When Maureen Flannery, a 33-year-old doctor who trained at Yale University School of Medicine, returned home to practice medicine in Appalachia a few years ago, she hadn't expected to confront a major ethical dilemma so quickly. Deciding in advance whom to save and whom to let die was an issue for big-city teaching hospitals with advanced life-prolonging technology, hardly the concerns of a doctor making house calls in a pickup truck through the hills of southeastern Kentucky.

Then, late in the evening last Oct. 26, after finishing her rounds in a bleak nursing home near the coal town of Hazard, Flannery was discussing her patients with a night nurse. The subject of resuscitation came up. In the past two years, the nurse said, six to eight patients had died while she was on duty. Some she had attempted to save; others she had not.

The nurse said she usually relied on her own judgment, trying to save any patient she thought had a good chance of recovery. She mentioned one man she would resuscitate: "He's up here and around. He'd be productive." Others, she would not: "They don't speak. They don't move," she told Flannery. "To me that would be torment. I'd rather be dead."

Astonished by the informal manner in which the nurse made her choices, Flannery started naming her own patients. If any of them suffered cardiac arrest, whom would the nurse attempt to save? "I would," she said about one person. "I would not," she said about another.

Flannery was shaken when she left the nursing home. "I didn't think it was right to put a nurse, and especially a middle-of-the-night nurse who has never seen that patient during the day . . . in that kind of position," she said later.

Flannery believes such decisions should be made jointly by a doctor with the patient and the patient's family. "It's not that I shirk the responsibility. I just don't think that the physician alone has any business taking that kind of power," Flannery says. "I see my role as more supporting the family's wishes."

Flannery decided it was time the local doctors, all five of them, devised a clear policy for the nursing home--similar to those being drafted in urban hospitals--on who should be resuscitated and who should not.

"The problem here," says Dr. Robert Constantine, a 29-year-old U.S. Public Health Service physician who recently began working at Hazard Hospital, "is that this hospital, this town, is on the verge of going from early 20th century medicine to the forefront of medicine and doing what everyone else is doing."

Dr. Robert Mason has practiced medicine in Baltimore since 1945. "Thirty years ago," he says, "each physician was in essence in control of his patients, and very few people were looking over his shoulder to tell him how he managed his patients from the standpoint of the basic ethical issues: who he would struggle to keep alive, who he would not struggle to keep alive."

Today the debate over withdrawing life-saving technology has expanded beyond the medical community to include lawyers, judges, insurance companies, the government, and hospital administrators. Last month, a President's Commission report entitled "Deciding to Forego Life-Sustaining Treatment" recommended that hospitals develop clear policies outlining standards for making and reviewing such decisions, and that they give patients a voice in those decisions.

The Rev. Clyde Shallenberger, head chaplin at Johns Hopkins Hospital, says: "I see the real ethical problems in medicine to be four-fold: who lives, who dies, who decides, who pays."

"Clearly we have to confront the cost," says Dr. Howard Champion, head of the Washington Hospital Center's Medstar trauma unit. "The availability of this high technology is opening the door to a spectrum of possibilities that did not previously exist. Ten years ago a guy who was 95 and dying of congestive heart failure would have been allowed to die in peace. Now you've almost got to tattoo on your chest 'Just leave me alone.'

"The technology is being provided to the medical profession without stringent rules about when it should be used."

Doctors once were taught that money should not be a consideration, according to Dr. John Lynch, a cancer specialist at the Washington Hospital Center. "We were never to think about the cost if it was a question of making someone better. The cost be damned. It's not that way anymore . . . . If a cancer-riddled patient's family is telling me to do everything possible, am I supposed to chew up resources if it means one less kid gets fed in Washington?"

Lynch cites a $300,000 bill for an elderly woman in an irreversible coma who, at her family's insistence, has been kept on a life-support system for a year.

"The hospital can't pay that bill," Lynch says. "At the end of the year you figure out the cost, tell the insurance companies, 'This is what it's going to cost . . . . ' The hospital that is going to survive takes the hernias and sore throats, and sends the sick people somewhere else."

The technology itself has made Lynch apprehensive.

"If we don't control it, it controls us. Its mere existence causes a whole range of problems, almost like Frankenstein in a way. We created it, and it runs a lot of our decisions."

Dr. Nisha Chandra, 30, is head of the intensive care unit at Baltimore City Hospital. She is an expert on cardiopulmonary resuscitation (CPR), the highly effective chest-compression technique that has saved thousands of lives.

Chandra is part of a research team developing a new form of CPR that potentially could save many more people. Yet she sees an ominous aspect to her three-year, $750,000 research project: CPR can restore a patient's heartbeat, but the mind does not always follow.

"My deepest concern," she says, "is that yes, we're going to save more, but we're going to save vegetables.

