A. Neil Pappalardo, “MEDITECH: Using Communication Technology to Eliminate the Middleman in Medicine” - MIT MechE Symposium: Mechanical Engineering and the Information Age

[MUSIC PLAYING]

PRESENTER: Pappalardo has asked me not to introduce him because he needs no introduction. He is a great benefactor of MIT, and his company, Medical Information Technologies-- you may understand that there is some relation to the initials of that. It is M-- Meditech.

PAPPALARDO: Right, they call us Meditech rather than what I prefer they call us.

PRESENTER: But Neil has very kindly agreed to talk to us tonight about Medical Information Technologies.

PAPPALARDO: I'll start very simply. I first and foremost have nothing to do with university teaching. I have nothing to do with university research. Therefore, I am completely unbiased, and everything that was talked today was spent basically about that subject, and I don't know that much about that subject. But what I do know is about software technology, or what is generally referred to today as information technology.

I started a software company called Medical Information Technology-- too big a mouthful. They chose to call us Meditech rather than MIT. That's fine. That's fine. I'll go with Meditech. It's a nice abbreviation. Perhaps the earliest software company in America, and therefore the world, started in 1968. I founded it. I continue to be the CEO of it. Couple of thousand employees. Private company, not a public company. Private company.

We make software strictly for hospitals. 98% of it is involved with the clinical process of delivering therapeutic care. A few percent is the administrative stuff that I get forced into doing for hospitals. Basically it's clinical information systems. We're very, very good at making software. Our existence, our life of 32 years attests to that.

We really know how to collect information-- clinical information. We really know how to store clinical information. We really know how to retrieve clinical information. Forbes Dewey made a comment earlier-- one of his talks about five megabytes for some sort of image. We need five megabytes to store 200,000 discrete pieces of information-- quite often numeric, quite often qualitative information, and certainly quantitative information-- about patients.

It's a big challenge. It's easier to put an image stored over there and retrieve it. It's very hard-- extremely hard to extract from 200,000 discrete pieces of information that are stored about a patient over a number of years to extract the right information, present it to the right person at the right time for them to make a clinical decision. That information is highly structured information, but it's extremely important to be able to get your hands on the piece or pieces of information that are needed here and now to effect therapeutic care.

We're very good at doing that. We've been doing that for 32 years. I like to believe we could do it for another 32 years. I myself am the principal architect of the company, in addition to be the founder and CEO, and I run the business, but I've got 10 officers who really run the business. And I can continue doing what I've always done, which is making software.

The software we make-- first and foremost, it's integrated clinical software. And I'll explain, show you a little bit about the integration concept. What we try to do is empower hospitals. The reason for the empowerment of hospitals is because in a community they are the major use of medical technology. That is the center of clinical technology in the community. It's the central point at which we would expect the most use of technology to improve diagnostic procedures to tell us what's wrong with our patients, and, just as important, therapeutic procedures to make our patients better.

And that's what hospitals are. They have a tradition of being high tech, and what we try and do is empower them so they become the focus of medical care for the community. They become the focus. They become the area where we would expect the proper use of not only machines like MRIs and things like that, which we think of as high tech machines, but also information systems. They become the focus for the whole community. We would expect the hospital is where you would go if there was a serious problem. Sure, we all might have primary physicians in a doctor's office that may be able to diagnose acute problems, may be able to diagnose chronic problems, but in all cases we feel comfortable that the hospital was there to cater to our needs in a very, very high tech situation.

What we try to do is orchestrate the clinical process. Orchestration is a great word. We think of an orchestra. We go to the symphony. The advantage the conductor has in a symphony is very simple. He gets to look at the 50 people that are playing the music, and those 50 people get to look at the conductor eye to eye. The conductor can orchestrate them to work in a way to produce great music.

In a hospital, we'd like to think that your surgeon, or your intern, or a resident is thinking only of you. There's no one else when you have a critical problem that they're dealing with, or worse yet, a family member of yours. We'd like to think the physician is only thinking of you. The physician is looking at you 24 hours a day and vice versa. You can see them-- not only the physician, the other millions that help-- the lab techs, the nurses, everybody involved in the clinical process.

