The solution of Dr. Kenneth T. Bird, who runs the unit, has been to provide a physician when the patient demand is heaviest, and to provide additional coverage by television. The system now used, called Tele-Diagnosis, is frankly experimental. It has been in operation for slightly more than a year. At the present time, eight to ten patients a day are interviewed and examined by television.

The Logan TV system is probably the first of its kind in the country, but Bird refuses to discuss priority. "The first to have it," he says, "was Tom Swift, in 1914."

Certainly there is a science-fiction quality about the station's equipment, for along with the Tele-Diagnosis apparatus, there is also a time-sharing station linked to the hospital's computer. Among other things, this computer can be used to take a preliminary history-to function as a doctor in questioning the patient about his symptoms and their nature. Some 15 per cent of the patients examined by Tele-Diagnosis have had their medical history taken by computer before they see the doctor himself. Like the cardiogram, the computer history can be sent directly to the physician.

Being interviewed by a machine is less bizarre than it sounds. Indeed, like the TV link-up, it is remarkable for the ease with which patients accept it. The most common complaint is boredom: the machine sometimes pauses three or four seconds between questions, and the patients get fidgety.

To be interviewed, one sits in front of a teletype console. The computer asks questions, which are printed out, and the patient punches in his answers. Whenever the computer gets a "yes" answer to some question, it follows it up with more questions on the same subject. If it gets a "no" answer, it goes on to the next topic. At the conclusion of the question, the computer writes out a medical summary. Unlike the questions, the summary is phrased in medical terminology. The entire process takes roughly half an hour.

The result of one such interview is reprinted in part below. The computer was given the same presenting complaint as that of Mrs. Thompson: chest pain. In an attempt to confuse it, the machine was first fed some false but suggestive information, namely, that there was a family history of cardiovascular disease, and that the patient was taking digitalis. However, in later questions, the machine was given a straightforward history for the type of chest pain most common among medical students-that of psychogenic, or musculoske-letal, origin.

A sample of the questions and answers ran as follows:

SEVERAL MONTHS

A FEW YEARS

DO YOU HAVE THIS COUGH EVERY DAY?

YES

DO YOU BRING UP ANY MATERIAL (SUCH AS SPUTUM, PHLEGM, OR MUCUS) FROM YOUR CHEST?

NO

HAVE YOU EVER COUGHED UP BLOOD?

NO

At the conclusion of these and other questions, the computer printed the following summary:

NO

DO YOU HAVE A COUGH? 

YES

HOW LONG HAVE YOU HAD THE COUGH? 

1. A FEW DAYS 

2. A FEW WEEKS

MEDICAL HISTORY SUMMARY DATE: MAY 27, 1969

NAME: MICHAEL CRICHTON UNIT #: DEMO

AGE: 26 SEX: MALE

CHIEF COMPLAINT: CHEST PAIN COMMUNITY PHYSICIAN: NONE OCCUPATION: MEDICAL STUDENT

MEDICATIONS: DIGITALIS DRUG REACTIONS: PAN ALBA HOSPITALIZATIONS: NONE

FAMILY HISTORY: HEART ATTACK, HYPERTENSION.

SOCIAL HISTORY 

PT. IS MARRIED, HAS NO CHILDREN. COLLEGE GRADUATE. PRESENTLY A STUDENT, WORKING 50-60 HRS/WK. HAS BEEN SMOKING 5-10 YRS, 1 PACK/DAY. ALCOHOLIC CONSUMPTION: 1 DRINK/DAY. FOREIGN TRAVEL WITHIN THE LAST 10 YEARS.

REVIEW OF SYSTEMS 

GENERAL HEALTH 

NO SIGNIFICANT WEIGHT CHANGE IN PAST YEAR. SLEEPS 6-8 HRS/NIGHT. HEAD INJURIES: NONE WITHIN PAST 5 YRS. EYE SYMPTOMS: NONE. HAS BEEN TOLD BY MD OF NO EYE DISEASE. NO TINNITUS. NO EPISTAXIS, NOTES SINUS TROUBLE, DENIES CHANGE IN VOICE.

RESPIRATORY SYSTEM 

PT. NOTES COUGH OF SEVERAL MONTHS DURATION, WHICH OCCURS DAILY. DENIES SPUTUM PRODUCTION, DENIES HEMOPTYSIS. NOTES NO DYSPNEA. HAS HAD HAY FEVER. HAS HAD NO KNOWN CONTACT

WITH TUBERCULOSIS. LAST CHEST X RAY -2 YRS AGO.

