Muddled AIDS Lab Tests Stir Confusion and Alarm

Reports written by the staffs of laboratories that test for the AIDS virus are often so muddled they confuse even physicians, according to a new study that suggests the garbled lab results may be alarming the uninfected and delaying treatment of the infected.

The study, by California and federal researchers, found that laboratories frequently misstated the significance of their results, used outdated interpretive methods, failed to do recommended confirmation testing and even misidentified the virus.

“Certainly, egregious examples of double talk such as those noted in the (study) should be eliminated,” two public health experts remarked on the study’s findings, published today. But they added, “Sadly, this may not be so easy.”

The problem stems in part from a lack of uniform standards on how tests for the human immunodeficiency virus (HIV) should be used and interpreted by private firms. Without clear guidelines, some public health officials say, companies have had to improvise.

Federal health officials moved this week to address that problem, publishing guidelines for interpreting one of the most common tests. The guidelines appear in the Journal of the American Medical Assn., along with the laboratory-testing study.

“However, since there are no federal agencies that can demand implementation of these criteria, adoption of these changes may be slow,” observed the authors of the laboratory study, headed by Dr. Abram S. Benenson of San Diego State University.

The importance of accurate and clear HIV testing has increased in recent months as new drugs have become available to treat HIV infection. Last year alone, laboratories performed between 23 million and 32 million HIV tests, federal officials estimate.

The most common test is the so-called enzyme immunoassay. But, because it is known to produce some false positive results, public health officials recommend that any positive finding be confirmed with a more sophisticated test called the Western blot.

But even the Western blot test can be ambiguous. The test searches for nine viral proteins. But until recently, there was no consensus on how many proteins would justify a definitive positive finding. The lack of clear standards, therefore, has resulted in a high number of indeterminate results.

In the case of indeterminate findings, labs in California can--but do not always--forward the blood sample to a state lab for one of several other tests. Those tests may well produce a definitive result, or a new blood sample can be taken and tested.

In the study published today, the researchers asked hospitals and clinics in San Diego County to send out three blood samples for testing. They knew that one was negative and one was positive; the third was negative but carefully selected because it produced an indeterminate result.

Problems Abound

The researchers then reviewed the reports sent back by the testing labs for accuracy and potential sources of confusion. They found many problems, “some of which compromised the value of the testing process and might have led to improper patient treatment.”

Benenson, head of the division of epidemiology and biostatistics at the university’s Graduate School of Public Health, said that almost every one of the 13 labs studied produced reports that were unclear in some way.

The researchers immediately sought publication of their findings. “It was so striking that as soon as we saw the results come in, we thought, ‘My God, we can’t sit on this,’ ” Benenson said.

The 13 laboratories, considered representative of labs in general, were located throughout California and in two other states.

“We expect that (the problem) is not restricted to Southern California,” Benenson said. “We think that the problem is probably a general problem.”

Among the shortcomings cited by the researchers:

- Five of the labs reported that the indeterminate specimen was actually positive; five reported it as indeterminate. One lab left the determination up to the physician, simply describing its lab results and giving the physician the criteria by which to measure them.

- Only one of seven reports of indeterminate results suggested the use of an alternative test, even though such a test is available through state laboratories and is used successfully to arbitrate ambiguous cases.

- One lab received the indeterminate specimen from four different hospitals and clinics. To two of them, it reported that the blood had tested positive for infection; to the other two, it reported that the results were indeterminate.

- One report failed to identify the virus by its accepted name--human immunodeficiency virus--using instead two terms no longer in use. Another report employed criteria for interpreting lab tests that the researchers said have been out of use for three years.

- Three of the labs reported that a positive test result indicates that the patient has been “exposed” to the virus--rather than what the researchers said is the correct statement, that the person is “infected.”

- One report of a positive finding was followed by a lengthy, boilerplate statement on how to interpret a negative finding. “Was this serum result positive . . . or was it negative, as the wording implies?” the researchers wondered.

- One lab reported a positive result without doing any confirmatory tests.

Dr. Michael S. Ascher of the state Department of Health Services, a co-author of the paper, said he knew of two cases in which people had been wrongly told that they were infected on the basis of tests that were never confirmed.

One involved a child whose family learned two years later that the child was not infected. The other, Ascher said, involved a 50-year-old woman who learned recently that her initial test was wrong when it found her infected four years ago.

“I think it’s difficult to measure those kinds of incidents,” said Dr. D. Peter Drotman of the federal Centers for Disease Control. “Someone might not suffer physically . . . but they may go through days or weeks of severe emotional distress.”

In an editorial accompanying the article, Drotman and another Centers for Disease Control official wrote: “The observations (in the paper) . . . may represent yet another impediment to timely medical intervention for treating HIV infection.”