Most women would have been thrilled to have received the report Nancy Seeger got last September from Oncormed, a genetic testing company then based in Gaithersburg. Seeger's mother and maternal aunt both had died of breast cancer, but the company's test of Seeger's blood indicated that the Bethesda-born woman had not inherited any of the common genetic mutations that would predispose her to the disease.

The problem was that Oncormed had come to a different conclusion eight months earlier, when Seeger's doctor first sent her blood to the company. At that time, Oncormed had told Seeger she harbored a mutation that could place her at a greatly increased risk of breast and ovarian cancer. As a result, and after consulting with her doctor, Seeger quickly underwent surgery to remove her ovaries.

Now the company was writing to say the first result had been a mistake.

"I was shocked, of course; then I was relieved, then I was horrified," said Seeger, who lives in Evanston, Ill. "Everybody had believed that this information was so true, and I'd had this surgery based on it, and I'd lived for the past eight months with this horror and anxiety and all the trauma of how to tell the kids. I felt like I'd been forced to confront my own death prematurely."

Seeger is one of a growing number of people in this country learning the hard way that the rapidly expanding field of genetic testing is not everything its gleaming molecular image would suggest.

Genetic tests -- in particular those that can predict a person's medical future -- are technically more difficult to perform and trickier to interpret than conventional medical tests. And they can have a far more profound impact on people's lives, exacting intense psychological tolls, disrupting family relationships and unfairly jeopardizing employment and insurability.

Yet genetic tests are subject to virtually none of the formal oversight required of standard medical tests. The Food and Drug Administration, which approves the safety and effectiveness of other blood tests, has opted not to regulate genetic tests such as the one Seeger took, saying it lacks the resources. And the primary piece of federal legislation that ensures a certain level of expertise and quality control at the nation's 170,000 medical laboratories does not have provisions for genetic tests.

That means that commercial biotechnology companies and other laboratories, which already are offering more than 400 genetic tests to a largely uneducated public, are free to decide how accurate their tests need to be before marketing them and what they will say in their advertisements to doctors and others.

As a result, some people are paying hundreds or even thousands of dollars apiece for tests that either give wrong answers or are medically useless because they are still experimental.

"Many of these tests are in the quasi-research category," said Edward R. B. McCabe, physician-in-chief of Mattel Children's Hospital at the University of California at Los Angeles, "but everybody wants to charge for them."

McCabe is chairman of a new advisory committee to Health and Human Services Secretary Donna E. Shalala that is to decide by the end of the year how to ensure the quality of genetic testing in this country. The committee faces the enormous challenge of devising a system that will protect people such as Seeger without stifling the nascent genetic testing industry, which could bring many medical benefits in the new millennium.

Unlike conventional medical tests, which help doctors diagnose diseases by looking for biochemical or cellular abnormalities, genetic tests probe for tiny alterations in crucial genes. The information they provide can predict, with varying degrees of reliability, a healthy adult's odds of succumbing to a deadly disorder later in life, in some cases providing an opportunity to prevent the disease. The tests also can detect genetic traits in embryos or fetuses, and thus can influence a woman's decision on whether to have an abortion.

Most genetic tests probably are being conducted skillfully and ethically, but there is evidence of shortcomings in how labs perform the tests and how doctors interpret their results, committee members said at their first meeting earlier this month. One recent study, for example, found that 15 percent of 245 genetic testing laboratories surveyed scored lower than 70 percent on a quality assurance scale, "suggestive of substandard laboratory practice."

Meanwhile, the field is growing exponentially. Some 4,000 genes already have been linked to hereditary medical problems, and new tests are emerging every week. Yet nobody knows exactly how many are on the market or how many people are taking them -- much less how many are providing wrong or misleading results.

It won't be easy to create a system of oversight, however. Many genetic tests predict the odds of contracting a disease over the next decade or longer, and companies say they cannot afford to wait that long to confirm that their predictive numbers are correct before marketing them. But how reliable and clinically useful should a genetic test be before it is offered to the public? And what agency or professional group should be responsible for enforcing those criteria?

"We can't really depend on the marketplace alone for responsible introduction of new genetic tests," said Francis S. Collins, a member of the HHS committee and director of the National Human Genome Research Institute, the federal entity spearheading an international project to identify every gene in the human body.

"The power of a predictive genetic test to alter a person's medical and life decisions is very strong," Collins said, "so in some ways the tolerance for error should perhaps be smaller than for other tests."

Seeger wishes it were so. The writer and communications consultant was 14 when her mother lost a gruesome, six-year battle against breast cancer. When her aunt died of the same disease a few years later, Seeger shouldered the burdensome reality that her family history placed her at high risk.

So it was with a mix of hope and trepidation 18 months ago that Seeger, then 54, took advantage of a newly developed genetic test that could tell whether she had inherited a predisposition to breast cancer. Her doctor sent a blood sample to Oncormed (bought last year by Gene Logic of Gaithersburg), and three weeks later the response came back.

"DNA analysis indicates the presence of [a] mutation" in the BRCA1 breast cancer gene, the report said.

The precise risk conveyed by that mutation is uncertain, the report continued. However, it stated: "The lifetime risk to develop breast cancer in women carrying a BRCA1 mutation, in general, may be as high as 85 percent and the risk for ovarian cancer 50 percent." That compares with an average lifetime risk of about 12 percent for breast cancer and about 1.7 percent for ovarian cancer among all American women.

"I went into shock. I was stunned. I started to cry," Seeger recalled. "I was very, very depressed and conflicted about what to do and who to talk to and what to say and what not to say."

Seeger and her doctor decided that careful screening with mammograms might allow her to find emerging breast tumors before they became deadly. But no test can reliably detect ovarian cancer early. So three months later, she had her ovaries removed.

