Health-Oriented Telecommunications

by Joe Flower


This article originally appeared in Wired, January 1994, Vol. 2.01

International Copyright 1994 Joe Flower All Rights Reserved
Please see our free downloading policy.



Leave Hillary and Bill out of it -- whatever they do in Washington, the way you keep yourself healthy, and what you do when you are sick, is going to change drastically over the next decade in about 15 different ways. And one big way has everything to do with smart cards and dumb terminals, big bandwidth and micro-probes, genetic markers and info-markets.

It's the next American healthcare, and it doesn't look like anything you've read in the paper.

Picture this:

All three of these scenarios are possible within a decade.

Every dimension of health care is changing, driven by powerful systemic feedback from both within and outside. Over the next decade, we will see it become less doctor-centered, and more community- and family-centered. Medicine itself will become less of an art, and more fact-based. The very discoveries and inventions that will continue to transform medical practice will push it to be less about hardware, less about vast and powerful machines watched over by highly-trained acolytes, and more about shared information. Healthcare will shift its center of gravity away from last-minute, traumatic, intensive, expensive, short-term hospital-centered care, and toward early-as-possible, preventive, long-term, intimate, inexpensive application of information in the community and the family.

Any way you measure it (per capita, absolute dollars, percentage of gross domestic product), we have the most expensive healthcare system in the history of the world, yet Americans are less satisfied with their healthcare than citizens of any other developed nation. We are far from the longest-lived people, our infant mortality rate is 22nd in the world (equal to that of Greece), our maternal mortality rate is 26th. I remember in 1990 giving what I thought was an alarmist speech at a healthcare conference, predicting that our healthcare costs would hit $1 trillion by 1995. I was wrong. It will hit just about that number this year [1994], and it's going up at a rate of $100 billion per year.

This situation has to change, and HOTS (health-oriented telecommunications) are likely to be a major part of that change

The fact is, of all the actions that make up this vast humble-jumble we call health care, the great majority are already transfers of information rather than shots, cuts, and pills -- diagnostic information, such as X-rays, vital signs, and the patient's look and feel; therapeutic information, such as prescriptions and orders for tests; consultations ("The pneumonia is one thing, doctor, but I don't like the look of these infiltrates"); professional education; patient education ("Mrs. Jones, for the good of your baby, you have to keep your weight up during pregnancy. This is not the time for a crash diet."), and billing and insurance information.

Much of this information, even today, is handled in ways with which Charles Dickens would be completely comfortable -- face-to-face, or on paper, in prescriptions, filled-out insurance forms, and notes scribbled on medical charts. Much of the information is never recorded at all. And, even if it is recorded, an estimated 25 percent of this information does not get where it is needed (it's lost, misrouted, or not sent in time) and must be generated again. "If you get referred to a hand surgeon," says Dr. Richard Satava of the Advanced Research Projects Agency (ARPA), "and the X-rays don't show up with you, he's not going to talk to you until he takes another X-ray."

Health-oriented telecommunications can fix that problem, on the way to making much larger changes possible.

Having the HOTS would do four major things in healthcare:

1. Develop new diagnostics and therapeutics (or: traditional doctoring, but with more cool gadgets)

For decades, medical technology has focused on ways to bring more and better information to the doctor. Now the focus is shifting to ways to project the doctor outward, to make more efficient use of the physician's time, and to distribute the expertise of specialists where it is needed -- including battle zones, rural areas, inner cities, prisons, and mental hospitals.

For two years already the Medical College of Georgia has been operating telemedicine links that allow doctors in rural hospitals to consult with MCG specialists over live video links. The same links transmit ultrasound and X-ray images, and the sounds from electronic stethescopes. To add the element of touch, MCG is working with engineers from the Georgia Institute of Technology to develop a sophisticated dataglove link: the rural doctor would don a glove and feel a growth that might be a tumor (for instance), and the specialist at MCG would put a hand in a sensory generator hundreds of miles away and "feel" the same tumor, with the same texture, the same temperature, the same feeling of pressure.

The biggest hurdle is not technical but bureaucratic: most insurers will not pay the specialist to consult on a patient who is not in the same room. Not yet.

