By Nicholas Regush

I remember one of the early meetings I had with Dr. Lawrence Weed -- doctor, computer specialist, inventor, philosopher, raconteur, critic of the Medical Establishment.

In a small office in a modest two-storey cottage on Mary Street in Burlington Vermont, Weed, then 62 -- it was 1985- almost immediately launched into his diatribe about how doctors, who rely largely on their memories to diagnose medical problems, take unnecessary chances with their patients, lives.

"Okay, let's take the patient who has high blood pressure," Weed said. " He basically expects his doctor to know all the possible diagnostic and treatment options and to choose an approach that fits his unique condition. Isn't that right?

"But it's wishful thinking," Weed continued, because there are about forty causes of high blood pressure. Most doctors don't know them. And, if they did, they can't possibly remember all they need to know about these causes, let alone the more than 150 questions to ask in order to rule them out. So what do doctors do in this particular situation? They play the odds. Doctors choose the four or five most common causes instead of the forty and begin to ask questions in support of what they have already started to believe. And they arrive at a diagnosis on the basis of what they believe is highly probable. Therefore, in the case of high blood pressure, causes considered to be rare may be overlooked or simply not remembered."

One example he cited is a hormone disorder called "hyperaldosteronism" that causes high blood pressure and is triggered by eating licorice.

Weed said guesswork is the rule throughout medicine and the public pays dearly for it, in terms both of personal health and of social costs. We know today as a result of numerous studies that a high percentage -- as much as 40 to 50 per cent -- of doctor-ordered procedures such as surgery, lab tests, X-rays and drug prescriptions are either inappropriate, excessive, or totally unnecessary.

Internationally acknowledged as a pioneer in the application of computers to health care, Weed had been trying for more than two decades to reduce the amount of guesswork in medicine. He had argued in medical journals, books, classrooms, and public lectures that doctors rely on a dangerously inadequate method of diagnosing their patients, problems -- namely, drawing on their storehouse of memorized facts, combined with gut instincts. He told me that such an approach is far too limited for what medical problem-solving requires, and he had been calling for the re-design of medical practice. The new methodology, he said, would incorporate the use of computers -- which have far superior memories to the human variety -- to assist doctors in their work.

Weed termed the computer "an extension of the human mind," pointing out that, in the same way, the microscope and the telescope are extensions of the eye and the car is an extension of our legs.

He argued that medical-school education must also be completely restructured, shifting emphasis away from the memorization of "repetitive and outdated" book and lecture data to the use of the computer as an encyclopedic guidance system that helps a doctor "couple the right information to the right problems." The new school system would also choose students on the basis of their interpersonal skills, manual dexterity, and interest in information science, rather than largely for their ability to pull in high grades, often evidence primarily of good memory work.

In Weed's vision of future medicine, some doctors would be selected to care for patients. Others, who would specialize in linking up computer systems to the very latest in medical knowledge, would become expert "map-makers" who would lay out main routes and options for traveler-doctors to follow.

This wasn't just theory. Weed was putting his ideas into practice. His vehicle to transform medicine was the "Problem-Knowledge Coupler," a computer program to help doctors diagnose various illnesses.

To demonstrate, Weed first inserted a diskette designed to display and process the large amount of information a doctor should consider during diagnosis. The doctor could run through sets of branching questions that appeared on the computer screen about the wide range of patients, possible medical conditions.

Weed offered an example of a hypothetical patient who had come to the office with abdominal pain. After entering basic information about the patient, such as age and sex, Weed began typing instructions on the keyboard. This would produce a display of questions to locate the pain: Is the pain in the right upper quadrant of the abdomen? The right lower quadrant? Or is the pain generalized? And so on. Weed made his choice on the keyboard and an arrow appeared on the computer screen beside "right upper quadrant." Then he called up the next set of questions. Gradually he worked through a detailed questionnaire about abdominal pain. " No two patients provide all the same answers to the questions," he said, " so we get a unique map of the patient's problem."

The computer program matched all the recorded answers to the pertinent medical literature and summarized the possible causes of the patient's illness. The suggested causes were listed in order of those with the highest absolute number of findings that match, along with comments from the pertinent medical literature. Weed said that at this point there might be a few potential causes of the illness that would invite further scrutiny." There is no substitute for taking the time to examine those possibilities with a reasonable number of matches, reading the comments and letting the patient in on the process," he said. "Patients will always be interested in what is a good match, no matter how rare the disease might be." The doctor may order medical tests. The computer program can even suggest which ones would be the most useful.

Weed's plan, back in 1985, was to revolutionize medicine by designing computer programs to help doctors diagnose illness and to involve the patient in the diagnostic process. Today, he is still very much involved in furthering his dream. The company he founded, PKC Corporation, now has fancier digs in Burlington and relationships with some health institutions. Weed has been widely honored -- even by the titans of Established Medicine - but most doctors still do not use computers, but rather use their memories. So overall, Weed's impact has thus far been limited.

Weed recently summed up the consequences of the mainstream medical decision-making system this way in the British Medical Journal:

"Because of its archaic intellectual infrastructure and its confusion about decision-making authority, the health care system produces decisions that frequently do not conform to the individual needs of complex, unique patients. Such decisions cause substantial harm, in both medical and economic terms. Moreover, the prevailing confusion and lack of patient autonomy block genuine reforms in many areas, including "consumer protection," credentialing, graduate medical education, analysis of clinical outcomes, and the framework for marketplace competition."

That archaic intellectual infrastructure to which Weed refers is modern medicine's death knell. But there is simply too much at stake to change. As Weed put it to me in one our many conversations over the years, "What computers do is take away the authority from the doctor and make the patient a potentially equal participant in the medical process -- and that's dangerous to the Establishment."

PKC Corporation

Weed's essays in the British Medical Journal