, because government needs equalization for all of the companies and all of the technologies with the result this kind of result nothing now.

Ron: Oh, yeah.

Kazuo: For example, in United States we can find HL7 and other formats. But that's standardized formats. You cannot use the format. I cannot use that format, and this kind of problem is same in Japan. Japanese standards is the same situation. System with communication it is a no problem. But the inside of the design is well compressed and the HL7 standard has nothing.

Ron: So the standards like HL7 really don't say anything about how you build the system. They just talk about the integration between things. Yeah, I think there's a lot of complexity that you have to deal with there and I have noticed whenever I deal with hospitals or doctors that it seems like every piece of medical hardware they get, they get imaging machine or whatever, is like a silo. It is very independent. It doesn't connect to anything else and so they have difficulty bringing all of this imaging equipment and everything together into an integrated system.

Kazuo: And all of our imaging system is quite high speed evolution. Therefore, two years later, the product and format is old style.

Ron: Yeah, yeah. OK.

Kazuo: This kind of problem is little difficult point for medical system. Another big point. How to choosing the people? It is important point because I want people here, but this people in this hospital, this hospital, this hospital. In this situation, how do you find these people in another hospital? A very difficult point. This kind of solution needs the be the same as the Internet system.

Ron: So you're talking about how you identify each person when they come in. So if they go to this hospital, you say, "We've seen them before. We know who they are." That is a very big problem. We need to give people a GUID when they're born, and tattoo it on their foot or something. "This is your universal identifier for the rest of your life."

But unfortunately, we don't have that. We don't have it in the United States, either. I've talked to people in the UK. Even though they have a National Health ID, they've found that that ID is not necessarily unique. Sometimes it's incorrect; people have been assigned duplicates and problems like this over time.

Kazuo: Same as the situation in Japan. That kind of ID is not unique or, every time, is changed by the social system. Therefore we made a "cleansing" server. This cleansing server uses people's property, age, telephone number, or birthday, or address. This kind of information shows which people are the same people.

Ron: So you can use fuzzy logic to quantify people who are the same based on these criteria?

Kazuo: Exactly. It is a ultimate care system, used by the hospitals, they register them and automatically find these people in other hospitals.

Ron: One of the biggest areas of concern in medical systems is always issues of privacy. In other countries, lots of laws around privacy---

Kazuo: Ah, very difficult point. Therefore, this cleansing server has no privacy information, basically. And, this server has no name! [laughs]

Ron: Oh, all right. This is a really big problem for a lot of information systems, because you have large amounts of data. You need to sometimes look at it as a whole, but then, anytime you do, you open that privacy problem. Because whoever has access to ALL the data, can see anybody's records.

Kazuo: This kind of system cannot make a centralized database. Why? It's very simple. All the information is called privacy information because it is inside. In Japan, this is protected by law. All doctors must protect customer's information -- I'm sorry -- PATIENT'S information. This kind of information must be protected by doctors. This must be inside hospitals, not outside. The result is that these databases cannot centralize to one place.

Ron: So that brings up an interesting problem, then. If you can't have a centralized database, but yet you need to do this process to see if you have duplicates or matching them. So when you get the data from this hospital, and this hospital, and this hospital, is this centralized server calling web services or something to get the data from these others?

Kazuo: Yes, that's right. The most important point is that the global information is just personal information. Patient information must be global to the hospitals, but other information is different. For example, hospital system has an auto-ordering system. The ordering system is used by doctors. The doctor is checking this [inaudible] under all the sections.

Ron: So, some of the diagnostic information about what blood tests they did, or medicines, or a diagnosis? So they have that kind of information, which is obviously very sensitive from a privacy point of view.

Kazuo: One point is privacy, but another point is making order inside the hospitals. This is very local information; it's very proprietary. We must redesign it for proprietary information and global information. Global information is patient information.

Ron: The global information would be the name, the address and those kinds of things.

Kazuo: This would also include the analyzed reports. That kind of information is global information.

Ron: OK.

Kazuo: But when it comes to linking that all up for the people, there is the local system. This local system is very interesting. One kind of information is local. Another kind of information is shared with the society. Therefore, the system must always check the information, such as address.

We have two kinds of websites: One type of website is for the society. The other type of website is for inside the hospitals.

Ron: So you have these open, public records on websites, and then internal stuff.

Kazuo: Yes.

Ron: So when you're putting these systems in place in the hospitals, obviously, many hospitals already have existing systems that you have to integrate with. What kind of problems do you run into when you integrate?

Kazuo: The most important problem is technology compliance. Many systems in Japan have been around for a long time, maybe 40 years. So the system is quite old. How do you update this kind of system? Every time, we become confused.

Ron: So it's not only a problem of the technology. It's also a problem of the people who run the old system. Do they cause problems?