But what comes next is even stranger: SRI International is already spinning off a company that will market a "telepresence" system that will allow online surgery. ARPA is up to its elbows in online surgery for both military and civilian uses: last June, at Fort Gordon in Georgia, surgeons in a MASH tent operated on animal parts from a slaughterhouse. The animal parts were on an operating table in another MASH tent a hundred yards away -- and the surgeons were mentored by a specialist at Walter Reed Hospital in Bethesda, Maryland, who watched the operation on video.

Dr. Richard Satava and other ARPA specialists have designed "telementoring" helmets that could be worn by battle medics -- or by civilian EMTs. Put the helmet on, with its imbedded video cameras and microphones, and the doctor back at the hospital can see and hear what the EMT sees and hears, coach the EMT, and help the EMT to do far more complex things in those first critical moments, to try and stabilize the patient, than is possible today. Prototypes of these helmets have already been deployed in Somalia and Macedonia.

The ARPA project extends the "teledoc" into battle conditions. Instead of throwing the wounded soldier on a Humvee and racing for the field hospital, medics would pull the soldier into an armored ambulance -- a miniature trauma center crammed into an armored personnel carrier. Guided by "telementor" doctors at the hospital, they would do everything possible to stabilize their charges on the spot -- including telepresence surgery -- before taking that long risky journey to the rear.

Civilian centers such as the Dartmouth-Hitchcock Medical Center and SRI International are developing new ways of seeing into the body. SRI has combined with Stanford, UC Berkeley, the National Library of Medicine, and the Digiray Company to develop a system that would combine MRI scans, CAT scans, traditional X-rays, and Digiray's soft-tissue X-rays into a single virtual image, which physicians could view from any angle, at any scale. Last June, physicians at Brigham Women's Hospital at Harvard fused a video image with an X-ray image to find a tumor that was otherwise invisible to the surgeons.

Combine "telepresence" surgery with virtual reality, and what you get is a surgical simulator. "Now," says Satava, "we have to practice laparoscopic surgery [see "Telesurgery: the evolution of an idea"] by sticking probes into plastic boxes, picking up grapes and cutting apart chicken wings. In the future, we will be able to practice on the very same instruments that we use for the actual surgery -- first you do it on a virtual human, then you do it on a human."

At the same time, expert systems (which emulate the heuristic "rule of thumb" knowledge of experts in a particular field) will expand and support the doctor's skill. One, the Apache system, is already in use in emergency rooms. Another, the 3M HELP Patient Care System, helps doctors work through the testing and decision tree for specific sets of symptoms.

All of these are ways of projecting a doctor's skill and multiplying a doctor's usefulness. "We are closing the loop on the digital physician," says Satava. "We can put a doctor in every foxhole, at every bed in every rest home, in every ambulance, and even in your home."

2. Immerse traditional medicine in a wider, deeper information net (or: get the docs and hospitals wired).

Medicine is still pretty seat-of-the-pants stuff. Most procedures, in most medical settings, have simply never been never been rigorously examined to find out whether they are even needed, whether they work, and what would make them work better. To do that, you need what has only recently become available: massive numbers of computers across the country, reporting huge amounts of data from actual cases. Put that information together in databases, and find ways to control for different ways of reporting the raw data, and suddenly you have a way to say, "These five hospitals do a lot better than everyone else on ectopic pregnancies (or premature infants, or renal cancer). Let's go see what they are doing differently."

This is called "clinical benchmarking." It is part of the "outcomes movement" that is spreading rapidly in healthcare. Many hospitals are paying as much as $100,000 per year for this kind of information.

On the patients' end, this is pushing hospitals and doctors toward fully computerized patient record and clinical systems. Most clinical information -- vital signs, doctors' observations, nurses recording whether they have given the proper dosage -- is still recorded on paper. And information from a stay in one hospital usually does not link up with information from a stay in a different hospital. And the hospital records don't link up with records from your doctor's office, a clinic you visited, or an outpatient surgical center. Most people do not really have a coherent "medical record" gathered in one place. That's one reason you have to fill out those huge forms every time you go to a new place. All that information is recorded somewhere, but if you didn't tell the new place that you are allergic to penicillin, or fought in Vietnam, or have had three children, they wouldn't know -- and if you're unconscious, you can't tell them.

