It is an irony of our age.

Millions of American in hospitals -- late-stage cancer patients, burn victims, accident victims and those recovering from surgery -- suffer unnecessary, sometimes agonizing, pain because their physicians are needlessly cautious about prescribing narcotics.

Millions more unhospitalized -- the numbers cannot be measured accurately -- are dangerously overdosing on painkillers often inappropriately prescribed for their chronic pain from headaches, backaches, pinched nerves and arthritis among other disorders.

These are two main conclusions from last week's National Institutes of Health consensus conference on the management of pain.

Many pain experts, including those at NIH, have speculated for years that as many as one third of the American population -- perhaps 70 million people -- suffer chronic pain severe enough to interfere with normal functioning. But the panel concluded that not enough hard scientific research exists to warrant using even that figure, much less speculate on the numbers abusing prescription drugs.

In its final statement, the panel said that "over and above . . . the limitations posed by current knowledge and technology, many informed observers, supported by some scientific data, perceive continuing deficiencies in the clinical management of pain.

"Surveys of hospitalized people with acute pain," the panel wrote, "have found that many of them continue to have moderate or severe pain because their treatment has been with doses of narcotic analgesics that have been too low" or were not given frequently enough.

At a press conference after the consensus statement was accepted, Dr. Ronald J. Dougherty, medical director of a chronic pain outpatient clinic and administrator of a drug abuse recovery service in Syracuse, N.Y., said that he objects to the "cookie cutter" approach to pain -- treating all patients the same -- in hospitals, but that on the other hand, "there is no safe prescription drug for the treatment of chronic pain."

In the hospital, the panel said, physician fears of addicting patients and, in some cases, of depressing respiration, has led to the conservative use of narcotics. Yet, said Dougherty, "in 17 years of practice, I have never seen an acute pain patient addicted.

"As I tell my medical students, I don't think it makes any difference even if we use heroin for treatment of acute pain. It doesn't matter what drug we use, because we generally treat it less than five or six days and you don't get addictions from that.

"I believe doctors should give acute pain and cancer pain patients whatever that patient needs to relieve his or her symptoms, regardless of how little or how big." Dougherty said.

Many patients continue to suffer because "doses of narcotic analgesics . . . have been too low."

-- National Institutes of Health Consensus Panel

However, he said, one of the biggest problems seen "in the 500 or so pain clinics across the country are people suffering not only from pain, but from prescription drug overuse."

"When the effect of the drug are becoming more crippling than the effects of the pain, it's time to do something about it." he said. Signs of drug problems include social withdrawal, appetite changes and behavioral changes even more intense than similar symptoms caused by the pain.

He mentioned specifically drugs such as Darvon, Percodan and Percoset and Tylenol-3 (with codeine) as often abused drugs for chronic pain. "Many people we see use up to 60 Darvon a day,'' he noted. "These may not be the right drugs for a specific chronic pain, but may help a little at first. Then, in frustration, as the pain increases, the patient, to deal with the pain, will take more and more.

"They don't know any better," said Dougherty, and "unfortunately their physicians who were not confronted with this in medical school don't know either."

"There are no good controlled studies," he said, "but there is definite evidence that people are overtreated for chronic pain, and there is definite evidence that people die from overusing prescription drugs."

Dr. Donald S. Kornfeld, professor of psychiatry at Columbia University College of Physicians and Surgeons and chief of psychiatric consultation-liaison service at Presbyterian Hospital in New York, said that the in-hospital narcotic apprehension stems from a physician obeying the Hippocractic admonition "first, do no harm."

And the other side of the coin is the desire to help -- the patient comes in and says 'I'm still suffering.' and the physician needs to feel he is helping, and this increasing pressure to provide help can be linked to the misapplication of narcotics."

The consensus panel, headed by Laurel Archer Copp, dean of the University of North Carolina School of Nursing, included, in addition to pain specialist Dougherty and psychiatrist Kornfeld, a neurologist, a clergyman, a dentist, two oncology nurses, a pharmacologist, a biostatistician, an anesthesiologist and several general nurses.

They heard testimony about many nondrug options for treating chronic pain urging more research into such techniques as biofeedback, acupuncture, the small electricMARTINELLI,COPY,SY,ACT,COPY,,,pack called the TENS (transcutaneous electrical nerve stimulation) machine, hypnosis and other approaches. (One witness urged that everyone be made aware that TENS, often used useful in treating musculo-skeletal pain, sells for about $600 in this country but can be purchased from overseas manufacturers for about $100.)

Oncology nurses have also made great strides in use of drug therapies, Copp said, ''achieving great innovations, tailoring drugs to an individual patient and the patient's individual pain by using different drugs . . . adjusting doses, trying new routes of administration, noting what works and what does not work.''

''Their work with physicians is truly collaborative,'' she said.

Another area, the panel said, where there is a serious lag in pain research and management is that dealing with pain in children.''

In their statement the panel wrote, ''the assessment of pain in children presents special problems and is a subject of current research interest. Clinical impressions suggest that children in pain may frequently be undertreated.''

And, one witness said, only about a third of pain-killing drugs have been tested for their effects in children.

Panelists agreed that the subject of pain is ''vastly complex,'' in part because pain varies from patient to patient. They wrote that changes in pain in the cancer patient probably have more to do with progression of the disease than tolerance to a drug, and urged that health care educators and workers combine forces to learn to evaluate each pain patient.

One new technology appeared to offer some hope for the estimated 4 to 6 million post-operative patients whose pain is now inadequately treated. That is the technique known as patient-controlled analgesia (PCA).

With this device a patient has an anesthetic available through a catheter fixed in a peripheral vein. Pain medicine may be injected as the patient feels it is needed.

According to Dr. Emanuel M. Papper, a professor of anesthesiology at the University of Miami School of Medicine, PCA ''is very popular in the Unite Kingdom and continental Europe, and is beginning to be used here.''

Physician fears that it would lead to overdosing have been so far not found to be a problem overseas, he said. Studies have suggested that patients are not only more comfortable with PCA, but tend to use less analgesia.

It is, wrote the panel, ''one of the innovative ways that may provide effective individualized analgesia and comfort.''