"When I used to pull somebody back five years ago I'd say, 'Okay! All right! We got him back!' Now when I get somebody back, if I get them back after the first three minutes, I am not elated. I'm very cautious. Because I don't know what I've got back."

One day last year Chandra visited a man in his fifties who had been hospitalized with repeated bouts of near-fatal heart failure. The prognosis was hopeless. Chandra sat by his bed. "You don't smile the way you did," she said.

He looked her in the eye. "Doc, I hate living like this," he said. "This is not a life . . . . I can't even go and watch my son play. I don't want to live like this."

Chandra returned to see him the next day. While she was there, the man suffered another cardiac arrest. She shouted for the resuscitation team, and as a reflex action, started pumping on the man's chest. "I was about to put my face on his and start mouth-to-mouth and I stopped," she recalls.

"I stopped because I remembered him saying how sad he was. And I knew I could perhaps get him back because I was willing to pull all my research stunts on him . . . and I stopped, and I said, 'Get him back for what? Only so he can die again?'

"It was so difficult for me to stand there because his cardiogram still showed a little . . . . Surgeons came pouring in . . . . I wasn't in a white coat . . . . They said, 'Okay lady, you can leave now.' I said, 'Wait, this is my patient.' They said, 'Okay, get me a central line, get me this . . . . ' I said, 'Cool it.'

"And it was very, very, very difficult standing there and watching him die . . . . He was viable. All my instincts were to work on him, and I felt guilty . . . . It's a feeling of such utter frustration."

At the Washington Hospital Center, Lynch and a group of doctors have formed an ethics committee to review difficult cases and to offer physicians and families a second opinion that may help resolve disagreements over continuing therapy.

"It can be a very lonely world out there for a physician who's managing one of these situations by himself, and is facing all sorts of pressures in his own mind and pressures from the family," Lynch says.

He says that some families insist on continued life support because they hold out hope for a miracle. "I tell patients, 'Miracles don't need machines . . . . If we're going to have a miracle, we don't need the machine. We don't need the tubes . . . . We don't need any kind of treatment at all.' "

But disagreements can escalate into painful confrontations. Dr. Alan Fields, associate director of critical care at Children's Hospital, recalls one emotional father who demanded that Fields resuscitate his dying infant son over and over until Fields says he finally told the father: "I understand you need this to survive . . . but you're now going to the point where you're abusing him. A human being . . . has some dignity and that's where I draw the line."

At George Washington University Hospital, Dr. Christine Wanke once watched doctors pumping on the chest of a feeble, 80-year-old stroke victim in the waning hours of his life, bringing him back each time his heart stopped, possibly breaking bones or injuring organs in the process, pumping until there were no heartbeats left to squeeze out. "My feeling after that was that finally the poor man was allowed to die," Wanke recalls. "He had been trying and trying and trying, and no one would let him."

Wylene Larsen, a physician at the Georgetown University Hospital ICU, says: "We just torture people because we--the medical community--can't make a decision to let somebody die."

In the modern intensive care unit (ICU), where the legal definition of death involves only one factor--lack of brain activity--at times the heart will continue to beat after someone is declared dead.

Dr. Rick Munschauer, a Johns Hopkins neurologist, conducted the rigid test for brain death on a 63-year-old patient last November. He pronounced her legally dead, turned off the respirator, wrote the official time of death, "3:15," in the medical chart, and prepared to leave.

On the monitor above her head, her heart continued to register 106 beats per minute.

Asked how long he expected her heart to pump, he said: "I don't want to know."

Munschauer clicked off the monitor and pulled a curtain around the woman's bed. He would never be entirely comfortable with this process, but he had seen enough cases to know that the heart would stop on its own, as it did in this case.

As disconcerting as the new definitions of death may be, the termination of medical treatment for hopelessly ill patients is still more agonizing. "All through medical school we were trained to save, save, save," says Elizabeth Streeten, a 26-year-old Hopkins physician. "I still have not gotten used to letting people die."

Over time, some physicians have even more difficulty. "The older he gets, the more he associates the death of his patients with his own death," says 38-year-old GW physician Andrew Anderson. "If there is one chance, he'll take it."

Another traumatic step for a doctor is in fulfilling a competent patient's request to be removed from a respirator that is providing life. "It's one thing to talk about. It's another thing to do," says Dr. Jack Zimmerman, codirector of the George Washington ICU. "You take that machine off, you throw that switch . . . . "

Zimmerman recalls one such case when it took 10 minutes for his patient to die. "That man's last moments, until his blood-oxygen level fell and his carbon dioxide level rose until . . . he was unconscious, were terribly uncomfortable. I watched this poor man suffocate."

In 1978 in Fort Lauderdale, Fla., Dr. Marshall Brumer refused a terminally ill patient's request to be taken off a respirator.