One Arthur D. Little study counted up $15 billion we could save every year just by digitizing everyone's medical records and putting them online, available to anyone with the right codes anywhere in the country. Eventually everything from insurance information to x-rays and MRI scans, will go digital -- and eventually you will carry all that information with you on a card.

On the research end, aggregating vast amounts of clinical information online will lead to "n-dimensional queriable research databases." A researcher will be able to ask "what if" questions, and answer them by dipping into databases with hundreds of millions of records, and aggregating all cases that fit. Currently, a researcher might ask a question with a single dimension and a longitudinal progression, such as: "Give me all the patients who took such-and-such drug, and search their records for six years afterward for these possible after-effects." But even the most powerful Sun workstations may take up to two weeks to provide the answer. So the rapid increases in computer speed and power that are expected by the year 2000 -- widely estimated at a 104 increase in computing power and a 107 increase in transmission speed -- will make a huge difference.

Combine such databases with expert systems, and you get expert systems that can learn: "The standard dosage is 10 cc, but the experience in the database shows that, with an elevated platelet count in a male over 60 with lung complications, 5 cc would be better."

Tie the entire healthcare system into a vast data structure, and you get a system that can learn how to do it better, faster, cheaper and easier.

3. Tie individuals directly into info nets (or: get everybody wired, top-down)

The health info-net is already beginning to spread beyond the hospital and the doctors' office. Some of the first tries are relatively simple. For instance, Healthy Start is already using a CD-based multi-media Community Services Workstation (developed by the D.C. government, Baylor and Emory Universities, Bell Atlantic and the Office of Disease Prevention and Health Promotion). A health worker sitting down with a client can access everything she needs to know to get the right kind of help -- to put together the housing with the medical care, and the chemical dependency program with the child care. It's a lot of work being poor. It's hard, not easy, to find the help and get to it. One of the principle things that keeps poor people sick and unproductive is simple lack of information, easily gathered, reliable, and understandable information.

In healthcare, you can cut costs, and do a better job at the same time, if you help people learn how to improve their health, if you catch disease as early as possible, and if you take care of people at the lowest appropriate level of intensity -- keep the colds and the aching backs out of the Emergency Room; don't tie up Intensive Care with something that a little prevention could have caught.

Telecommunication has powerful potential to help. A two-way, digital, switched broadband fiberoptic network that extended into people's homes would allow a lot more of the frail elderly and the chronically ill to live independently, rather than in convalescent homes. Relatively less-trained practitioners could visit them at home, and tap directly into the technological support and expertise of the hospital, hooking up to share vital signs, advice, and a direct look at the patient with the patient's doctor.

Does this work? In a Dartmouth study, doctors cut clinical visits and hospital days by nearly a third just by making a follow-up phone call after someone had visited the clinic the first time. Information and communication are powerful health measures.

Similarly, some Alzheimer's patients and people with certain types of acute schizophrenia could live with more independence and less constant supervision if remote monitoring -- much like that now used on some parolees -- warned when they had wandered outside their "safe zone."

Many people at serious risk of heart attacks, strokes, or other sudden, lethal shifts in their body systems would welcome such monitors, if they could give them the freedom to live a normal life. A miniaturized body monitor, either worn or implanted, connected to an Iridium-type global telephone system, would do the trick -- since the global phone must establish its position as part of its normal operation.

The ARPA project is already making such monitors part of the battlefield HOTs of the future. Satava pictures every soldier carrying four small devices: 1) a global positioning system locator, 2) a friend/foe identification broadcaster ("We get tired of losing 20 percent of our casualties to `friendly fire,'" says Satava.), 3) a vital sign monitor that can sense blood pressure, temperature, and other indicators, and evaluate whether the soldier is all right, injured in some minor way, in shock, or dead, and 4) an automatic communications link that would start broadcasting as soon as the vital signs went beyond normal parameters.

Our information net is developing rapidly, but so far one of its most powerful possibilities -- helping us stay alive and healthy -- has been lagging far behind.

