Feature Story

Rapid HIV test a weathervane for waived tests

April 2004
Karen Lusky

The Food and Drug Administration’s waiver of the OraQuick rapid HIV-1 antibody fingerstick test last year sent shock waves through the clinical laboratory community. Many feared the agency’s move signaled a crack in the regulatory dam that would bring a flood of other waived tests that could hardly be viewed as posing "no unreasonable risk of harm if performed incorrectly," as CLIA ’88 requires for the least regulated category of testing.

"It’s hard to imagine anyone could look someone in the eye and say a false-negative or -positive HIV test could not cause potential harm," says Jared Schwartz, MD, PhD, director of the Department of Pathology and Laboratory Medicine at Presbyterian Healthcare, Charlotte, NC, and a member of the government’s Clinical Laboratory Improvement Advisory Committee, or CLIAC.

Health and Human Services secretary Tommy Thompson pushed through the OraQuick waiver despite protests from CLIAC and the CAP. The government was in the market for a portable kit to test people at high risk for HIV in nonclinical settings where rapid turnaround of results is paramount. And OraSure Technologies’ OraQuick fit that bill, requiring only a fingerstick sample and anteing up an answer in just 20 minutes.

One year later, it’s still too early to tell if OraQuick’s CLIA waiver represents an exception driven by public health interests—or a harbinger of things to come on the waived testing front. Yet some industry watchers are now pointing to the waived rapid HIV fingerstick test as a "poster child" of sorts, as one observer put it, or a case study in the emerging trend toward finding ways to balance access to new waived testing technology with quality assurance and other safeguards.

The FDA approved OraQuick’s waiver status with a number of restrictions, so the test isn’t as waived as the category normally indicates. For starters, the FDA restricted sale of the OraQuick test to clinical laboratories, meaning any setting that has a CLIA certificate of waiver. And OraSure is informing all purchasers of the test that they must be CLIA waived at a minimum to use it.

"The rapid tests are also restricted to clinical labs with an adequate quality assurance program and where there’s assurance that operators performing the test will receive and use instructional materials," says Bernard Branson, MD, chief of the Centers for Disease Control and Prevention’s Laboratory Determinants and Diagnostics Section in the Division of HIV/AIDS Prevention, Surveillance, and Epidemiology. At HHS’ request, the CDC developed a model quality assurance plan for the OraQuick test (www.cdc.gov/hiv/rapid_testing) that labs or other sites using the test can adopt.

Furthermore, "CDC has taken steps to ensure that recipients of its programmatic money for HIV testing receive training in how to properly perform rapid HIV testing," says past CLIAC chair Toby Merlin, MD, associate director for laboratory medicine in the CDC’s Division of Laboratory Systems Public Health Practice Program Office. OraSure Technologies also works with OraQuick purchasers to help them coordinate a training plan suited to their level of sophistication and needs, a company spokesperson says.

The FDA restrictions on rapid HIV tests appear prescient of CLIAC’s new recommendations for categorizing and regulating waived testing. The committee endorsed the proposal at its February meeting and was finalizing it at CAP TODAY press time. The FDA has agreed to use the recommendations as a framework for developing a clear regulatory process for waived tests.

"The new CLIAC waiver proposal would allow the FDA to waive some tests that have more criticality to them," says Carolyn Jones, spokesperson for AdvaMed, which represents device makers’ interests. In such cases, however, "the manufacturer may have to engage in some after-market activities to ensure the quality and performance levels of the testing are maintained," she adds. "The proposal also encourages more shared responsibility among manufacturers, the government, labs, and accrediting organizations for post-marketing activities, such as training."

Might the FDA impose the same restrictions on other high-stakes tests as it did on the OraQuick rapid HIV test? Steve Gutman, MD, director of the FDA’s Office of In Vitro Diagnostic Device Evaluation and Safety, says he doesn’t know "how and if the restrictions will be applied to the new waiver guidance for manufacturers and/or regulations about the waiver process. But they are certainly interesting and worth considering," he adds.

CLIAC’s new proposal recognizes there are important tradeoffs to consider in determining waiver of tests, Dr. Merlin says. "Glucose testing is an excellent example. We know that waived glucose testing is associated with some errors. But expanded availability of glucose testing permits diabetics to attain better control of their disease, with improvements in morbidity and mortality."