But in reality, no one's looking at you very often, certainly not the physician. At best, he might look at you in the eye if you're a patient-- I don't know-- five minutes a day . There are literally 25 other people 24 hours a day that are performing diagnostic and therapeutic procedures on you-- and by the way, unlike the symphony, they're doing that for 25 other people. Each person in the hospital at the time you're there is doing procedures for you and procedures for 24 other people simultaneously and at different points in time even.

You may be in a preoperative stage. There may be someone with the exact same problem as you who's in a post-operative stage, and the people that have to do work for you quite often are not looking you in the eye. They're dealing with the numerous other people simultaneously. Now, if symphonies were conducted that way, where the musicians couldn't see the conductor and the conductor couldn't see the musicians, and the musicians were playing not only the music you're hearing, but 24 other pieces simultaneously, the music wouldn't sound so well.

And it's even worse. At least with the symphony the author of the symphony has written the score from beginning to end and placed the score in front of the musician and as well in front of the conductor. And the conductor's problem is to simply lead that process through. In the hospital setting, and, in fact, in the individual doctor's office, music isn't performed that way. Instead, the conductor picks one note, and he gives it to one musician, and says play that note. And after that note is played, he picks another note, and hands it to another musician, and says, play that note. And then, of course, picks another note for totally another piece and gives it to someone over there.

Now, of course, I hate to admit this is about the worst possible process that we could ever imagine human beings being involved with, and yet we all expect a successful-- I don't know-- procedure being done when we go to a hospital. We really accept that. Thank you.

Now, we're good at communicating, storing, and retrieving information. Well, let's get some examples. My talk, of course, has to do with eliminating the middle man. The reason the middle man has to be eliminated is the middle man is the source of error in procedures. A recent study was done indicating that we kill about 100,000 Americans each year through medication errors, and the primary reason we kill 100,000 people a year with medication errors is there's too many middlemen in the process.

A physician orders something. The likelihood of that order being properly conveyed to a pharmacy is 50/50. The likelihood of the pharmacy filling it out and sending the medication back to the floor, 50/50. The likelihood of the nurse grabbing the medication and giving it to the right person, another 50/50. And you compound all these things, it's no wonder we kill 100,000 people. Having said that--

AUDIENCE: [INAUDIBLE].

PAPPALARDO: Having said that, the importance is to eliminate the middleman. I'll give a couple examples of eliminating the middleman. My first example is patient education. We expect the nurse to actually educate patients and patient families about what's going on, because generally physicians have no time to do that, so it's treated as a don't worry about it, you're in good hands. I'll take care of you. Don't ask too many questions, and we'll get through this all right.

We obviously believe in empowering patients. I put down here pulling information and pushing information. For 30 years and as a continuation, we believe the importance is to pull information. People use the internet and browsers as if these are the epitome of information. We, of course, suffer from a total information glut.

So if we are to provide patient education in a pull form, it would look like this. Here's a browser. It's an internet browser. It's looking at patient education that Meditech provides. Just grabbed our product and link to it. We can look at procedures done on the brain-- strokes. We can look at diabetes, type 1 and 2. We can look at heart things-- various types of information that a patient would want to know. OB/GYN, lungs, surgical procedures-- this is the classic pull. We rely on the individual-- in this case, the patient or patient's family to pull the information to them that they need to know.

Well, how do they know what information? There's so much information out there. How would they possibly know? But of course, it's kind of neat. If I look at-- let's take a surgery. Let's take a gallbladder removal.

This is what we would present. I'll read this. We're talking about a gallbladder. I don't know if anyone's ever experienced a gallbladder operation. This is the part of the--

[VIDEO PLAYBACK]

- The gallbladder is a small, pear-shaped organ that sits under the liver in the upper right section of your abdomen. Its job is to store bile, a fluid your liver makes, to help digest food. As needed, the liver and gallbladder send bile through ducts or passageways into your intestine.