4. Tie individuals directly into decision nets (or get everybody wired, top-down, bottom-up, and sideways)

A few read-'em-and-weep statistics on our $1 trillion-a-year bankrupting-the-country healthcare industry:

Prognosis? We could save money in big buckets if we gave people an easy way to grab good information about their own health at home. Michael McDonald, Chairman of Communications and Computer Applications in Public Health (CCAPH) calls this the Personal Health Information System. Picture the kind of TV/telephone/computer information appliance that people widely expect will a big part of the home in the future. Give it a home version of a medical "expert system," fitted with a highly interactive graphic interface. It can ask the kind of questions, and give the kind of answers, that the doctor at the clinic would: "No, if the boy didn't pass out, it's not a concussion. Here's what to watch for . . ." Or: "If it's round and has definite edges, it's not skin cancer. Here's what skin cancer looks like . . ." Based on what you tell it, the system can triage the cases you can take care of yourself from the ones the require a doctor's care ("Better make an appointment") and the ones that require instant attention ("I've dialed 911.")

Such a system would know you. You would give it your digitized medical history and answer its questions. It would ask you different questions, and give you different advice, if you were a 48-year-old woman with a family history of breast cancer, than if you were a 28-year-old gay male, or a black man thinking about getting married.

Or this system, with all its interactivity, could be put online. According to McDonald: "Part of the reason self care, prevention, and health promotion are so undeveloped is that traditional print and mass media do not allow the individual to access health information when they need it in a form that aids appropriate decision making. Health-oriented telecommunications are likely to revolutionize this part of the health system by making available anything a person needs or wants to know about their health 24 hours a day, seven days a week in the home, school, workplace, or through public terminals."

The Health Promotion Resource Center of the San Francisco Health Department already has proposed the first stage of this: bringing people basic information about health through interactive terminals in all their facilities, and through the San Francisco Library, which is developing one of the most advanced library-based telecommunications systems in the world.

But in many ways the most revolutionary part of all these ideas is the part that allows sideways, many-to-many discussions, just as today seniors can talk on SeniorNet, homeless people can tap into HandsNET, and MCHnet deals with matters of maternal and child health. One bulletin board system, CHESS, was designed solely to deliver information to people with HIV -- but the people with HIV wanted more than access to the latest information. They wanted to talk to each other. And the people who ran CHESS changed the system to meet the demand.

Many of the factors that can actually build healthier communities -- factors such as diet, street crime, environmental problems, access to information, good child care and housing, and substance addiction -- can be most powerfully affected at the community level, but the energy has to come from the community. Systems that allow open "forums" provide a matrix on which this kind of community energy can grow.

Developed and used properly, health-oriented telecommunications carry possibilities that go far beyond gee-whiz Buck Rogers romanticism, and even beyond more toys for the big boys. They carry the possibility of providing a major assistance in revolutionizing health care, making it both cheaper and better, spreading it wider, involving people in making decisions about their own lives, helping America (and eventually the world) build truly healthier communities.

How close is this stuff?

How close these ideas are to reality varies enormously. Some are already being put in place, at least in their first-generation versions. These include expert systems, telemedicine links, and using massive databases for outcomes management and research. Others are in the prototype stage, including dataglove links, image fusing, telesurgery, and telementoring. Still others, such as the battlefield "armored ambulance," are just ideas. Some of the ideas will require not just technical work, but large-scale funding, a major expansion of the infrastructure, and a long-term political commitment -- including especially a universally-available personal health information system.

How can we afford this?

Several ways:


Medical records can be revealing -- with notations of electro-shock therapy, HIV status, abortions, plastic surgery, treatments for depression, impotence, or chemical dependence. In the hands of a tabloid journalist, a political opponent, an insurance company bureaucrat, or an employer, the bits and pieces in an ordinary medical record can be damning.

Many people would be horrified at the prospect that their medical record will be digitized and made available online, and privacy advocates are already speaking up. According to Janlori Goldman, director of the American Civil Liberty Union's Project on Privacy and Technology, "While a hospital storage room full of paper files may raise some very serious worries, the threat to privacy is at least limited to those who are physically there and can get into the room. With remoter access from around the country and around the world, electronic data interchange might make possible multiple invasions at the same time by people scattered across the globe."

