College of American Pathologists
Printable Version

Putting some teeth into anti-HCV testing

April 2003
Feature Story

Karen Titus

Supporters of the CDC’s new guidelines for hepatitis C virus testing sound a bit like politicos when they speak on the topic. The thrust of their observations could fit smoothly into campaign slogans: Spend more to save more. Put the testing burden back where it belongs. End false-positives—and false hopes.

But unlike empty political rhetoric and certain fuzzy economic proposals, the new guidelines have some meat on them.

"It’s economic good sense; it’s medical good sense," says James Versalovic, MD, PhD, who helped draft the guidelines. "There are really no major disadvantages to this."

"It’s an important development," he adds. "But it’s one that could be easy to overlook."

There’s some basis for such concern. By and large, most agree, HCV simply doesn’t draw the same interest or serious level of discussion as HIV. Add to that persistent misunderstandings about the value of and need for confirmatory testing, and HCV antibody testing arrives at its current muddle.

The new guidelines, which appear in the Feb. 7 issue of "MMRW Recommendations and Reports" [52 (RR03); 1-16] (also available online at, take clear aim at that muddle.At their core lies a recommendation that anti-HCV testing include reflex supplemental testing by RIBA or nucleic acid testing, and, to reduce costs, that this testing be based on screening-test-positive signal-to-cutoff (s/co) ratios. For the two currently licensed anti-HCV enzyme immunoassays, the guidelines say that supplemental testing needs to be performed only for EIA-positive samples with ratios less than 3.8. By using these ratios, laboratories should be able to keep the need for supplemental testing to a minimum while making reported test results more reliable.

This option could substantially change laboratory testing practices, says Miriam Alter, PhD, the lead author of the guidelines and by some accounts the primary force behind their genesis. "Only a minority of laboratories provide any routine supplemental testing for anti-HCV," says Dr. Alter, associate director for science at the CDC’s Division of Viral Hepatitis. "These guidelines, we hope, will encourage laboratories that don’t do supplemental testing to adopt this alternative use of signal-to-cutoff ratios to improve the interpretation of the result."

That, of course, implies there is a problem with results interpretation—an implication with a fair amount of truth to it, says Dr. Versalovic, director of microbiology laboratories, Texas Children’s Hospital, and assistant professor of pathology, Baylor College of Medicine, Houston. The overall goal of the guidelines, he says, is to reduce the number of false-positive HCV test results being reported. The problem stems from the testing method. An initial HCV test is done using a simple ELISA method to screen for the presence of antibodies, says Dr. Versalovic, who is a member of both the CAP Microbiology Resource and Molecular Pathology committees.

"The problem is that many laboratories are reporting the screening result as positive or negative." By their very nature, however, screening tests are "very sensitive, but they’re not as specific as we’d like them to be. The problem then becomes, many people may get a positive test result but not have the virus."

That’s because the screening assays were developed without a so-called gray, or buffer zone, says Frederick Nolte, PhD, professor of pathology and laboratory medicine, Emory University, Atlanta, and director of the clinical microbiology and molecular diagnostics laboratories for Emory Medical Labs. This zone, which encompasses weakly positive results, isn’t acknowledged in package inserts. "So labs are left on their own on how to proceed," says Dr. Nolte. "It’s a simple matter, but a problem nonetheless."

Dr. Versalovic makes the inevitable comparison to HIV to explore the screening ELISA’s limitations. While an ELISA is also used to screen for antibodies to HIV-1, it’s rarely the last word. "As a rule, because HIV is such a sensitive issue, patients who are positive by the screening test have a followup immunoblot. No result is released without that," he says.

HCV is different, though it shouldn’t be. "It’s an important problem, and in terms of numbers is a much greater problem than HIV-1," says Dr. Versalovic. "And of course long-term there’s the risk of cirrhosis and liver cancer." Yet followup testing is uncommon.

Chalk it up, in part, to attitude. "I have participated in many discussions in which comments were made that getting a false-positive test for hepatitis C isn’t serious, or isn’t as serious as getting one for HIV," Dr. Alter reports. "That’s appalling. I think it’s appalling that anyone would have that attitude toward giving a patient a false-positive result, and then subjecting them to not only the psychological stress but also the expense of additional evaluation when it isn’t necessary."

