Doctors could reduce the use of antibiotics in hospitals by 20 to 25 percent, cutting $100 million to $200 million from the nation's medical bills each year, without adversely affecting their patients, a new study indicates.

A team of researchers, writing in this week's New England Journal of Medicine, found that 30 percent of all antibiotics given in hospitals are administered for prevention of illness in operations or nonsurgical procedures.

But 80 percent of these are administered more than 48 hours after the procedures, when, the researchers conclude, they are of no value.

" . . . If prophylaxis were discontinued 24 to 48 hours after a procedure, from 20 to 25 percent of all antimicrobial-drug use in hospitals would be saved," the report said.

Such a limitation would "undoubtedly reduce the number of drug reactions, would probably diminish to some degree the rate of emergence of (drug) resistant organisms and would save not only the cost of unused drugs but also the considerable costs of ordering, storing and dispensing such drugs," it added.

Dr. Edward H. Kass of the Harvard Medical School, one of the authors, said in an interview that while his efforts to assign a dollar value to the savings were "crude," "We estimate we're talking about somewhere about $100 million and $200 million," and those numbers are "in dollars of three years ago."

The study was based on an examination of the charts of 5,288 patients selected at random from 20 hospitals in Pennsylvania.

Based on those charts, the authors calculated that 157,000 patients a year received antibiotics for prophylaxis in Pennsylvania, and, if that state is representative, 3.3 million are so treated nationwide.

Kass noted that "roughly 70 percent of all antibiotics" go for "strictly therapeutic use" -- curing disease -- and "most of that 70 percent is strictly legitimate by anybody's criteria."

The problem arises in the administering of antibiotics when the patient is not sick, but is thought to be especially susceptible, such as after an operation.

The report said that only in certain specific procedures, such as surgery on the large bowel or hysterectomies, is there good evidence that prophylaxis "has been shown to reduce the rate of postoperative infection."

In other procedures, there is not sufficient evidence to determine whether there is benefit, the report said.

But in any case, there does not appear to be any value in extending antibiotic treatment more than 48 hours after the operation.

Still, the report found, "once prophylaxis had been initiated it tended to be continued for the remainder of the hospitalization of the patient . . . The familiar tale of the nurse rushing to give the departing patient the last dose of antimicrobial drug before discharge has much credibility."