Michael McDonald of Communications and Computer Applications in Public Health (CCAPH) acknowledges such concerns ("The privacy and confidentiality of all health records must be maintained."), but dismisses them: "That's like saying `Cars are dangerous.' Of course they are. That's why we have brakes, stoplights, doorlocks, and ignition keys. There's no reason why digitized medical records connot be kept safe. With proper computer security, your records, with your name attached, would be available only to the people who have access to them now -- your doctor, for instance -- plus emergency medical technicians who have the proper access code."

McDonald points out that, at present, individual medical records are comparatively easy to steal by ordinary, non-computer means. "What would be a lot easier to steal by computer is population stuff -- aggregations from vast numbers of records, such as lists of all HIV-positive people. The answer is simply not to aggregate that type of information with the names attached."

Telesurgery: the evolution of an idea

Since 1989, surgeons have been performing a sophisticated new version of "laparoscopic" operations on the abdomen. Instead of slicing you open to work on that gall bladder, the surgeons push several pencil-size probes through your abdominal wall. One of them carries a video lens and tiny lights. Then the surgeons inflate the belly so that there is lots of room to work. "It makes you look nine months pregnant," says Richard Satava, M.D., one of the surgeons who pioneered the idea. "It's like working in the Astrodome." The surgeons stand over the patients, watching a television screen several feet away, while they manipulate the instruments.

"I've trained a lot of surgeons," says Satava. "Among the young ones, I can always tell the ones who've played a lot of Nintendo -- they never look at their hands."

Patients love laparoscopic surgery: little or no pain, little or no need for stitches, almost no risk of infection, go home the same day or the next day, back to work in a week. Hospitals love it. Insurance companies love it.

Surgeons hate it. "Ever eat with chopsticks?" says Satava. "Fine, but with one chopstick in each hand? There is no dexterity, the movements are counter-intuitive, there is very little sense of touch, the TV image is hard to see."

An Army surgeon stationed at the time at Fort Ord near Monterey, California, Satava had applied several times to be an astronaut ("Since the day I was born, I wanted to do the first appendectomy in space."). So when he set himself three goals for improving laparascopic surgery -- 3D vision, dexterity, and a sense of touch -- he went straight to NASA Ames in Mountain View. They sent him to Jaron Lanier's VPL, and from there he ended up working with Phil Green at SRI International.

Green and other SRI wizards developed a system that gives surgeons all three: 3D vision (by looking at a monitor through field-sequential liquid-crystal shutter glasses), dexterity and force-feedback for a sense of touch (by substituting computer-driven servo arms for the simple "straight sticks" that the surgeon held).

"We were several months into this project," says Satava, "evolving this idea with the surgeon standing over a monitor with his back to the patient, when suddenly we realized that there was no need for the surgeon to be in the room. You don't even need a doctor to stick in the probes -- an OR tech or a nurse could do that." The surgeon could be blocks away, or miles away. In fact, with a good broad-band fiber-optic hookup, the surgeon could be as much as 500 miles away (beyond that, seemingly unsolvable time-lag problems would make the scheme unworkable).

Changing the shape of healthcare

Besides making healthcare in many ways both more effective and cheaper, the telecommunication revolution may well shift its shape, making it:

  1. less doctor-centered and more patient-centered

  2. less scattered and more integrated (encouraging the kind of large, integrated healthcare systems that many experts feel are necessary to both bring costs down and increase quality)

  3. less local, more regional, national, and even global (allowing doctors and laboratories to do actual work across far greater distances than ever before)

Does information make a difference?

"My first daughter, Mikayla, was born with cystic fibrosis (CF) . . . . If we, her parents, and her doctors had had access to information systems which allowed us to quickly and easily explore information on alternative diagnostic and treatment approaches, she might have been saved from lifelong disfiguring scarring.

"I don't mean to sound ungrateful for the medical care she received. Since the first month of life, she has lived a virtually normal life . . . . We were lucky to have been in an urban hospital and to have had an extraordinary resident who was able to catch her distress early. What would have happened if Mikayla had been born in a rural area or had physicians with less up-to-date expertise? She would surely have died -- as infants and children do of preventable and treatable illnesses every day because of a lack of appropriate information and decision support."

-- from the testimony of Michael McDonald, chairman of Communications and Computer Applications in Public Health, before the Science Subcommittee of the House Science, Space and Technology Committee, February 2, 1993

Articles | Healthcare | Technology articles | Main Page