Sometimes bile hardens and stones form inside the gallbladder. These gallstones can lie quietly and not cause any problems. However, if gallstones get stuck in a bile duct or cause inflammation of the gallbladder, your gallbladder may have to be removed. This procedure is called cholecystectomy. Because the gallbladder is just a storage bin, the bile your liver makes simply goes directly into your intestine.

[END PLAYBACK]

PAPPALARDO: What we clearly try to do is eliminate the nurse or the doctor from telling you about stuff like this, because they don't have the time to tell you stuff like this.

[VIDEO PLAYBACK]

- The surgery to remove your gallbladder is called cholecystectomy. During laparoscopic cholecystectomy the surgeon inserts the laparoscope through a small incision near your navel and finds the gallbladder.

We use animation so that the blood doesn't spurt out.

- --just below your ribs. Blunt instrument helps grasp the gallbladder. Another instrument carefully closes off the gallbladder duct and artery. And an instrument such as a laser detaches the gallbladder. The surgeon then guides the gallbladder toward the incision, where it is emptied and slipped out. The entire surgery usually takes from one to three hours.

[END PLAYBACK]

PAPPALARDO: Let me exit this and hop to something else-- lungs, asthma for kids. Of course, what I'm showing you is something that expects the user, the patient, the patient's family to know what they're supposed to look up.

[VIDEO PLAYBACK]

- Your chest may feel tight. You might cough a lot, and your lungs may feel squeaky or wheezy. What is happening inside your body? Asthma causes the breathing tubes in your lungs to close up and get clogged with a lot of mucus, sort of like your nose when you have a cold. If your breathing tubes close up too much and make too much mucus, air has a hard time getting in and out of your lungs. Why do you have asthma? Asthma--

[END PLAYBACK]

PAPPALARDO: I'll stop this, point out the condition is the second choice. It tells you about what's wrong with you. Other things-- how do we treat that? How to use an inhaler-- if I went to that one, presumably it would teach me the proper way to use inhalers.

I won't let these things go on. You can spend a half an hour learning about things-- but simply to point out that we're big on patient education, to eliminate the middleman, freeing up nurses to give you more direct care, trying to answer as much questions. We break down education-- that's the second time I hit this button here. There'll be a slight pause while this warms up-- the on/off switch. It's not my display. There it is. That was quick.

Treatments-- self care-- what you, yourself, should do for yourself. I don't know how to make that go away. It'll probably go away when I press another button, or it'll go away when I pause, or it'll go away when I hit-- it went away by itself. Modern technology-- indistinguishable from medicine. On the pull technology, we don't know exactly what your physician wants you to do. We don't know what test the physician might want to do for you, so we educate you about all possible tests for this thing here.

We don't know what buzzwords we should educate you on to understand various-- I'm just zipping through here-- buzzwords de jure of all the different things they say around. There's a sound here to pronounce these words, and I've sort of suppressed that.

The point is this is pull technology. We present to you a full menu of all possibilities. The classic argument is too much information is presented to you. You, as an individual, don't know what to do. We contrast this with the pull version of the same information. Sorry, I lied-- the push information of the same information. That was a demonstration of pull technology.

You can go and look at a vast repertoire of knowledge if you know what you should be looking for. The push version is very different. The push-- try again. Push version is--

AUDIENCE: You didn't press hard enough.

PAPPALARDO: Excellent. I'll use that as my next joke. Talking about a joke, I have to interrupt for a joke. [? Naum ?] was talking about-- which blew me away. 80% of these Asian-Americans aren't going into mechanical engineering, or something like that.

AUDIENCE: I was thinking maybe you would joke about the gallbladder.

[LAUGHTER]

PAPPALARDO: Well, that, too. But in reality, I'm a full blooded Sicilian. All I can find is 80% of Sicilians go into crime, and I was thinking only a few of us go into what we would consider honorable businesses.