CARDIOVASCULAR SYSTEM 

PT. NOTES CHEST PAIN OCCURRING LESS THAN ONCE A MONTH, LOCATED "ON BOTH SIDES," WHICH RADIATES TO NEITHER ARM NOR NECK. PAIN IS NOT AFFECTED BY DEEP BREATHING, IS NOT ASSOCIATED WITH EATING, EMOTION, OR EXERCISE. PAIN IS NOT RELIEVED BY RESTING. PT. NOTES PALPITATIONS ON RARE OCCASIONS. DENIES ORTHOPNEA. DENIES PEDAL EDEMA, DENIES LEG PAINS, DENIES VARICOSE VEINS, DENIES PERIPHERAL REACTION TO COLD. CARDIAC MEDICATIONS: NONE. HAS BEEN TOLD BY MD OF NO COMMON CARDIAC DISEASE. NO ECG IN PAST 2 YRS.

This is only half the total report. Analysis of gastrointestinal musculoskeletal, genito-urinary, hematologic, endocrine, dermatologic, and neurological systems followed. This particular computer program draws no conclusions about diagnosis; it only summarizes answers to its own questions, and it does not cross-check itself. Thus, while the computer was told the patient took digitalis, it later accepted the conflicting statement that the patient took no cardiac medications.

This program, which was devised at the MGH, is a rather simple example of the way that computers can and almost certainly will be used in the future. But it is the least sophisticated of the medical-history programs available; more complex ones already exist.

When Mrs. Thompson arrived at the MGH emergency ward, which had been expecting her, she was taken down to the EW X-ray department. In doing so, she passed a door near the front of the EW which is unmarked, without a label. Over the door is a lighted sign that says, incongruously, "On Air."

Dr. Murphy was behind that door, sitting in a corner of a small room, surrounded by equipment. Directly in front of him was a camera and a large TV screen, on which he watches the Logan patients. Built into his desk were two other screens: one, a small monitor of the larger screen, the other, a monitor that showed him his own image being transmitted to the patient. This second monitor allowed him to check his own facial expressions, the lighting in the room, and so on.

To his right was a panel of buttons that controlled the various remote cameras-two in the examining room and one in the laboratory. The examining-room remote camera is operated by a joystick: by pushing the stick right or left, up or down, the camera moves accordingly. In addition, there are buttons for focusing and zoom control.

Before going out to check on Mrs. Thompson, Dr. Murphy continued a study of Tele-Diagnosis

capability: reading a series of 120 chest X rays that are set up for him at Logan. He planned to read these by TV and later reread them in person, to compare the accuracy and consistency of his diagnosis.

The nurse at Logan set up the next X ray.

"What's this one?"

"Jay-nineteen," the nurse said, reading off the code number.

"Okay." He moved the joystick and touched the buttons. The camera tracked around the X ray, examining the ribs, then scanning the lung fields. "Wait a minute." He zoomed in to look closely at the right-upper lobe; he watched the little monitor, because resolution was better, but by glancing up at the large screen, he could also get a magnified view. "No. Well, on second thought…" He zoomed back for an over-all view. He zoomed in on another part of the upper lobe. "Looks like a small cavitation there…" He zoomed back, touching the buttons. He turned to the joystick, panned across the rest of the lung field, occasionally pausing to look at suspicious areas. "Nothing else, not really…" He finished his scan, and returned to the right-upper lobe. "Yes, there's cavitation. I'd have to call it moderately advanced tuberculosis. Next, please."

He was working with considerable rapidity.

"You get to be pretty good at this," he said. "At first, it all seems clumsy, but as you get more accustomed to the equipment, you move faster." The average time for a patient interview and exanimation by Tele-Diagnosis is now twelve minutes, less than half the average figure a year ago.

"What I'm doing now," he said, "is really just a test of our capability. It has no immediate practical use, because we can't take X rays at Logan-that's one of the main reasons we brought Mrs. Thompson into the hospital. But it's important to know if X rays can be read at a distance with accuracy. Our impression is that you can read them as well on TV as you can in person."

"Jay-twenty," the nurse said, putting up another film.

Murphy began his scan. "Ah. What's this? Looks like a rib fracture…"

One can argue that for the past twenty years technology has defined the hospital, has made it what it is today. That is, once a range of expensive, complex therapeutic and diagnostic machinery became available, the hospital assumed the role of providing a central location for such equipment. This was inevitable: private practitioners and even large group practices could not afford to buy such equipment, nor maintain it, nor pay the personnel to operate it. Only the hospital could do this. It was the only institution in existence that could possibly absorb the expense. Other possible institutions, such as nursing homes, were wholly inadequate.

Furthermore, because the hospital was already oriented toward acute care of critically ill patients, the technology that it absorbed was precisely that

which helped in this area. Monitoring machines and life-support equipment are clear examples. Thus technology reinforced an already existing trend.

Now, however, the pressures and forces acting upon the hospital are social and of a nature that is changing the meaning of technology within the hospital. As C.P. Snow has said, "We have been letting technology run us as if we had no judgment of our own." But such judgment is now required, and one can argue that in the next twenty years the hospital will define technology. That is, it will create a demand for new technological applications- and in certain ways will itself produce the new technology.