A majority of U.S. physicians responding to a survey say they "favor strongly" withdrawing life-support systems from hopelessly ill or irreversibly comatose patients, and most say they have faced such situations.

"Withdrawing treatment sounds harsh, but it can be a a benevolent act," said Dr. Kenneth Vaux, a professor of ethics at the University of Illinois Medical School in Chicago.The survey, conducted by the American Medical Association, asked doctors: "Would you favor or oppose withdrawing life-support systems, including food and water, from hopelessly ill or irreversibly comatose patients if they or their families request it?"

Fifty-eight percent of those surveyed answered "favor strongly," 20 percent "favor," 5 percent "oppose," 10 percent "oppose strongly" and 7 percent "unsure."

One thousand randomly selected physicians in various specialties responded to the survey, which was published in the AMA weekly, American Medical News.

Of those physicians, 67 percent said they have been directly involved in treating a patient where the issue of refusing or withdrawing life-sustaining treatment arose.

"While physicians should never directly cause death, they must always act in the best interest of their patients, and that sometimes includes allowing them to die," said Dr. James H. Sammons, executive vice president of the Chicago-based AMA.

Dr. James Childress, a medical ethics professor at the University of Virginia, said he perceived a shift toward considering cost when deciding whether to continue life-sustaining treatment.

"During the time of the Karen Ann Quinlan case in the mid-1970s, there was a greater risk of overtreatment. But now with pressures of cost containment, patients may be in danger of undertreatment," Childress said.