What happens is an email is sent to the patient-- an email containing a link to this first page and instructions on what the password de jure for that patient is to enter. I was told the password de jure for me was neil.neil. That's my name. It's certainly not a Sicilian name. Same kind of thing-- patient education.

There's a big difference, though. My care-- I am presumably going to the hospital for heart failure. Heart failure doesn't mean your heart stopped beating, by the way. It's a clinical term. It's troublesome. It doesn't work the way it's supposed to work, but it doesn't mean it died, and [? you ?] died. I also-- because I happen to have diabetes. It's a chronic disease that's also there.

And there are other important items. I no longer will see the general information associated with all the clinical information. I will see information that's specific to me. My physician has pushed this information to me from the vast collection of information out there-- chronic heart failure. It goes on, tells you a little bit about chronic heart failure. The point is now we're using a push technology through the email. We're emailing to the patient or patient's family specific information that they have to know for their visit to the hospital.

This is how we're going to treat your chronic heart failure. These are the drugs we're going to give you-- the specific drugs. These are specific tests we're going to perform on you when you're at the hospital. So we've taken a whole vast amount of information-- by the way, the voice is simply a voice over. So if you can read, you can know what they're talking about.

If I look here--

[VIDEO PLAYBACK]

- --including eating less salt. Rest often.

[END PLAYBACK] Back to viewer care-- specifically the drugs that you're going to be given while you're at the hospital. You're going to be given Coumadin. This is the actual monograph-- exactly what it's there for. I'm just going to page through some of this information.

The point is, by using push technology, we take the vast amount of information that's possibly out there and stylized it specifically what we want you to know. That's an example in patient education. I won't belabor it anymore.

The self care are the exact procedures that you're supposed to do. Thank you.

AUDIENCE: That's push technology.

AUDIENCE: No, it's pull. It's pull.

PAPPALARDO: He told me that the first time he stood up it was 10 minutes. The second time he stood up it's 20 minutes, and the third time he'll stand up would be 30 minutes.

AUDIENCE: But Neil asked me to do that, so--

[LAUGHTER]

PAPPALARDO: That's an example of pushing information versus pulling information. Browsers-- the epitome of pulling information. Email-- the epitome of pushing information. Shows how we can educate patient, patient's family without a middleman. For the woman here, I could have said a middlewoman, but I think women would prefer the middleman be eliminated.

Doctor's orders-- pulling information-- well, we've been very good for many, many years of being able to let physicians or clinicians pull information. Wow. Such a big deal. Stupid. It's a bad idea.

How do I know that I'm supposed to look at information? I'm a physician. How do I know it's time to look at one of these patients? I don't know how I know. Assuming I select him, I can obviously have some buttons here. I can look at patients where I physically am, or I can look at my own patients, but in all cases, how do I know what patient I'm supposed to be looking at information for?

So I'll guess that this guy, Fred Jones, got a lot of-- whoop. I've got some demographic here. We're very good. We give very simple buttons around the peripheral to let the clinician easily select what they want to look at, but I ask you-- how do they know what it is they want to look at? Who told them to go look at Fred Jones? And in particular, if they told him to look at Fred Jones, who told him to look at what information about Fred Jones? Should I look at the flow sheet? Should I look at the chemistry? Should I look at blood bank? Medications?

Based on that, should I look at that information in items by time? Should I look at times by item? Should I look at more information? Should I look at less information? Who knows? And yet we're very good at putting out simple buttons to let people look at information. There's a little scroll bar here, so I'll give you one without a scroll bar.

Who knows what to look at? This is the classic pull. We all think the epitome of computer technology is the browser. I submit to you the browser is worthless in an information glut society. What possible mechanism there is to know what you should look at?

Well, I compare this one. This is simply numeric data, but we can look at it in graphs and every which way, but the point is, how do we know what to look at? Let me deep six this one and go to the next page, which is the push version. Now, the push version means we can take any data that we're so good at storing, and we can wrap it around with HTML type stuff to take any display we've ever been able to do with any elements. We can wrap it up in HTML stuff, but what good is it to look at if you don't know what you want to look at?