Kazuo: Yes. All of people can understand.NET technology very easily. But in the country, most of the people cannot understand.NET technology now. For example, those people using Visual Basic 6.

Ron: Yeah, a lot of applications have been written in that.

Kazuo: That application is not current. I was surprised.

Ron: It's funny because we talk about that like it's ancient. Well ten years ago, I was writing Visual Basic applications, but some people are still writing them even to this day, on a regular basis.

Kazuo: Yeah.

Ron: Well it sounds like a very large and challenging project. You've been working on this a number of years now, right?

Kazuo: Only four years.

Ron: Only four.

Kazuo: Yes.

Ron: For a lot of people, that's a very long time to work on one project.

Kazuo: Really?

Ron: Yes.

Kazuo: I think it's a very short time.

Ron: And how long do you think this will take?

Kazuo: Sorry?

Ron: How long do you think it will take for you to complete this project?

Kazuo: Maybe another 10 years.

Ron: Another 10 years. Wow.

Kazuo: Yes. There needs to be a very, wide range solution for the system. Also, I'm making applications for doctors' system. The functionality is complete and the design is perfect, I believe.

Ron: OK.

Kazuo: But this isn't just about doctors. I'm also working on a nursing system. A nursing system is much more difficult because nurses are always quite busy. How do you make the system?

Ron: You know my wife is a nurse.

Kazuo: Really?

Ron: Yes, she is. She hates the information system.

Kazuo: I know, all the time.

Ron: Like you said, they're very busy. Nurses are always going here and there. In fact, her hospital invested in putting in hardwired terminals all over the place. It seemed to me it would be better to have a wireless tablet, or something they can carry with them.

Kazuo: That's fine for large information, but the tablet PC is not small.

Ron: Yeah.

Kazuo: Therefore, Japanese nurses like PDAs. PDAs are quite small and low functionality. Now we are talking about bedside [inaudible]. All of bedside one time. You could do more better solutions, I believe so.

Ron: You do think so?

Kazuo: Yeah. We talking about this kind of solution, in Japan. That big hospital... they host it now. And, in this case one good situation is patient communication to nurse and doctor. Doctor is every time a good guy, but every people cannot talk friendly. [laughs] Therefore, if a patient afraid of a test, or if they are afraid to [inaudible], in that case is a very serious situation, but almost all doctor and nurse has no idea how that kind of patient are feeling.

Ron: So you're saying...

Kazuo: In that case, it must be picked up by a modified terminal.

Ron: Oh OK, so you're saying, actually, that if a patient was not comfortable talking to their doctor or something, they could actually send a message from their bedside terminal.

Kazuo: Yeah, and not messages only. It is making a menu, if afraid of operation or death, in that case just push the button and in that result is the most important information for doctor and nurse. Do you know nursing diagnosis? Nursing diagnosis is how to solve, how to choosing to nursing problems. This kind of information needs this kind of feeling and another viewpoint.

But, this kind of information cannot pick up now by normal communication. But, personally they can choose to [inaudible] desk. In that case, very important information enough. In that result, system automatically changes to nursing diagnosis and choosing a nursing prompt and they can keep that feeling and they can send him to a therapist or another people.

Ron: You know, I think there's something maybe cultural there that... I don't know if its true or not whether in some cultures people are happy to tell you if they want to complain about something, they will complain.

Kazuo: I understand that.

Ron: But, in other cultures, people don't feel free to complain to a doctor or to say "I'm really afraid of dying." Is that the case, do you think in Japan?

Kazuo: Yes, Japan is different. Some people cannot talking that word, therefore it is very big problem sometimes.

Ron: But that's really interesting, because what you've done is taken a technology and adapted it to the culture, really to get the right result in the end. That's a very good point and I think this is one of the reasons why sometimes people say, "Well, eventually maybe, all software development will be done in one country or another country, " and that will never happen, I don't think, because some software concepts don't cross cultures very well.

So, well, this has been very nice talking with you about the health care system and the work you're doing there. Thank you so much for joining us on ARCast today.

Kazuo: Thank you very much.

[music]

Ron: I bet you haven't thought about some of the challenges of keeping data - health care data for a long time, a lifetime perhaps. I think about data that I would like to save and how much data I've lost over the years because it was in a file format that I don't have the program for anymore, or it was stored on media that I don't - you remember floppy disk things? I used to store a lot of data on floppy disks.

In fact, I still have some and I have a hard time laying my hands on a floppy disk drive right now to read it, provided even that the disk hasn't somehow deteriorated to the point where it can't be read. It's a very, very interesting challenge to think about. How can we make data - software records - that could live for a hundred years or more? I'm sure that a hundred years from now we'll have all kinds of amazing technology, but all that amazing technology doesn't really help you too much if you don't have the floppy disk drive, or the equivalent that we'll have at that time.