A while back, physicians in one of the nation's outstanding hospital intensive care units were asked to predict which patients they were admitting to the unit would live and which would die.

Thirty percent of the patients the doctors predicted would live, died. Twenty percent of the patients the physicians said would die, lived.

In order to better understand who can and who cannot benefit from the technological wonders of the modern intensive care unit, the Federal Health Care Financing Administration has given George Washington University $500,000 to conduct a three-year study of all the patients in its hospital's intensive care unit.

The study is being conducted, said Dr. William Knause, conductor of the unit, "to better define what we're able to do with technology, to be able to prognosticate a little bit better the outcome of patient care.

"Right now," he continued "we're doing everything for everybody and we think there might be a better way to at least help us make some of the decisions we're forced to make anyway as far as prognosticating illness and as far as distributing resources."

"I can tell you right now that between 10 and 12 percent of the patients admitted here in the next three years are going to die," said Knause's codirector, Jack Zimmerman. "The trouble is, I don't know what 10 or 12 percent.

"The emphasis I would want to come out of this study is on quality, rather than quantity," said Zimmerman, who said such data would enable him and his staff to give relatives a better idea of their family member's chances of survival.

The launching of the study comes 10 years after the completion of the historic "Report of the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death."

"This is where the brain dead people were when the people at Harvard started working on their study." said Zimmerman, referring to intensive care units. "The brain people were in the intensive care units."

The release, and eventual acceptance, of the Harvard report drastically affected the practice of intensive care medicine. No longer were brain dead individuals maintained on respirators for days, weeks, or even months, with no hope for recovery.

The question, now, is how the results of the GW study will be interpreted and used.

A spokesman for the Health Care Financing Administration said yesterday the study is intended to allow the development of procedures that help determine whether and how to use intensive care units and to establish a "patient classification system by illness."

Could the results of the study be used to determine which Medicaid or Medicare patients would be admitted to intesive care units? the spokesman was asked.

"That could be one of the things though that is not the primary goal of thoug that is not the primary goal of the project.

"We very rarely do triage" - determine on the basis of their chances of surviving who should or should not be admitted to the intensive care unit, said Zimmerman, "mainly because our beds are such that we" have no need to.

"The dispersal of the technology" in the intensive care unit setting "has outstripped our ability to evaluate it," said Knause. And because no one knows whether an individual patient will pull through, the futile attempts to save him may continue on and on, "disrupting people's lives when this goes on for weeks and weeks."

Zimmerman, who said he never wants to be in such a position, also pointed out that "there may come a day when somebody says "you have so many dollars to run this unit for a year, go ye out and run it."' And decide who can and cannot benefit from the expenditure of those dollars.