"We're trying to help people, not trying to kill them," Brumer says. "By prolonging life, or sustaining life, how is that harm? . . . How do I know that in those pain-wracked months new chemotherapy will not come down the pipeline that will be used to fight cancer?

"When you're dead, you're dead a long time.. . . One of the Commandments says, 'Thou shall not kill.' One thing was certain: God was not a lawyer. He put that on a tablet of stone. There's not a modification or qualification or anything . . . . If you remove the life-support system from a patient who is alert you're killing him prematurely."

The patient's family ultimately sued for the patient's right to die, and a judge ordered Brumer not to interfere. The respirator was disconnected and the man died.

In the Washington Home Hospice on Upton Street NW, Dr. Joanne Lynn routinely gives morphine to ease the pain of dying cancer patients. One side-effect can be to depress the drive to breathe, which may hasten a patient's dying. In recent years Lynn has also used morphine on several patients suffering from "respiratory insufficiency," a terrifying, drawn-out suffocation process in which the patient struggles to breathe, gasping 50 to 60 times a minute. Unless treated with morphine, Lynn says, the gasping could continue for as long as a few days until the patient is overcome by exhaustion--and death.

Lynn says: "If you decide that the right thing to do is give enough morphine to depress the breathing--thereby relieving the suffering--then there is an almost direct correlation with an earlier death, as close to one-to-one as man gets, as close to a borderline case as I would want medicine to get to.

"When we talked with the patients and their families, they were all 100 percent in favor of giving the morphine, knowing it was likely to shorten the patients' lives. In the first couple of cases, they were pleading with us to do something before we got around to recognizing what could be done.

"Now we can be much more straightforward in presenting this option and helping patients and families in realizing the importance of this decision and the paucity of other options."

Lynn, who helped draft the recent presidential commission report, cites the section that says doctors may give morphine to such patients after having "considered the full range of foreseeable effects, have knowingly accepted whatever risk of death is entailed, and have found the risk to be justified. . . ."

Says Lynn: "After clear public debate, if the law said that giving morphine in these situations was illegal, then I think that would require me not to do it.

"It would be agonizing to all concerned, however, to know that the morphine is there in the cabinet but the law required us not to use it and the patients therefore suffered so greatly in their last hours."

The most difficult cases may be those involving infants. "The ethical decisions are daily decisions for me," says Children's Hospital surgeon Martin Eichelberger. "The baby can't make a decision ever."

At Children's Hospital, doctors withdraw medical therapy from about 35 to 40 infants each year who are dying and for whom no treatment is expected to reverse their condition, according to Dr. Anne Fletcher, director of the ICU for newborns. Fletcher says that accounts for about two-thirds of all the deaths each year in the ICU.

Before terminating therapy, the hospital policy requires a finding by a team of doctors of the "hopelessness of the child's condition," as well as parental consent. Social workers, nurses and doctors help families prepare for the death of a child.

"Sometimes," says Children's Hospital intensive care physician Alan Fields, "I engineer death. I absolutely do. I am an engineer of death. I time it. I delay it. I put it to the time of day you want, the hour you want, the minute you want. I get the people in who are necessary. When you talk about playing God--in that sense, it is to an extent.

"The family's got to survive. They've got to be comfortable. And that's the one thing I can do for them. I mean, I can't have the child survive, but the one thing I can do is at least make that most final, important act somewhat meaningful."

One case last year illustrated that no matter how carefully drawn, even the most well-intended rules can't cover every situation.

A boy was born prematurely and not resuscitated for the first 10 minutes of his life. He was taken to a local hospital, then transferred to Children's. During the time he remained there, breathing for three months on a respirator, he had brain damage, heart failure, lung disease, kidney failure, a hernia, rickets, and seizures. Doctors struggled to keep him alive, resuscitating him, performing surgery, keeping him warm, and nourishing him.

After three months, Dr. John Rose wrote in the medical chart that the baby would probably never breathe on his own, or be conscious of the world around him."

Fletcher agreed with the prognosis. She decided to seek the mother's permission to remove the baby from the respirator and allow him to die. But the mother was an unwed teen-ager who didn't want to see her son, much less make a decision about his survival, according to hospital records.

"Your child is critically ill," Fletcher said to her in a sternly worded telegram. "It is urgent that we see you. Until you free him for adoption you are the parent, and must contact us . . . . "

It took weeks of canceled meetings and unreturned phone calls before the mother agreed to see her son.

" The mother did ask what dying was like," Fletcher wrote in the chart. "She also expressed her fears of coming in, feelings of guilt that she was at fault for his being here."

The mother held her son as Fletcher disconnected the respirator. But instead of dying as Fletcher had predicted, the baby breathed on his own and continued to live. Several weeks later his confused and upset mother surrendered custody to a social services agency, telling doctors: "I don't want to fool with this anymore."

