Pediatricians and pediatric ophthalmologists are concerned that there is a new epidemic of blindness and serious eye damage among babies who were born prematurely. But it is an odd epidemic, if indeed it is one at all, for no one knows how many children are affected nor what the chances are today that a premature baby will become blind.

This is in sharp contrast to most epidemics, such as an unexpectedly large number of influenza cases or the epidemic of acquired immune deficiency syndrome. In such instances, figures on the numbers of cases are readily available. But the blindess epidemic is different because medical scientists have been reluctant even to speak out about the disorder for fear of attracting the attention of lawyers and medical malpractice suits.

One study done at the University of California at Los Angeles suggested that of the 20,000 in- fants who weigh less than 3.3 pounds at birth in the United States each year, about 2,000 would be partly blind and more than 500 would become legally blind. But those estimates are based on births at one only hospital, so no one knows whether they really can be used as an estimate for the entire country.

The story of blindness among premature infants is a story of the uses and abuses of the charge of medical malpractice. And it is a story of the helplessness and frustration faced by physicians who care for very tiny and very sick babies who seem doomed to develop eye problems if they survive at all.

Late last year, a group of pediatricans, ophthalmologists, obstetricians and other medical scientists gathered at the National Institutes of Health for a symposium sponsored by the National Children's Eye Care Foundation to discuss what to do about what appears to be an epidemic of baby blindness.

When a baby is born prematurely, its eyes are still developing, and blood vessels continue growing into the retina. The retina is the innermost coat of specialized nerve cells at the back of the eyeball that capture light rays and convert them into electrical impulses that can be carried by the optic nerve to the brain.

These blood vessels grow from the center of the retina to the outside edges, and their growth in premature infants sometimes is disordered.

The vessels can proliferate wildly and can bleed. Scar tissue may form, causing the retina to buckle and, eventually, roll up like a scroll. Once the retina detaches, the child is blind.

In the mid-1950s medical researchers thought they had solved the problem of this eye disorder, which is called retinopathy of prematurity, or ROP. (This disorder was called retrolental fibroplasia until a few years ago, when the name was changed to show it is primarily a disease of prematurity.)

A large NIH study had shown, apparently, that oxygen was the culprit. At that time, it was standard medical practice to give all premature babies high concentrations of oxygen for 28 days, whether they needed it or not. The NIH study showed that if the babies were given oxygen only when and if they needed it, the numbers that got ROP greatly diminished.

So physicians stopped giving oxygen routinely and, said Dr. Jerold Lucey, a pediatrician at the University of Vermont, they all thought the problem had gone away. "Everyone thought, 'ROP is gone forever, and when it does occur, it must mean we are doing something wrong.'

"But then we would see these cases and we would know we were not doing anything wrong and the cases would keep coming."

No one publicized the continued existence of ROP cases, Lucey said, because "it's not something you're proud of."

But gradually, as physicians spoke to each other, they began to realize that ROP seemed to be back in force. And excessive use of oxygen, even though it may have been a leading cause in the 1950s, is not the answer today.

"We are very uncertain of what really causes ROP," said Dr. Burton Kushner, an ophthalmologist from the University of Wisconsin in Madison. "But it is clearly not just a matter of too much oxygen."

As smaller and smaller babies are kept alive, doctors find they must give some oxygen to most of them. They give the minimum amount for the shortest possible time. However, said Lucey, "if the sick eye is a reflection of a sick baby, as every study seems to show, I don't think we can do much about ROP in the next few years."

It is not, however, as though no one is trying. One controversial hypothesis after another is advanced, and highly speculative treatments are being offered, but it is not clear whether the treatments work, because the data on their success rates is poor.

For example, some physicians advocate giving high doses of vitamin E to premature babies in order to prevent ROP. The idea is that vitamin E, which soaks up extra oxygen molecules and highly reactive forms of oxygen called free rad- icals, might prove protective if indeed oxygen is the source of the problem.

But high doses of vitamin E can lead to other difficulties. The vitamin can cause an erosion of the lining of the intestines, for example, leading to serious infections. Or it can cause gangrene of the bowel.