So what we choose to do instead is to take the right data, wrap it up-- both diagnostic and therapeutic data-- and email it to the physician in charge. Very simple concept. I'll give some examples of it. Here, I really should be looking at my front screen of my email, which gives me messages, because I am assuming those four messages-- something about pain, something about drugs, something about x-rays, about various patients-- came to me in my email system. You know the you have mail? Someone wants my attention, and I pick the first one.

55-year-old female John-- Jane Doe. She weighs 80 kilograms. I know this information was initiated by a nurse because the first chunk of data is the qualitative information that the nurse is the one who initiated this email. Your patient and presumably Dr. Pappalardo-- your patient has been complaining of increased pain-- because I happen to know this particular patient-- came in for pneumonia. Chronic condition of arthritis of the knee has been causing her increased pain. We've currently have her on a current pain medication. That's existing data. I didn't have to go search for this data. It came to me in one email message-- the crisp information that I want to know.

Patient's complaining about pain. This is the pain medication I'm currently on. Here's my order set. What do I want to do about this? I can continue treating with this drug. I can treat with that drug. I could treat with this drug.

AUDIENCE: Can you do two at the same time?

PAPPALARDO: I'm a physician. I can do anything the hell I want to do.

[LAUGHTER]

AUDIENCE: [INAUDIBLE].

PAPPALARDO: Should. And we have pharmacists involved. The pharmacist's job is to keep the physician under control. The order is always sent to the pharmacist whose job it is check for medication, problems with other medication, drug-- whole bunch of stuff. So no, it doesn't get by the pharmacist.

AUDIENCE: It's not legal with the medical health plan.

PAPPALARDO: When it's paid for?

AUDIENCE: Yeah.

PAPPALARDO: No, not at this level. Not at this level. I can send also a qualitative measurement back if I want-- a message back. Tell me how she feels lighter, blah, blah, blah, blah, whatever I want it to say. You have a case. This is an attachment to an email. I submit it. I'm done.

That information is sent back to a computer. Computer would deal with this [? mess. ?] We have intelligent email systems to deal with things of that nature.

Next example. This is not from a human being. There's no signature here. This is from a computer. A computer sent me this email. Hospital policy is, after I give a narcotic, 72 hours later that narcotic will expire. I have to renew it. It's a computer generated email specific to me, telling me exactly what I want to know, giving me the opportunity to do whatever I want without having to search around the neighborhood for whatever it is I'm supposed to do. Same argument.

Could decrease the dosage with or without a message, submit it, and done. This is presumably a patient who came in for total hip arthroplasty. Do a hip surgery. We move the hip. Put it on her. She's in pain. She's in serious pain and needs some really hard drugs to deal with it. I'd submit it, and again go on to the next thing.

Another email-- x-ray data. Here, I-- it's the primary physician-- have a patient-- outpatient. He's here because I suspect smoker. I suspect cancer in the lungs. I've asked for an x-ray. I've ordered it. I want it interpreted by the radiologist. I, as a primary physician, wouldn't know how to interpret an x-ray if my life depended on it. Thank god patients don't know that.

Dr. Lemke here presumably has said, I want you to reorder this. It's got some shadows in it. I can't tell if the spots are real or the spots are shadows. Wants me, as the primary physician, to reorder it. I'm sure if I look down later I'll see how to order the x-ray again.

I can choose the position. These are just different ways they look at x-rays. I can tell whether stat means do it immediately. Routine is when you get around to doing it. Once again, check what I want, add a message, submit it. I'm done.

Once again, I have pushed the on/off button. Modern technology-- why is there an on/off button?

AUDIENCE: If something goes wrong, do the lawyers come after you, or do they go after the doctor?

PAPPALARDO: Boy, that's a very interesting question. Being Sicilian, I have a very simple answer for it. Do I have to embellish that answer anymore, or is that it?

[LAUGHTER]

We, by the way, are simply the providers of information. The ultimate decisions are made by physicians. Physicians don't in any way delegate their responsibility to someone else. So we've been in business 32 years. We've never been sued, let alone lost a suit on issue of that nature.