The new guidelines could do much to quell such attitudes. They have roots in the blood bank community, according Dufour, MD, chief of pathology and laboratory medicine service at the VA Medical Center, Washington, DC, and professor of pathology, The George Washington University Medical Center. Blood bankers first reported the implications of weakly positive anti-HCV test results; namely, those with weakly positive results tend not to be HCV RNA positive or positive on a recombinant immunoblot assay, or RIBA.

These observations were largely ignored in the clinical setting, says Dr. Dufour, who helped write the CDC guidelines. "Probably the first time anyone paid any attention to this was at our institution, when my predecessor published an abstract in 1994 indicating that even in the clinical setting, those with weakly positive antibodies were typically negative when further testing was done.

"But even after that, nobody paid much attention to it," he says ruefully.

In the late 1990s, Dr. Dufour continues, the blood banks provided CDC with data showing that s/co ratios could be used to predict supplemental test results. In response, the CDC began assembling samples from various populations, including high-risk and low-risk individuals to evaluate the use of s/co ratios in other settings.

At about the same time—October 2000, to be exact—Dr. Dufour and his colleagues also began routine confirmation testing of all weakly positive results. They found that the vast majority-approximately 86 to 88 percent-of patients with weakly positive antibodies were negative on all confirmatory testing. (Their study is published in the March issue of Clinical Chemistry. 2003; 49: 479-486.)

Following up weakly positive anti-HCV results in this manner put Dr. Dufour’s laboratory in the minority. As the guidelines report, two 2002 surveys on anti-HCV testing practices—one of public health laboratories, one of VA medical centers—revealed the practice of confirmatory testing is rare. "Very, very few laboratories are doing confirmatory testing," Dr. Dufour says. "I would venture to say that in hospital laboratories, it’s even less common."

The CDC guidelines note that the majority of public and private-sector laboratories rely primarily on requesting physicians to order the appropriate tests and correctly interpret their results. The guidelines also suggest such reliance is not wise.

It’s a view shared by Dr. Nolte. "Labs, in a cost-conscious environment, are abdicating some of their responsibility for producing reliable results, in the false hope that the ordering physician will be knowledgeable about the test’s performance characteristics. And that may be true of hepatologists, but it certainly [isn’t] true of primary care physicians and other people who may be doing the screening," he says.

Hepatologists, in fact, may be among the few subsets of physicians who understand the nuances of anti-HCV testing well enough to argue that the CDC’s guidelines are unnecessary. "Gastroenterologists and liver disease specialists have a certain perspective on this," says Dr. Nolte, "and that perspective is, in a high-risk patient population, RIBA testing doesn’t add very much, because almost every time they see a patient who is EIA positive, the RIBA’s positive." This group of physicians, he says, will likely find the CDC guidelines "not very helpful."

"But that’s not the point," he continues. "The point is, there’s a lot of testing that goes on in other settings. When you have 1.8 percent of the U.S. population seropositive, some 2 million to 4 million people, you realize there’s a lot of testing that goes on for hepatitis C outside of the liver disease clinic." And, as the CDC guidelines point out, laboratories don’t know, on an individual patient basis, who is high risk and who is low risk. "The way the guidelines are constructed, they allow you to target the supplemental test to that group of specimens that would benefit from having the supplemental test done, regardless of the pre-existing risk factors or disease prevalence of the population you’re testing," Dr. Nolte says.

The developers of the guidelines assessed how they would work in various populations, with different disease prevalence. They also covered virtually every possible testing algorithm. "We didn’t want to launch a surprise, or issue recommendations in a vacuum. So we did a lot of work with everybody we could think of before we even drafted the guidelines, to deal with potential barriers," Dr. Alter recalls. "The last thing I wanted to find out after publication was that there was some barrier that couldn’t be overcome for some reason.

"But we didn’t find one," she continues. "We found difficult ones, but not insurmountable ones." The two biggest barriers were (1) reprogramming laboratory software to generate an appropriate s/co ratio, and (2) reimbursement. The CDC worked closely with anti-HCV test manufacturers on the former issue, and contacted the Centers for Medicare and Medicaid Services to explore reimbursement possibilities.

