To Test, Or Not to Test

By Jamie Talan

Newsday 20 Jan. 2000

Doctors say quick screening now used is too often wrong, may do more harm than good

The young mother had just given birth at Montefiore Medical Center in the Bronx. The newborn had been carefully placed in breastfeeding position to obtain his first ounce of nourishment when Dr. Ellie Schoenbaum, an infectious disease specialist, entered the room.

"This baby may have been exposed to HIV," the doctor told the young mother, explaining that the rapid screening test she had consented to during labor had come back HIV positive. Schoenbaum explained that it was imperative that the baby be treated immediately with the powerful antiretroviral drug AZT to reduce the risk that the infant would become infected as a consequence of its exposure during delivery. Confirmatory testing would later show that the mother was indeed HIV positive.

But Schoenbaum and other health professionals say they have become concerned about how to handle such cases, in which pregnant women arrive at the hospital in labor without a record of their HIV status.

Under a state health department regulation implemented in August, hospitals and birthing centers are required to ask such mothers to consent to a rapid test during labor. They may refuse, but their infants' blood will then be tested at delivery.

What health professionals have found is that the rapid HIV test, which provides results within 30 minutes -- the only quick HIV screening assay approved by the Food and Drug Administration -- is often wrong.

Based on the sometimes false-positive results of this test, women are requested to begin taking infusions of AIDS medicines to prevent the transmission of the virus to their unborn child, and the antiretroviral AZT will also be administered to the baby once it's born. It can take up to 48 hours for results from further blood tests that are required to confirm the diagnosis in the mother. Given the unreliability of the rapid screening test, health professionals are questioning the mandated procedure, occurring at the highly emotional time of labor and delivery. They are also concerned about exposing healthy infants unnecessarily to the powerful drug AZT, whose possible longterm effects remain unknown.

"The timing of this couldn't be worse," said Mayris Webber, an epidemiologist at Montefiore. "I am concerned that this is a psychologically vulnerable moment for the mother."

Webber and Schoenbaum are embarking on a study at Montefiore and a handful of other New York City hospitals to test the feasibility and effectiveness of the rapid testing program. But doctors witnessing the statewide program in full-swing say the number of false positives returned from the rapid test, called SUDS or Single-Use Diagnostic System, is much too high.

"A woman who hears that she may be HIV-infected when she's in labor or first bonding with her baby may forever link the receipt of this horrible news with the arrival of the baby," Schoenbaum said. "For the mother who is false positive, will she ever be free of the nagging doubt that something is wrong with herself or her baby? I'm not saying that we shouldn't go all out to prevent perinatal HIV transmission, but we need to be sure we have the necessary support in place for these potentially

traumatized families. At present, we don't."

Without therapeutic intervention during pregnancy and in the first hours of life, one in every four babies born to an HIV-infected mother will harbor the deadly virus. If the HIV-infected mother receives AZT during pregnancy, as is the recommendation for these patients, and the baby receives AZT during the first six weeks of life, the odds of becoming infected drop from 25 percent to around 9 percent.

While Montefiore's Webber points out that "three-quarters of the babies exposed to AZT would not have become infected" even if their mothers were positive for the virus, most doctors and patients have gotten over this hurdle with the realization that the benefits to those who would have become infected outweigh the risks of a relatively short course of AZT therapy for mother and child, or the child alone. (Adults can refuse treatment.)

But what about mothers and babies who test positive but in fact don't have the virus in their system? With the use of the popular rapid HIV test, Webber argues, "you don't even know for a fact that the mother is positive. This is a major ethical problem. Would I want my [healthy, uninfected] baby exposed to AZT? Absolutely not."

According to Dr. Guthrie S. Birkhead, director of the state health department's AIDS Institute, statistics from the first three months of the mandatory testing program -- from Aug. 1 to Oct. 31 -- show that at least 17 women, 24 percent of the total number of preliminary HIV-positive tests collected by the state health department's Wadsworth Laboratory in Albany -- had received a false-positive result. Birkhead's report, issued last month and sent as a "Dear Colleague" letter to doctors in the state, said that 13 of these 17 newborns were started on AZT until subsequent testing of the mother's blood, which can take a day or two, proved the initial test was wrong.

And a preliminary study conducted earlier this year at Montefiore found that the false positive SUDS rate was so high that the hospital decided to forgo using it, opting instead to order an ELISA screen that can take about four hours.