The argument that less-than-perfect waived testing is better than none in some cases carries special weight when battling HIV. "If you have a population where 30 percent of people don’t come back to get their results, as the CDC says is true for people testing for HIV in certain settings, then you can never have a sensitivity [for any test] that’s greater than 70 percent," says Sheldon Campbell, MD, PhD, a member of the CAP Point-of-Care Testing Committee and clinical pathologist at the VA Connecticut Healthcare System, West Haven, Conn.

In other words, "If you have an inferior test that is only 90 percent sensitive—and OraQuick is much better than that—then you’re going to be doing better than using a test [with 100 percent sensitivity] where 30 percent of people don’t hang around for the result—no matter how perfectly accurate that result is," Dr. Campbell says.

The OraQuick fingerstick test is highly accurate: 99.6 percent sensitive and 100 percent specific, according to published studies submitted as part of its premarket approval. "In studies done in other places, OraQuick was shown to have a sensitivity at least equivalent to the enzyme immunoassay [the test’s reference method]," the CDC’s Dr. Branson says. "The studies also showed more false-positives with the EIA than with OraQuick."

OraQuick also has an internal anti-IgG control so that if the person performing the test does not add the specimen, the test provides an invalid result.

In terms of accuracy, the CDC’s post-marketing surveillance conducted over the past year shows that OraQuick has a strong track record. But the CDC has also found that training nonlaboratorians to do the test helps to improve their performance. According to a CDC study, nine percent of people with no laboratory background who performed the test after simply reading the manufacturer’s written instructions made a mistake in the procedure. They got an invalid result thanks to the test’s fail-alert mechanism, which means they would not have reported a false result to the patient. "Yet 97.2 percent of those who performed the test correctly interpreted the result correctly," Dr. Branson says. "After viewing a demonstration on how to do the test, only 0.6 percent of people doing the test got invalid results, and 99.7 percent of the tests were interpreted correctly."

The CDC has, however, found shortfalls in the confirmatory testing for reactive OraQuick results. (Negative results generally require no additional testing.) "The QA protocols for confirming a reactive test result from rapid HIV screening aren’t new, and they are quite explicit," Dr. Branson says. "EIA testing is optional after a reactive rapid test result, but regardless, the lab has to do one of the two approved confirmatory tests: Western blot or immunofluorescent assay." In the March 19 Morbidity and Mortality Weekly Report, the CDC emphasized that "reactive rapid HIV tests must be confirmed with WB or IFA, even if a subsequent EIA is nonreactive."

For various reasons, labs are not following the CDC’s recommended protocols for confirmatory testing after a reactive rapid test, Dr. Branson says. "In some cases, the lab may not have a mechanism to identify a specimen that had already tested reactive with a rapid test. But some private commercial reference labs are deferring to their own algorithms, testing specimens that have shown to be reactive on the OraQuick first with an EIA. If the EIA is negative, they do no further testing. CDC also had discussions with some state labs that had taken that position."

The CDC continues to perform post-marketing surveillance and has documented at least four instances where OraQuick was reactive, EIA was negative, but Western blot was positive or indeterminate. The agency believes there may be other instances where people who are HIV positive have been misdiagnosed as negative due to false-negative EIAs.

The solution to the problem, in Dr. Campbell’s view, is for "any place that uses OraQuick to have a pathway for directly confirming results that uses a Western blot or IFA."

Even with the restrictions on OraQuick’s waiver and CDC post-marketing surveillance efforts, the jury is still out on how the test will perform over time, given the logistics and costs of training nonlaboratorians in how to do the test correctly and the small number of inspections of waived labs.

"Yes, OraQuick performs well in a limited number of published studies, or the studies wouldn’t be published," Dr. Campbell says. "But whether that performance will carry over to practical settings over time is quite unclear."

"Waived tests are prone to procedural drift where people start cutting corners a bit here or there," he adds, "and after two years, what they are doing has no relationship to what they are supposed to be doing. It takes supervision and competence assessment to keep people doing the test correctly."

Yet the Centers for Medicare and Medicaid Services inspects only about two percent of CLIA-waived labs. "CLIA ’88 doesn’t provide for routine inspection of waived laboratories," says the CDC’s Dr. Merlin. "Because of concerns raised about the quality of testing in some waived laboratories, however, CMS has begun a pilot program of inspecting a limited number in order to assess the quality of testing." These inspections are generally intended to gather data and educate these laboratories about quality assurance of waived testing.