So far, said Dr. John Flynn, an ophthalmologist at the University of Miami School of Medicine, the studies of vitamin E have not been promising. If it does have an effect in preventing ROP, it is a small one.

"I'm very skeptical," Flynn said.

Among the treatments for ROP are cryotherapy -- a technique of destroying the outer part of the retina by freezing -- and microsurgery designed to reattach the retina.

Cryotherapy, said Dr. Gordon Avery, a pediatrician at Children's Hospital in the District, is a last-ditch effort to prevent retinal detachment. The idea is that since the blood vessels are growing wildly toward the outer edge of the retina, perhaps they will stop if the outer edge were destroyed.

So far, said Flynn, there are 50 articles on cryotherapy in the scientific literature, but "they are all poorly done. There is not a good one in the bunch."

And there are risks. The cold treatment could destroy a part of the eye that is necessary for vision, such as the optic nerve. The National Eye Institute is initating a large study at 24 hospitals to determine whether this treatment helps.

Once the retina detaches, the only hope is a delicate and controversial surgery to attach it again. This surgery, said Dr. Gary Abrams, an ophthalmologist at the Medical College of Wisconsin in Milwaukee, "is the most difficult retinal attachment in ophthalmology."

And even if the surgeon attaches the retina successfully, there is no guarantee that the child will see. Frequently there are other problems with the eyes that preclude vision. There are no good figures on the success of the operation.

But even when a procedure is still experimental, malpractice lawyers can be expected to try to use it to prove that physicians are negligent, participants in the NIH meeting said privately.

For example, Penny Glass of Washington's Children's Hospital recently did a preliminary study suggesting that some cases of ROP might be prevented if the babies were shaded from bright lights.

Medical researchers generally agree that the study results are intriguing but by no means conclusive. Yet ophthalmologists at the meeting said that as soon as the study was published, they were visited by malpractice lawyers wondering if their clients might be blind because they were exposed to bright light.

Wisconsin's Kushner said he saw three families and an attorney asking about the results within the first three weeks after the study came out. Kushner had recently told the attorney that, in his opinion, the attorney's clients had no cause for a lawsuit. But the attorney had returned to Kushner on the chance that he could use the light results to demonstrate negligence.

And attorneys can easily find doctors to testify that the mystery of ROP was solved 30 years ago:excess oxygen causes the disease, and the disease is entirely preventable.

Flynn, who has done an informal survey of ROP lawsuits, said half of them were won by the plaintiffs, often on the grounds that the doctors were negligent in administering oxygen. The settlements, said Flynn, are enormous, averaging nearly $5 million per child.

"It really is disturbing," Flynn said. "The system is not working when physicians and hospitals that are rendering the highest levels of care are subject to these suits. And it is the best hospitals and the best nurseries in the country that are involved, because those are the places caring for the sickest babies."

According to attorney Marshall Simonds of the Boston firm Goodwin, Proctor and Hoar, a 50 percent success rate in malpractice cases is unusually high. Simonds attributed it to "the uncertainty in the profession coupled with the dramatic impact of a blind baby." He said that people today expect compensation, whether or not injuries could have been prevented.

One of the consequences of the ROP lawsuits is that medical investigators have been reluctant to even publish figures on the incidence of the disorder.

"It's a funny business because everyone's a little ashamed of having incidence figures at all," said Lucey. And although the physicains think there are extraordinarily many cases of ROP these days, "I don't believe there are any figures that allow us to establish the epidemic we all think we are seeing," Lucey said.

Most physicans tend to keep quite about ROP, Abrams said. "They think it is safest not to talk about it and attract lawyers' attention. Out of sight, out of mind."

Even basic research on the disease has been hindered by the constant fear of lawsuits. "It has not been a popular area for research," said Lucey. "People think there is something wrong when they have ROP cases." No one wants to publicize failures.

No answers to this problem are forthcoming. Simonds urged the physicians to publish in the medical literature statements saying there is no standard of care for very premature babies and, short of preventing premature births, there is no known way to prevent ROP. Lucey is confident that physicans are coming out of the woodwork and are at least talking to each other about their problems in preventing ROP.

Still, Lucey said, "I'd love to see a national effort to get a registry of ROP cases."