One more example. Here's an example, which is the wave of the future. In the past, mistakes are made because we have a big feedback loop. I don't know if you know how medicine works. We measure things about you. Based on what we measure, we change therapeutic procedures.

Here's a patient, John Doe. 45-year-old male came in. Fundamental problem is a blood clot in the leg. Well, simple way we treat blood clots is we give them a medication, which should increase the time it takes to coagulate blood. Medication is heparin. We give him this, and, of course, we want very much to increase the time it takes to clot blood.

So we have a test. It's called PTT. If anyone wants to know what it means, it's up on top there-- partial thromboplastin time. I like to call it the time it takes for your blood to clot.

This patient came in. The first thing we did is gave them a major-- inject amount of heparin-- major amount, and at the same time, ordered a PTT test. But it's now come back-- is the result from that PTT test. 124 seconds-- that really says it took 124 seconds for your blood clot. Presumably, it took 12 seconds for the blood to clot when they had-- because it was clotting too fast, and that could have caused that. So it came back to 124 seconds.

Well, what we submit to the physician is you should not in health care go back to the old fashioned method of I give you a drug. I order a drug for you. I order a test for you, and when the test comes back-- the PTT-- I look at it-- possibly in real time, possibly when I get around to it-- and based on that value, modify the drug. Increase the dosage if we're not clotting-- we're clotting too fast, or decrease the dosage if we're clotting too much. If I said it right or wrong, I don't know, but it doesn't matter.

That feedback loop, which is continual, does the same thing that one of the speakers said today. It creates immense overshoots and under shoots, but who cares? Human beings-- we assume their physiological system can deal with that. Anyways, in this particular case, we said, rather than involve 30 distinct orders of therapeutic procedures and diagnostic procedures being done over time, I will order one thing for you. I will order a protocol, but since I'm responsible for you for a patient, I want to be kept abreast.

And here, I'm getting a result back through my email again. The latest measurement of the PTT test was 124 seconds. We are currently dripping heparin into your body. That's 30 minutes. I am just about ready to finish. We're dripping heparin into your body at 1,350 units of whatever you want to call them per hour. The automatic protocol says, if you're at 124 seconds, somewhere in here we're supposed to-- it's kind of high. We're supposed to hold the drip. We've obviously overshot.

When the patient came in, you can believe they were clotting very fast up here. We overshot our mark, the no change one, the desired goal. We're going to hold the drip by 60 minutes. Automatic protocol-- and wait an hour and then decrease the drip by 300. So it'll be dropped to 1,050. You, as a physician, should be aware of this.

You can do either of two things. You can ignore everything we've sent. You make believe it never happened. You can order a PTT test right now. You could actually change the protocol, which I've chosen not to do, but in either case-- and then you do that in this case by telling them change the protocol, because the protocol is what you originally ordered, and we want to make sure you would document changing from that protocol.

But here's an example of an overshoot due only to the rough measurement of that algorithm. The algorithm is based on the weight of the patient, and it shouldn't be in such things like 100 or 200. We can do with the actual amount. We can change things like that.

That completes what I want to say to you. Let me summarize again. All aspects of medical care involve middlemen. We want to obviously remove them from the process. They simply introduce errors.

We have very good algorithms. We would prefer ordering protocols, rather than individually ordering therapeutic and diagnostic procedures in some apparent way, but in reality a haphazard way, which quite often introduces overshoot and undershoot. We obviously want to make the physician in charge of the patient acutely aware of the information they should know about the patient without having to wander through a browser or a pull type technology to find it.

That's the end of my comments. I'm more than happy to--

[APPLAUSE]

If there's any short questions, because I know it's late, and people want to go home, and-- yes?

AUDIENCE: How do you back up the information, and who does it?

PAPPALARDO: And who does what? Pick up the information?

AUDIENCE: Back up the information.

PAPPALARDO: I missed the question.

AUDIENCE: Back up.

AUDIENCE: Back up.