At the heart of the problem is testing. SUDS is a screening test, and not a good one when you are testing a broad number of people who fall outside of high-risk groups, which for HIV include homosexuals, IV-drug users and their partners, experts say. SUDS is a blood screening test that measures the body's antibody response to HIV. How well the test performs is based on a mathematical equation. A test has a standard sensitivity and specificity. It will identify a certain number of true positives and a certain number of false positives. If the test is normally wrong 1 percent of the time and you have a high-risk population, you will have fewer false positives. But in a nonrisk population where the percentage of true positives is much lower, then the number of false positives will be much higher.

The Montefiore researchers found that when given to women outside of the high-risk group, the SUDS test was wrong two out of every three times. Once the initial SUDS test is positive, another is often ordered. But even a repeat SUDS test could be incorrect. The next step is the ELISA, another antibody-screening test. and a result could take up to four hours. Again, experts say there is a risk of a false positive when testing a broad population of people. If the ELISA comes up positive, a confirmatory test called a Western Blot is ordered. This is now the gold-standard diagnostic test, and a positive result, which can be available within 48 hours, normally means that the person has HIV.

State health officials are so concerned about the problem of false positive tests that they requested the FDA allow them to use several unapproved quick assays that would lower this false positive rate and spare a healthy child from spending its first 48 hours on AZT until confirmatory tests are done.

"This situation is not ideal. But we are trying to prevent HIV transmission, which is a lifelong problem," Birkhead said. "It [the false-positive results] is not something that we want to have happen. We are urging the FDA to let the state use these unapproved assays." FDA officials say they are actively working on this problem to expedite testing of these assays.

There are no figures on how many pregnant women or their newborns across the country have received a false positive HIV test -- or more than one. The majority of pregnant women receive the test much earlier in pregnancy, well before they enter the hospital to give birth. In the privacy of a prenatal visit, rapid tests are generally not done. The ELISA test is ordered from a lab, and if it tests positive, a Western Blot is automatically done before a person is even told the status of the tests.

Problems emerge, however, for women who are not properly counseled during pregnancy or for those who do not seek prenatal care. Preliminary results from the Bronx study suggest that, depending on the population of people, as many as 40 percent of pregnant women arrive at the hospital without having received an HIV test during the pregnancy.

Webber and her colleagues have just begun a study to figure out how many false-positive results actually occur during mandatory HIV testing of pregnant women during labor and delivery, and whether the test results are obtained fast enough and newborns started on AZT early enough to prevent the spread of HIV from mother to child. They will also follow mothers to see whether they continue providing their children with

AZT for the six weeks that scientists have determined is the optimal dose to prevent transmission.

So far, 84,200 pregnant women have been tested since the law went into effect in August statewide. More than 13,500 women arrived at hospitals in labor only to be told they would need to be tested for HIV infection. Of around 44 positive test results, 26 were confirmed on a Western Blot to be positive. Seventeen of the confirmatory tests were negative, which means that neither the women, nor their babies, were infected.

The SUDS blood test is performed manually and the reading of the result is highly subjective, says Sara Beatrice, director of retrovirology and immunobiology at the New York City Department of Health. Room temperature can also skew a result, timing is key, and reading the small color-changing tab, much like a home pregnancy test, can lead to many of the errors.

In a study conducted by Beatrice and her staff, 45 percent of the SUDS specimens sent in by area hospitals turned out to be false positive.

"It's important to have the information, but we need the right information," Beatrice said.

Some worry that the number of false-positive tests will increase as more states seek mandatory testing. Now, all states except for New York and Connecticut (which adopted a mandatory program in October), have voluntary testing laws.

But sometimes, a voluntary program can itself exert a great deal of pressure.

Kathleen Tyson, a 40-year-old mother from Eugene, Ore., says that she is happily married and doesn't have any risk factors that would have led her to believe she has HIV. When her midwives recommended it as part of a routine pregnancy screen in 1998, she said sure, why not. Both the ELISA and Western Blot tests came back positive.

"It was a terrible shock," recalls Tyson. "I was worried but I have always been so healthy. It took a couple of weeks to sink in."

The midwives urged her to seek medical help, and Tyson, wanting the best for her unborn child, and not knowing a thing about HIV infection, entered her second trimester on a cocktail of three AIDS drugs. She was also urged to question her husband about hidden drug problems or homosexual encounters. (Soon after, her husband tested negative for the virus, and so did her 11-year-old daughter who had been nursed for the first three years of her life.)

After six weeks of antiretrovirals and an equal amount of time digging into her strange, new diagnosis, Tyson and her husband agreed that she would stop taking the medicines. There were no signs she was ill, and tests for viral markers of HIV were all negative.

