One patient heard a crunching noise when the surgeon cut through his bones. Another felt "white-hot fire pain" as a physician probed his insides, but he was unable to move.

It sounds like grist for a lurid movie, and indeed last fall, Hollywood explored this notion of patients waking up while under general anesthesia. But the episodes above - though highly unusual - are all too real, and they are described tomorrow in the august pages of the New England Journal of Medicine.

For many hospitals, the answer for this "anesthesia awareness" is to use a brain monitor, so doctors can give more drugs to patients who seem to be waking up. Yet authors of the new study concluded that in fact, the most common such device may not reduce the risk of that problem at all.

The manufacturer of that brain-monitoring technology, called Bispectral Index (BIS), disputed that interpretation. And anesthesiology chiefs at three Philadelphia-area hospitals said the study results would not keep them from using the device, which they said also helps them on the other end of the spectrum - making sure they don't administer too much anesthetic.

Though not perfect, it is one of several tools at a vigilant physician's disposal, all of which should be part of the mix, said Michael Goldberg, chief of anesthesia at Cooper University Hospital.

"What harm does it do? It's a sticky on your head," Goldberg said of the brain monitor, which detects electrical activity via electrodes mounted on the forehead. "Until they come out with something better, I will continue to use it."

At Drexel University, engineers and physicians already are trying to develop a better device, one which uses near-infrared light to measure changes in oxygen levels in the brain's prefrontal cortex. The team is presenting its findings at a conference this month in San Francisco.

The BIS "is a good monitor," said Jay Horrow, chairman of the anesthesiology department at Drexel's College of Medicine. "We think we can do better."

One thing is clear: though some physicians have been known in the past to dismiss reports of awareness as simply a bad dream, the consensus is that it exists.

Anesthesia awareness - regaining some level of consciousness during surgery - is thought to occur in perhaps one or two out of 1,000 surgical patients in the United States, for a total of 20,000 to 40,000 cases a year. The bulk of them do not feel pain.

Still, for some it is so disturbing that they suffer from post-traumatic stress syndrome, and must undergo counseling.

For general anesthesia, patients typically are given a mix of drugs - including one to "knock them out," and often another called a paralytic.

This relaxes the muscles in order to make surgery easier. But in the rare case that a patient starts to wake up, the paralytic effect can be horrifying. That's because the patient cannot move or speak.

In the study, conducted by researchers at Washington University in St. Louis, 2,000 patients were randomly assigned to one of two groups. All were hooked up to one of the brain monitors, but the physicians could see the machine's measurements only in one group.

In that group, an alarm bell went off if the monitor's readings dipped below 40 or went above 60, on a scale of 0 (no brain activity) to 100 (fully awake).

In the other group, an alarm went off if levels were too low or high on a different sort of monitor - one that measures the concentration of anesthesia in the patients' exhaled breath.

They were then interviewed three times: within a day after surgery, then one to three days later, and finally a month later. The interviewers did not know which group the patients were in.

The researchers reported two definite cases of awareness in both groups. The brain-monitor group had four additional cases described as "possible," while the second group had one possible case. There was no statistically significant difference between the groups, said lead author Michael S. Avidan.

He said even larger studies are underway, by his team and others.

"I am quite sure that our study is not the final word on this subject," Avidan said.

Officials at Aspect Medical Systems, the Massachusetts company that makes the BIS brain monitor, said the results were encouraging.

That's because even though there was no statistical difference between the results of the study groups, the overall rate of awareness - two per thousand - was lower than might be expected.

That's roughly the rate of awareness in the general patient population, but these patients were specially selected because they had a high risk of awareness, for one reason or another. For example, some had to be given lighter doses of anesthetic because they had cardiac or other medical conditions.

In such high-risk populations, the rate of awareness has previously been estimated at up to 1 in 100.

Cooper Hospital's Goldberg agreed that the two-per-thousand rate was encouraging, and said the BIS device was used on almost all Cooper patients. Zvi Grunwald, chair of anesthesiology at Thomas Jefferson University Hospital, said he limited his use of the device to high-risk patients.

Speaking of risk, some doctors have expressed a negative view of the Hollywood take on the subject, worrying that it could inaccurately raise patient fears about surgery. Without revealing the plot, a couple of the physician characters in last year's Awake are less than model citizens.

Avidan said only that the movie was "really bad."