Keep a shock patient warm.

That's long been standard first aid practice for someone with low blood pressure and a weak pulse, whether it be someone suffering internal bleeding after a car crash or someone with an oozing stab wound. It seems like common sense. But it may be wrong.Odd as it seems, cooling a patient suffering hemorrhagic shock might be beneficial, researchers at the University of Pittsburgh say.

In animal experiments, they have found that lowering body temperature to about 93 degrees can significantly extend patient survival time - the so-called "golden hour" in which trauma surgeons try to begin treatment.

"It flies in the face of what we think of as standard care for trauma patients," admitted Dr. Samuel Tisherman, who has led the research program along with Dr. Peter Safar at Pitt's Safar Institute for Resuscitation Research.

But mild cooling protects oxygen-starved internal organs and thus may prevent the multiple organ failure that ultimately kills many patients days after they have been successfully revived, the researchers say.

Following a few more animal experiments, Tisherman said he and his colleagues may be ready later this year to test their theory in a clinical trial.

They will have some persuading to do among trauma surgeons, however.

"If you are truly in shock, you can't be cold or you'll die," said Dr. Larry Gentilello of the University of Washington in Seattle. "There's not a single study in humans saying you do better cold."

Trauma surgeons have long noted that the colder patients are, the more likely they are to die. And a study by Gentilello published last year showed that patients who were rapidly rewarmed were significantly more likely to survive - at least in the short-term - than patients treated less aggressively.

The therapeutic use of cold is by no means new, said Safar, 70, who helped develop cardiopulmonary resuscitation in the 1950s and pioneered critical-care medicine in the 1960s. Brain surgeons and heart surgeons have long cooled their patients as a means of extending the time they could survive on the operating table.

But the use of cold in resuscitation was largely ignored until the mid-1980s, when Pitt research demonstrated that cooling could save brain cells in emergency situations.

Mildly cooling patients with brain injuries has been shown to improve outcomes, as reported last year by Pitt's Dr. Donald Marion in the New England Journal of Medicine. Animal experiments suggest similar benefits in protecting the brains of people who suffer cardiac arrests. In Japan and many European countries, cooling of emergency patients with cardiac arrest has become standard, Safar said, though it has yet to be studied in a U.S. clinical trial.

In the case of hemorrhagic shock, however, the rationale for cooling is a bit different. Cold isn't necessary to protect the brain or the heart in such cases; as long as the heart keeps beating, the brain can protect itself by relaxing its blood vessels, and the heart can sustain itself.

By contrast, blood vessels feeding the visceral organs, such as the liver, kidneys and intestines, constrict and limit the amount of oxygen that reach them.

As a result, a patient that remains in shock for more than the golden hour may suffer irreversible damage to these organs. Even if these patients are successfully revived and their original injury is repaired, these patients may die days later as their liver and other intestinal organs slough away, Safar said.

In these cases, cooling the body may help preserve these visceral organs. Safar said mild cooling can be accomplished by using bags of ice placed around the head and neck. In many trauma cases, Tisherman added, shock patients may need no additional cooling but simply not be rewarmed.

In either case, this cooling, or hypothermia, must be controlled, with the patient maintained at the desired temperature. Allowing patients to cool uncontrollably may be just as dangerous as health-care workers have assumed it to be, Tisherman said.

In their animal experiments, Safar, Tisherman and their colleagues have induced shock by removing blood from anesthetized rats. In one study, 54 rats were kept in shock for 75 minutes before being resuscitated; one-third were maintained at normal temperature, another third at 93 degrees and the rest at 86 degrees. Of the 36 rats that were mildly or moderately cooled, 29 survived at least 72 hours. Of the 18 rats in the normal temperature group, just three survived as long.

But the University of Washington's Gentilello questions whether these conditions adequately mimic the stress of a patient who not only is in shock but has suffered some traumatic injury and may suffer uncontrolled hypothermia before help arrives.

In a study at Seattle's Harborview Medical Center, Gentilello and his colleagues tried to determine whether cold patients did poorly because they were cold, or if the cold simply reflected the fact that they were sicker.

One barrier to studying this problem is that conventional techniques, such as blankets, may require four hours to rewarm a patient to normal temperatures, Gentilello said. That, he argued, isn't fast enough to make the role of temperature clear. So Gentilello invented a device that would re-warm patients in less than an hour by running their blood through aluminum tubing immersed in warm water.

He and his colleagues studied 57 trauma patients who had temperatures of 94 degrees or less. Half were rewarmed using conventional means, while the other half were rewarmed using his device.

They found that 12 of 28 patients receiving conventional re-warming failed to reach at least 97 degrees and all 12 died. By contrast, only two of the 29 patients who were actively rewarmed failed to reach 97 degrees, and both died.

But although rewarming had a dramatic effect on short-term survival, the advantage over the long-term was less pronounced. Half of the 28 patients who had conventional rewarming survived to be discharged from the hospital, compared with 19, or two-thirds, of the 29 patients who had active re-warming.

The bottom line, he argued, is that warming is valuable even if it doesn't reduce the later organ failures.

"You have to be alive to have a late organ failure," he added.

But Tisherman and Safar maintain that the therapeutic hypothermia that they advocate is different than the uncontrolled hypothermia that Gentilello was trying to counteract. It's true that patients who cool uncontrollably do worse than other patients, Tisherman said.

In controlled hypothermia, they said, health-care workers monitor patients to make sure they don't shiver, which can cause metabolism to shoot up, and they make sure the body doesn't cool too much.

Safar and Tisherman said the protective effects of cooling might be extended over longer periods, perhaps doubling the golden hour to 120 minutes. Experiments are under way to determine if the addition of certain drugs can provide additional protection for the visceral organs.

Safar maintains that controlled hypothermia is likely to continue to grow in value for many types of emergency care, just as it has already proven itself in brain trauma. In addition to use of mild hypothermia for hemorrhagic shock, Safar and Tisherman also are exploring the use of extreme cold to create a state of suspended animation in patients who suffer uncontrolled bleeding that would otherwise kill them in minutes.

In the field of resuscitation, Safar said, "Hypothermia is hot."