The mistakes largely result not from surgeon fatigue, but from the stress arising from emergencies or complications discovered on the operating table, the researchers reported.

It also happens more often to fat patients, simply because there is more room inside them to lose equipment, according to the study.

Both the researchers and several other experts agreed that the number of such mistakes is small compared with the roughly 28 million operations a year in the United States. "But no one in any role would say it's acceptable," said Dr. Donald Berwick, president of the Boston-based nonprofit Institute for Healthcare Improvement.

The study was done by researchers at Brigham and Women's Hospital and Harvard School of Public Health, both in Boston. It was published in Thursday's New England Journal of Medicine (news - web sites).

The researchers checked insurance records from about 800,000 operations in Massachusetts for 16 years ending in 2001. They counted 61 forgotten pieces of surgical equipment in 54 patients. From that, they calculated a national estimate of 1,500 cases yearly. A total of $3 million was paid out in the Massachusetts cases, mostly in settlements.

Two-thirds of the mistakes happened even though the equipment was counted before and after the procedure, in keeping with the standard practice.

Most lost objects were sponges, but also included were metal clamps and electrodes. In two cases, 11-inch retractors — metal strips used to hold back tissue — were forgotten inside patients. In another operation, four sponges were left inside someone.

The lost objects were usually lodged around the abdomen or hips but sometimes in the chest, vagina or other cavities. They often caused tears, obstructions or infections. One patient died of complications, but the researchers withheld details for reasons of privacy.

Most patients needed additional surgery to remove the object, but sometimes it came out by itself or in a doctor's office. In other cases, patients were not even aware of the object, and it turned up in later surgery for other problems.

The study found that emergency operations are nine times more likely to lead to such mistakes, and operating-room complications requiring a change in procedure are four times more likely. A rise of one point in body-mass index, a measure of weight relative to height, raises the chances of such a mistake by 10 percent.

The length of the operation or the hour of day does not appear to make a difference, suggesting that fatigue does not cause such mistakes.

"It tends to be in unpredictable situations," said lead author Dr. Atul Gawande of Brigham and Women's Hospital.

Some other researchers said fatigue could promote such mistakes in a way undetected by this study.

The Boston research team suggested that more X-ray checks be done right after those operations where such errors are most likely. Metal instruments and radiologically tagged sponges show up in such checks.

Eventually, wands similar to supermarket bar-code readers might be developed to detect missing equipment, researchers said.

Dr. Sidney Wolfe, health research director of the public-interest lobby group Public Citizen, said the real number of lost instruments may be even higher, because hospitals are not required to report such mistakes to public agencies. He said they should be.

However, some others said such mistakes are so rare — occurring about 50 times in 1 million operations — that figuring out how to prevent them could be difficult.

"Something has to be done about this. It's just a very tough balance to decide. Do we really want to add this hoop for every patient to jump through?" said Dr. Kaveh Shojania, author of a 2001 federal study on medical mistakes.

Lori Bartholomew, research director at the Physician Insurers Association of America, said: "I find it's going to be difficult to make much more improvement, because some of the risk factors are things that are hard to control." The Rockville, Md., group represents medical malpractice insurers.