ORLANDO, Fla.— "There's a very fundamental and serious flaw in the infrastructure of medicine," Lawrence L. Weed, M.D., said to close the Institute for Healthcare Improvement's 18th annual National Forum on Quality Improvement in Health Care. "You couldn't design a better system to create errors in medicine."

Those statements may sound obvious now, but it's a message that Weed, 83, has been trying to convey since the 1950s. (See http://www.health-itworld.com/emag/070104/289.html?page:int=-1.) But if the enthusiastic audience reaction was any indication, he may have reached a few thousand new healthcare quality advocates.

Bounding around professorially between the lectern and two easels set up on stage in lieu of a chalkboard, Weed, creator of the problem-oriented medical record (POMR) and the subjective, objective, assessment, planning (SOAP) format of progress notes, forcefully stated his case for IT in healthcare.

"We bring people up [in medical school] to think with their clinical judgment," said Weed. "People, it's not working."

Weed, whom IHI president and chief executive Donald M. Berwick, M.D., calls his hero, explained, "You're looking for the combination of finding in that patient that best match each chapter of the textbook. Since the mind can't do it, what are we doing? ... We're teaching probabilities."

Probabilities might be fine for some conditions, but, "If you have the uncommon disease, do you want a workup for 10 things you don't have?" Weed asked

"We're moving knowledge from its source. We've got all this knowledge from the medical literature and from the patients themselves. We're trying to move that to our everyday actions in medical care. And we said the way to do that is to put it in human heads and the human heads will take the actions," Weed said, to hearty laughter.

Comparing the chain of knowledge to an electric utility, Weed said, "The voltage drop across that line is pathetic."

The solution, according to Weed, is to let computers do the thinking, coupling knowledge to specific problems. (Problem-knowledge couplers just happen to be the thrust of PKC Corp., the software Weed founded in 1982 to develop clinical decision-support software.)

"New tools are everything," Weed says. "It's the tools we use that advance our civilization."

The tools aren't exactly new to Weed, who first wrote about electronic medical records in 1968, but they have to be applied properly.

"What's the point of outcomes data?" Weed wonders. So what if there are four times the rate of prostate surgeries in Salt Lake City as in Denver? "I wouldn't know whether I should move to Salt Lake so they don't miss my cancer of the prostate or move to Denver so I wouldn't have unnecessary surgery."

Berwick, for one, thinks medicine finally is ready for the change Weed has envisioned for half a century. "It's inevitable," said Berwick.

"Computing power is so much greater now," Berwick explained. "We also have a whole new generation [in medicine]. Computers are totally integrated into their lives."

"This is going to be a much easier sell now."