AUDIENCE: Back up the information.

PAPPALARDO: How do we back up the information?

AUDIENCE: Yes. [INAUDIBLE].

PAPPALARDO: Well, computers actually work for years at a time today. We use very thin server technology, which is extremely highly reliable, and we constantly back up information because it's distributed information through an institution. Remember, I said before we'd like to empower the hospital to be the primary reservoir of information for the community. We've got something like in the Houston area-- medical information and about 300,000 patients. That's accessible in any medical institution in the Houston area. Yes?

AUDIENCE: There was a contrast I recently became aware of the simulators for health care education. How do you view those in contrast to--

PAPPALARDO: We don't-- simulators are presumably to help educate the clinicians.

AUDIENCE: Right.

PAPPALARDO: Not my job. I'm only interested in educating patients. I assume the clinicians already start out educated, so I would like to believe that I can help them do their job more effectively with less errors by providing them with the right information at the right time to do their job. I don't educate physicians. Yes?

AUDIENCE: How many hospitals are using this?

PAPPALARDO: How many using this? At any point in time-- I currently have 1,500 active hospital accounts. At any point in time, there is my prior technology, my current technology, my next technology. This, by the way, is my current technology. We started selling this kind of stuff six months ago.

We probably have today a dozen hospitals using this out of the 1,500. If you asked me the question tomorrow, I'll just add another zero to the dozen. We've been fortunate that we've had 1,500 active hospital accounts, since, for 32 years, we've lost under 20 of them. So we constantly upgrade and continue making our hospitals--

AUDIENCE: Do they tend to be more in urban areas where the patients are more educated?

PAPPALARDO: The type of hospital I love are the community hospitals. The type of hospitals I hate are the major teaching hospitals, especially places like Boston. I used to work in Mass General Hospital. I would never sell anything to Mass General. I was in the university business. I'd never sell anything to MIT either.

AUDIENCE: What about the impact of HMOs? You talk about quality of care--

PAPPALARDO: Do you mean by HMO an insurance company? Is that what you mean by an HMO? Because that's what an HMO-- is an insurance company.

AUDIENCE: The primary care is driven by cost, as opposed to quality, perhaps.

PAPPALARDO: Well, what I submit to you, especially the push technology has an intrinsic benefit on cost. The current way to make medical care more cost effective is to lay off nurses and cut back the size of the staff, because people are 80% of your cost.

AUDIENCE: Well, the alternative is that you have successive layers of intervention and approval you have to go through.

PAPPALARDO: But I still say the way to reduce costs right now is to cut back staff and don't do things, because you cut back the staff, of course they don't do anything. We can't do it because you cut back the staff to do it. The first way, if you were a business, you want to control your costs-- first order effect-- lay off people.

The trouble with laying off people-- it decreases the quality of care. Notwithstanding, it's the people that caused the errors in the first place, but I'll still submit it decreases the quality of care. What we're trying to do is eliminate the middleman, eliminate the need for people, and at the same time improve the quality of care, because we're making the process more efficient. That's what we're trying to do. Yes, over there?

AUDIENCE: We just had a new drug approved, and I'd like to put it first on your list. What would you charge me?

PAPPALARDO: You'd like to put it first on my list. Well, that list there for the drug one is this physician's favorite pain medications-- that one there. This one here.

AUDIENCE: [INAUDIBLE].

PAPPALARDO: This is that physician, whoever that physician is. Dr. Pappalardo-- this is what I tend to use. It's my favorite set of pain medications. If you, a physician, wishes your own set, of course that's what you would see-- your own set.

AUDIENCE: So you interview me, or you watch what I've--

PAPPALARDO: We do the simple thing. We give you the hospital standard. We let you, through your minions-- because god forbid we would let you modify your own set of orders on a computer-- through your minions, we would allow you to change to your favorite order set. Yes?

AUDIENCE: I have two questions.

PAPPALARDO: I only have one answer. Go ahead.