The rest of her pregnancy was uneventful. On Dec. 7, 1998, Tyson's son Felix was born. Delivery room doctors asked whether they should start intravenous AZT. Tyson said no.

"Everyone seemed to respect my decision," Tyson said. An hour later, a nurse approached her and asked whether the baby was going to receive AZT. Again, she said no.

The next day, an infectious disease doctor begged the mother not to breastfeed and to begin AZT therapy for the child immediately. "We explained our position, but she felt that she was right, that she was the expert," Tyson recalls. The doctor returned a few hours later and explained that she would have to start the child on AZT or call the hospital's attorney and the head of the ethics board.

That evening, Tyson's husband returned to the hospital to find security guards standing at all exits on the maternity ward. A petitioner from juvenile court arrived with a uniformed police officer and said the hospital had reported she was endangering her child. While she initially had refused to stop breastfeeding, she told the petitioner that she would formula feed Felix. Two days later the couple would be forced to appear before a judge, and the state would assume custody of her child. "A case worker recommended that our son continue to stay with us," Tyson said. "But I promised the court I would not breastfeed and would give my son AZT."

Felix went through a six-week course of AZT. Every week, a social worker would arrive at feeding time to ensure that formula, bottles and medicine were in sight. By April, another hearing was held and Tyson's lawyers brought in Dr. Roberto Heraldo of New York University School of Medicine to discuss the inadequacies of HIV tests. Tyson had agreed that her son Felix could be tested four times -- at birth, four weeks, four months and one year. Tests like the ELISA and Western Blot pick up maternal antibodies in the newborns for at least six months, and therefore even a positive test doesn't prove that the child has been infected with HIV, experts say. Each time, Felix' tests came back negative. In December, the state's custody case against Kathleen Tyson was closed.

"My hope is that no mother goes through what we did," said Tyson, who continues to feel perfectly healthy. She has worked closely with the International Coalition for Medical Justice, an organization started to bring attention to problems with HIV test results. "Do I have HIV? That is the most ridiculous thing ever considered. None of this makes any sense."

The goal of testing during pregnancy is to find newborns at risk for infection and prevent transmission, explains Dr. Robert Boorstein, a pathologist at Bellevue Hospital in Manhattan. "These tests are part of a process." He said the hospital uses the rapid SUDS test, and provides the service to area hospitals. Bellevue screened 3,000 blood samples and identified 26 positives on the SUDS screen. Half of those results turned out to be wrong on confirmatory tests, he said, but "If you miss treating people who are positive, you will not have another opportunity to treat them."

Christine Ginocchio, director of the microbiology unit at North Shore-Long Island Jewish Health System, doesn't agree that it's worth the risk of the high false-positive results from the SUDS screen. She has decided not to use the SUDS test. Instead, her lab runs ELISA and immediate confirmatory Western Blots. Of the 100 positive tests from all patients during the past six months, only

two were found to be false positive. In those who are pregnant and delivering, the results are returned with enough time to start the baby on AZT, if a result is positive. "Are you going to tell a brand new mother that she may have AIDS when she doesn't? From a clinical and emotional point, this is not acceptable."

Ginocchio helped spearhead a program to educate the physicians in the North Shore-LIJ system so that they test their pregnant patients well before they end up in the delivery room. At the start of the mandated testing in August, less than half of the system's doctors had provided HIV testing during prenatal visits. Now, 96 percent of the doctors in the North Shore-LIJ system say that they test women early in their pregnancy.

Compared with the city, the suburbs have relatively few cases of pediatric HIV. Nevertheless, Dr. Boris Petrikovsky, chairman of obstetrics and gynocology at Nassau County Medical Center, which uses the SUDS test, said that it's "to everyone's advantage to err on the false-positive side rather than miss an HIV individual. What other choice do we have now?" According to a hospital spokeswoman, there have been no false-positive SUDS cases there since the mandatory program went into effect in August.

Boorstein argues that the state's commitment to child health is worth taking the risk that a few children will receive unnecessary doses of AZT. One of the major side effects is anemia. Doctors don't think that a few days or weeks of antiretroviral treatment will be detrimental, but possible long-term effects are not known. Some new research suggests that AZT may damage mitochondria, the energy packets within each cell, but this has not been proven.

"But we don't have a conclusive answer," said Dr. Bernard Branson, a medical epidemiologist at the U.S. Centers for Disease Control and Prevention. He is aware of the problem of the false-positive SUDS results and is now testing other rapid screening assays with hopes that the findings will expedite the federal approval process for a number of assays used outside of the United States.