The baby survived without the respirator for three months. Late one night his tiny lungs and heart finally gave out. There was no attempt to resuscitate him. His six-month ordeal, and the hospital's, was over.

"To the people who say we are playing God," Fletcher says, "I think we are playing God if we keep on going . . . . We are being mischievous . . . . That is something I use with parents: 'Just to be alive because you have four tubes in you, and you're on a respirator , that's not living. We're just prolonging the dying process . . . . '

"I just read through the Declaration of Independence again. It says, 'All men are created equal and are entitled to life, liberty and the pursuit of happiness.'

"I guess I tend to feel that all people are not created equal, to some extent, if they are too small to be born, if they have multiple and severe congenital abnormalities, or are so sick on a respirator that we can't keep them alive. We've said they're created equal, and yet we can't do it. The courts have said we can't judge the quality of life, and yet we do," Fletcher says.

"Looking at a baby's death, it is peaceful. There is the fear that death is awful, and parents must have this fear . . . . So I describe to the parents what will happen: The baby's color will slowly change. It may have a few twitchings of the extremities. Suddenly he will be very quiet . . . . Sometimes, afterwards, the heart will go on for another 45 minutes, still beating, not functional. Certain infants we stop on, it may be days or weeks. So you can never predict . . . . ." In these cases, Fletcher says, nursing care and nourishment continue until a baby dies.

"There are times that I sit back and I have wondered, 'Is this the correct thing? Why shouldn't I think like a doctor who thinks that everybody should be kept alive? . . . . Am I being too quick? Am I expecting too much? . . . '

"But I have learned this--having watched my father and my mother die--that death is not bad for the individual who is dying. It's what's left for the other people to get through the grieving process."

Children's Hospital adopted its written "no resuscitation" policy (known as "No Code 99") in 1975. Legal counsel C. Richard Cornelius says the hospital wanted to reform the "secretive" manner in which "no resuscitation" decisions were being made.

"The old policy," Cornelius recalls, "was whispered in the hallways, so to speak . . . . You'd write it in pencil maybe, and then later erase it . . . . A lot of deception."

George Washington, the Washington Hospital Center and Children's Hospital all have or are developing such policies. The Georgetown University Hospital ICU has none.

Dr. Edgar Reed of the Veterans Administration, says "no coding" was banned four years ago at all 172 VA hospitals nationwide because he was concerned about the public perception. "I had fears," Reed said in a recent speech to a medical group, "that somewhere, sometime, some headline or '60 Minutes' report would say 'Veterans Administration Killing Veterans.' "

The prohibition spawned its own problems. Several doctors at the Washington VA Hospital say they sometimes disregarded the rule. There soon may be a change: Dr. Dorothy Rasinksi says a new set of guidelines permitting "no code" orders is being drafted for the VA. "I think it's a good way to go for modern American medicine," she says.

Medical practice in Appalachia takes you back 20 years, to a place where most people still don't want to die in hospitals, to the tiny towns of southeastern Kentucky such as Hazard, where the acceptance of "no code" practices has been slow.

After Dr. Maureen Flannery learned about the ad hoc resuscitation decisions being made at the nursing home, she proposed a formal "no code" policy for the nursing home, to place the decision where she thought it belonged--on doctors and families.

But Dr. Denzil Barker, 66, who has practiced in rural Kentucky for 35 years and is Hazard's senior physician, said it would not work. Some families were so guilt-ridden about sending relatives to the nursing home in the first place that raising the question of whether their loved ones should be saved or allowed to die would be insensitive.

Says Barker: "The feeling prevails here, perhaps on a wider scale than in urban areas, that sending a father, mother, an older person, a relative, to a nursing home is tantamount to abandoning them. A lot of people feel badly about it; won't do it except as a last resort . . . . Some would look upon it as abandonment of the patient if you didn't do everything possible to keep them alive--almost negligence possibly--in not making every attempt to resuscitate them."

Flannery understood Barker's objection. She, too, was concerned with the families as well as the patients. She was born near Hazard; her grandmother had been the local midwife; her father a revered minister. She shared the Appalachian values and traditions, the emotional attachment to the land, the church and the family. She knew the same strong ties that bind these families in life also bind them in death. Most of her elderly patients wanted to die at home.

Flannery and the other doctors agreed they should not publish formal guidelines for the nursing home. But they also agreed that such decisions were medical decisions, and should be made by doctors, not nurses. To avoid confusion, they would instruct the nurses to consult with the doctors when there was a question about resuscitation.

That night in a small town in rural Kentucky, a compromise was made between what seemed right and what seemed possible. And if the decision did not go as far as Flannery had hoped, it was still compassionate and fair. Certainly there would be problems with an unwritten policy, but at least it was a policy, and that was a start.