AUDIENCE: The first one is what's the level of competition for your business here? Are you--

PAPPALARDO: Well, we've got 1,500 active hospitals in the US and Canada. There's about 4,500 total. So we've got a third of the market.

AUDIENCE: So there are other products that give you the same type of service?

PAPPALARDO: You're asking the wrong person. Can I ever admit that someone else could do as good a job as we could do? I would say this, though. I'm one of the old timers in the medical information business. My reputation is such that anybody in medical information systems generally knows who I am. Everyone else who is either in the business is either if they're old, they're dead. And if they're young, they're young Turks. So I'm still alive, all right?

[LAUGHTER]

I'd like to believe that I have pushed the state of the art continually in this business. My life cycle, being 32 years old in the business, attests to that. This is obviously something that the physicians are a little concerned about. We're in a sense sort of telling them what you should look at, what you should choose from-- a shortlist. We're constraining freedom of choice. Going to McDonald's and ordering off a menu is constraining freedom of choice. Physicians would like to be held in a somewhat higher esteem than being chosen-- reducing freedom of choice.

AUDIENCE: Are you prepared to take the next step, like pushing information on the products? Your site is an excellent site for physicians to get pharmaceutical information.

PAPPALARDO: I don't educate physicians. I've said that earlier, and I'll say it again. It's not my job. What I want to do is make physicians deliver higher quality care at a lower price. That's all I want. Better care, cheaper price. Yes?

AUDIENCE: What are the key software tools that you've used to develop the impressive architecture behind all this?

PAPPALARDO: This is all trivial. Everything we do is trivial. I said what we're very good at is taking a collection of data for a patient and organizing it in a such a way-- constantly appending. We don't erase it. We always append to the end of the file.

AUDIENCE: Well, my comment there is you've got a lot of data. How do you manage it all?

PAPPALARDO: Well, that's my business. It's what I do. I don't know how I do it. It's what I do. I don't know. To me, it's trivial. We're good.

I said earlier we're excellent at gathering data, storing it away, and allowing ourselves to retrieve subsets of it. I don't consider this rocket science. What you heard most do today is rocket science. This is as trivial as it can. I submit to you-- I mean, it's so much easier. I said it takes some data out of the person's file. I wrap some HTML around it, and I sent it to them in an email attachment.

Email is so-- I mean, how many people don't use email? Nobody here. The waiter? I'm sorry. The camera man-- do you use email?

AUDIENCE: Of course.

PAPPALARDO: Hey. Come on, give me a break here. This is taking commonsense methodology. There's a big difference between pull and push, and that's as trivial a concept as we could ever hope to explain. But most people I heard altogether-- I didn't hear anybody mention email today at a conference. They mentioned using browsers for teaching everything else, but I didn't hear anybody mention the most fundamental reason that the whole computer technology started doing something useful. Yes?

AUDIENCE: On that note, I recognize that you use passwords, but do you have a general concern what the push technology has on the privacy threat associated with it? Especially if my email address happens to be a company email address.

PAPPALARDO: Well, first, we never allow passwords to be conjured up by the user. Passwords are always generated by the computer. They're generated at least every 30 days, and you get a new password every 30 days whether you like it or not. And so anytime we push stuff to you-- especially in the patient education one it was clear we had to push it to you because you're going to get on a browser to go access the information, but anybody could get on that. We don't want to identify who you are, so we simply tell you type in-- this is the password de jure-- to find out what's happening.

AUDIENCE: If I worked for Lockheed Martin, which I don't, the email that comes to me [INAUDIBLE].

PAPPALARDO: Well, if you're a physician, we would send you private encrypted email.

AUDIENCE: From a patient?

PAPPALARDO: Sure, but that doesn't mean-- I bypassed the sign on. I hope you didn't mind that-- for the demonstration. There's always a way for them to properly identify themselves. Now, is that the end solution of security? Of course not. But the point is, the more secure we make information, the harder it is to deliver good care. Anyone else? Hey, [INAUDIBLE]. I'm all done here.

[APPLAUSE]

AUDIENCE: Thank you, Neal.