The guidelines explain that documenting physician ordering of the reflex supplemental testing—on which reimbursement may hinge—can be done via a printed requisition form that clearly identifies the level of screening-test results that will trigger supplemental testing, and what type of testing that will be. The guidelines also recommend that each of the supplemental tests offered (RIBA or nucleic acid testing) be listed separately, enabling clinicians to order them as needed.

Thanks to this advance work, the guidelines have been met with no resistance so far, Dr. Alter reports. "I’m hoping it’s because we did include everybody we needed to include, and that we gave them a lot of time to think this over." She and others are getting the word out via teleconferences and at annual meetings, and the guidelines are being widely distributed via mail and the Internet.

Dr. Nolte’s laboratory has already begun applying the guidelines. "We were one of the few laboratories that were confirming all of the positive screening tests with RIBA," he says. This gave them an opportunity to do a lookback study to assess the CDC’s s/co ratios. They looked at 1,000 test results, in a population with a 7.5 percent seroprevalence rate. Of the 75 positive EIAs, 65 had a signal-to-cutoff ratio of greater than 3.8; of those, 64 (98.5 percent) were RIBA positive. Of the 10 positive specimens with an s/co value of less than 3.8, 30 percent were RIBA positive.

Based on those results, Dr. Nolte’s laboratory adopted the guidelines and now performs RIBA testing only on samples with low signal-to-cutoff ratios. "As you can see, that substantially decreases the amount of RIBA testing that we’re doing. That’s a tremendous cost savings for the laboratory, and it doesn’t compromise our antibody results at all."

The laboratory report informs physicians that if a result has an s/co of greater than 3.8, it’s strongly positive for antibody to HCV. The laboratory doesn’t report the actual s/co ratio. It does note that greater than 98 percent of strongly positive samples are also positive on the supplemental RIBA. Strongly positive samples are followed up with a qualitative HCV RNA test, ordered reflexively, to distinguish active from resolved infection.

Samples that are weakly positive for HCV antibody are sent for supplemental RIBA. The laboratory informs clinicians that a qualitative HCV RNA test is recommended for all confirmed antibody-positive samples to distinguish active from resolved infection, but that such tests must be ordered separately. "If the RIBA comes back positive on that weakly positive screening test, we don’t automatically reflex to PCR—it’s too complicated to keep track of that in the laboratory to make sure it gets done effectively," Dr. Nolte says.

In Dr. Nolte’s laboratory, the s/co is determined manually. "It’s not a difficult calculation, and it’s not something that will present a problem for most laboratories. Remember, you’re only doing it for samples that are positive." Positivity rates rarely run higher than 20 percent, and more frequently they fall into the one to five percent range, he says.

The s/co can also be calculated automatically (depending on the instrument used) with either of the two enzyme immunoassays (Abbott HCV EIA 2.0, Abbott Laboratories; Ortho HCV version 3.0 ELISA, Ortho-Clinical Diagnostics). It is automatically calculated on the enhanced chemiluminescence immunoassay (Vitros anti-HCV assay, Ortho-Clinical Diagnostics). The CDC guidelines indicate that for the licensed EIAs, an s/co of greater than or equal to 3.8 would be highly predictive of the true anti-HCV status. For the CIA test, an s/co of greater than or equal to 8 was recommended.

The beauty of using s/co ratios is that it makes the process more efficient, says Dr. Versalovic. "What you’ve effectively done is increased the value of confirmatory testing. That s/co, if it’s high enough, is sufficient for confirmation without additional testing. Only the low positives are submitted for confirmatory serologic testing." The net result, he says, is an approach that doesn’t necessarily add another test, but rather improves the accuracy of current testing.

That’s where the spend more-save more theory comes into play. By increasing the efficiency and value of confirmatory testing, the guidelines should help curb or stabilize costs. "And the costs of unnecessary followup testing will drop," Dr. Versalovic says.

Dr. Alter sounds equally realistic as well as upbeat. She doesn’t expect the guidelines will be an overnight success. "But I’m hoping within six months to a year, it will really have taken hold," she says.n Karen Titus is CAP TODAY contributing editor and co-managing editor.

Karen Titus is CAP TODAY contributing editor and co-managing editor.