Oncormed's error probably would never have come to light had Seeger not donated blood to a local university hospital for genetic research. That lab could not find the mutation and contacted Oncormed about the apparent contradiction. In September, the company mailed Seeger a letter saying it had sent her "an erroneous result." A $350 refund check for the cost of the test was included, "as a small token of our concern."

Mistakes happen with conventional blood tests, of course. But companies that develop and sell those tests must prove to the FDA that they only rarely give the wrong answer and that the information they provide is useful. What's more, laboratories that conduct those tests must prove their proficiency on a regular basis.

By contrast, the only genetic tests the FDA regulates are so-called "kits" that are mass-produced for use in commercial laboratories and that have not yet been developed for most genetic diseases. The agency neither assesses nor approves any of the hundreds of "home brew" tests that most genetic laboratories develop and use, such as the breast cancer tests that Oncormed conducted.

Similarly, the Clinical Laboratory Improvement Amendments of 1988 require that medical laboratories wishing to retain their licenses correctly analyze blood and tissue samples sent by a central testing authority. But there is no nationwide proficiency testing or even licensing process for laboratories that perform genetic tests.

"We're talking about people having treatments they may not need, or not having a kid that would have been fine, or getting a false sense of security that you or your kid won't get a disease," said Lori Andrews, an ethicist at the Chicago-Kent College of Law.

Seeger is negotiating with Oncormed's former insurer about the test error, and officials at Gene Logic, which assumed Oncormed's liability when it bought the company, said they could not comment on the talks. But the case illustrates many of the troublesome issues in genetic testing.

For one, genetic information can be extremely disconcerting, because it is often probabilistic and uncertain -- "85 percent odds of getting breast cancer" is scary but something short of a diagnosis -- and because in many cases there is little one can do to prevent the disease.

For another, the information can lead to even healthy people losing their insurance. And genetic information can cause intense family strife. Seeger worried that she may have passed the mutation to her 18-year-old daughter, but couldn't bear to tell her at first. When she told her son, he had to deal with keeping a critical secret from his sister.

Negative test results can lead to just as much trouble as positive ones, as an Illinois incident recently revealed. That case, which has been wending its way through the courts since 1992 but came to light last year, involves a healthy couple whose first child was born with cystic fibrosis, a fatal lung disease.

The parents sought to make sure in advance that their next child would be unaffected by the disease, which occurs in one of four offspring conceived by couples that carry mutations in the cystic fibrosis gene. So they signed up at a Chicago fertility center to create multiple test-tube embryos by in vitro fertilization and had each tested for the gene defect.

Only those embryos that tested negative were to be transferred to the wife's womb. But nine months later, she gave birth to a child with cystic fibrosis -- apparently the result of an erroneous test result on one embryo.

That case is still unsettled, and the couple is not speaking to news organizations, said Jeffrey B. Sussman, the couple's lawyer in Chicago.

Separate from the question of test accuracy is the issue of accuracy in advertising. At least a few testing companies have been less than honest in their assessment of who stands to be helped by their tests, the certainty of their results and the balance of potential risks and benefits of getting tested, some HHS committee members said.

How best to balance the interests of gene-test companies, doctors and patients? There are two basic oversight options available for endorsement by the HHS committee, experts said.

One is to have the FDA regulate genetic tests. The agency has experience overseeing the safety, efficacy and marketing of tests. And it could demand that companies wishing to market their still largely experimental gene tests continue to collect data for many years, so the tests' usefulness can be determined over time.

Many commercial testers and others oppose that option, however, out of fear that a rigid system of pre-market testing and approval might slow the field. They say that professional organizations, such as the American College of Medical Genetics, should set guidelines for the use of genetic tests (it already has for some) and that the Clinical Laboratory Improvement Amendments should be beefed up to cover genetics labs.

On one point, at least, everyone seems to agree: No system of oversight will work without a large-scale effort to better educate doctors about genetics. A 1995 survey found that one-third of medical schools still did not require course work in genetics. And a 1997 study of 177 patients who underwent testing for an inherited colon cancer gene found that doctors misinterpreted the results about one-third of the time.

Patients, too, are going to have to learn that genetic tests are not perfect predictors of the future.

"The way people perceive these things is almost mystical," said advisory committee member Elliot D. Hillback, senior vice president of Genzyme, a biotechnology company in Cambridge, Mass. "Unless we demystify this, we can beat our heads against the wall with regulations and other things forever and still not be able to help people."

Detecting Disease

Scientists have developed more than 400 tests to detect genetic mutations associated with hereditary disorders. Most conditions are extremely rare. Here are some of the more common ones for which tests are relatively widely available:

Breast cancer

Five to 10 percent of all cases, and some ovarian cancers, are caused by defects in the BRCA1 or BRCA2 gene.

Cystic fibrosis

An inherited salt imbalance that causes poor lung function and other metabolic problems.

Huntington's disease

A neurological disorder that causes severe dementia, movement problems and death in adulthood.

Duchenne's muscular dystrophy

Causes a progressive decline in cardiac and skeletal muscle function in boys.

Hemophilia A

A clotting disorder that can lead to uncontrolled hemorrhaging.

Achondroplasia

A growth disorder that results in dwarfism.

Tay-Sachs disease

A fatal childhood neurological disorder that emerges during the first year of life.

Fragile X

The most common congenital cause of mental retardation.

SOURCE: GeneTests: Genetic Testing Resource and staff reports

CAPTION: Nancy Seeger, shown on recent camping trip, had her ovaries removed after getting what turned out to be erroneous results from a genetic testing company.

CAPTION: After Oncormed told Nancy Seeger that tests showed she had inherited a genetic mutation that put her at risk for breast and ovarian cancer, she had her ovaries removed. The company later told her that it had made a mistake.