Nearly half of the 400,000 Americans who have heart attacks each year could safely leave the hospital after seven days, a week earlier than most are let go now; a Duke University medical team maintains.

By "conservative" estimate, they say, this revolutionary step would save $360 million a year in hospital bills and might produce less anxious, less depressed patients, spared many long days in a mechanical environment.

The Duke report - the lead article in today's issue of the New England Journal of Medicine seems certain to speed a trend toward shorter hospital stays for heart patients.

Duke Medical Center is a national leader. And more movement toward briefer hospitalization is endorsed by Dr. Richard Ross, Johns Hopkins University medical deam, in an editorial in the same issue of the widely read journal.

Cutting hospital stays short for some patients might result in "a few" late complications and deaths, he wrote, but the benefit of a long stay "must be considered in relation to its cost" at a time when society can "probably" no longer afford to keep low-risk partients in hospitals longer than a week.

In a interview yesterday, he said the price of sending patients home that early might be one death in 500 or 1,000 cases. This low a risk assessment commonly forces doctors to decide some patients should have less, rather than more, of some costly kind of care.

Dr. Myron Weisfeldt, Johns Hopkins cardiology chief, said current Hopkins practice for low-risk heart patients without apparent complications is "10 days' or two weeks' hospitalization, or eight days' at the least." But "I think we are progressing in the direction" of a week's stay, he said, and the Duke study "lends support to using these criteria on our patients."

The Duke team - Dr. J. Frederick McNeer and six colleagues at the Durtheory on 67 patients. All had acute myocardial infactions (blockage of the arteries that feed the heart), but no serious complications for the first four days in the hospital.

Thirty-three patients were discharged after a week. The rest stayed two to six days longer.

There were no serious complications in either group in the next three weeks. There were no deaths in either group and no differences in health after six months.

The Duke conclusion was that "it is feasible and ethically justified" to discharge such patients after a week.

As to savings, they said 1975 nation wide data showed a 17-day average stay for such patients. A 1973 study done differently, showed an average 15-day stay.

Taking the shorter estimate and using a probable saving of $2.03 a patient (the difference between continued hospital care and the lower cost of some continued home care, at 1977 prices), the Duke doctors calculated that at least $360 million a year could be saved.

At George Washington University here, cardiology chief Patrick Gorman agreed yesterday that "the trend is toward shorter stays." But he wondered whether "hospital logistics" would permit one-week stays, since "a patient spends three or four days in a coronary care unit, then needs a period of adjustment and education before discharge." He said G.W. "typically" keeps heart patients 11 or 12 days.

But Johns Hopkins' Ross said "the next step" may be partial or complete home care for many patients. He cited a 1976 British study in which 455 heart patients without obvious complications were randomly assigned to hospital or home care - and those at home did as well as those hospitalized.

"The best features of both hospital and home care" might be achieved Ross said, by a system in which technicians monitor a heart attack victim at home for the first wo hours after the attack and then decide if the victim can stay at home safely or must go to the expensive and complicated hospital.