Dr. Campbell says: "The argument by proponents of waived testing is similar to the one used by the National Rifle Association about gun control legislation. The NRA says that it’s not that there aren’t regulations—the issue is that no one enforces the existing regulations. Similarly, if there were enforcement provisions for waived testing, the existing regulations would be adequate."

The flip side of that argument is that complying with regulations is burdensome and can chill health care providers’ willingness to do higher-stakes waived testing. "And the whole point of the waived category is to increase access to testing" where needed, Dr. Campbell says. "So what we really need is good research that examines the tradeoff between access and enforcement. At this point, we really don’t know what that cost-benefit tradeoff is."

The cost-benefit ratio of the waived rapid HIV test continues to unfold as clinical sites grapple with the realities of using the test. For example, because of its high volume of diagnostic HIV testing, Chicago’s John H. Stroger Jr. Hospital of Cook County is reserving rapid testing for source patients involved in employee needlestick injuries, says Kathleen G. Beavis, MD, chair of the hospital’s Divisions of Microbiology and Virology and vice chair of the CAP Microbiology Resource Committee. "Our virology laboratory can get all the HIV test results—which average 75 a day—in one EIA run. From a workflow perspective, it would increase the time to results if our routine HIV testing converted from one EIA run to rapid testing," Dr. Beavis says.

Stroger Hospital’s emergency department has, however, been using the OraQuick test with much success under an investigative protocol. "We have trained two health educators who have never even worked in health care to do the test and obtain informed consent and perform phlebotomy," says Robert Weinstein, MD, chair of the hospital’s Division of Infectious Diseases. "They have been doing the test exceedingly well, and the results are very accurate." In Dr. Weinstein’s view, OraQuick isn’t much different from a home pregnancy test.

The CDC is finding through routine screening in some emergency rooms that about two to three percent of people who agree to be tested have undiagnosed HIV infection. And many are in the later stages of disease and therefore qualify for antiretroviral therapy.

"In one study performed in an emergency department setting, the CDC found that 76 percent of people diagnosed with HIV through initial rapid HIV testing have entered into care—that is, they were eligible for antiretrovirals, and not on the basis of early therapy," Dr. Branson says. "And we definitely have evidence to support that people who know they are HIV-positive are much less likely to spread infection—and that people on therapy are probably less infectious."

OraQuick’s waived status certainly makes it easier for nonclinical testing sites to perform the test, "especially given that state regulations are, in some cases, more stringent than CLIA in areas such as personnel requirements for performing moderate-complexity tests," Dr. Branson says. "And it’s much easier for a test site to get a certificate of waiver than certification for moderate-complexity testing."

Ironically, however, a few states’ regulations governing waived or rapid HIV testing practically preclude using OraQuick in the way public health initiatives envisioned. For example, New Jersey recognizes only eight waived tests originally identified by CLIA as meeting waived criteria. "New Jersey clinical laboratories cannot perform any tests subsequently added to this category [including OraQuick] without possessing a clinical laboratory license," says Sindy Paul, MD, MPH, medical director of the Division of HIV/AIDS Services for the New Jersey Department of Health and Senior Services.

But the New Jersey health department is now proposing new regulations that would authorize the state to issue a limited-purpose lab license to department-sponsored, nonprofit, freestanding counseling and test sites to conduct rapid point-of-care HIV testing. In the meantime, rapid testing is available at five of the state’s publicly funded counseling and testing sites that have received a clinical laboratory license to conduct OraQuick rapid testing.

Illinois state law allows health care providers to perform the waived rapid HIV testing but prohibits them from reporting a positive result without confirmatory testing (ultimately Western blot), Dr. Beavis says. "If the result is negative, we can release that," she says. Since the FDA recently approved OraQuick for testing of venous samples, health care sites can now simultaneously draw a venous specimen for rapid and confirmatory testing, rather than having to do a fingerstick and phlebotomy. "But venous samples are more complicated to obtain in a clinic setting than the fingerstick," Dr. Beavis says.

The CDC’s initiative to use the OraQuick as a point-of-care waived test to give people immediate access to their possibly positive HIV status may give states the impetus to reexamine regulations that impede that goal, Dr. Beavis says. In terms of the Illinois regulations, "it comes down to which mistake is more harmful: withholding a positive result until it is confirmed, thus allowing someone to continue infecting others, or releasing a false-positive unconfirmed result and causing significant mental anguish."

The problem of false-positives was resolved at a consensus conference in 1998, Dr. Branson says. "People acknowledged that false-positives might occur after a rapid screening test and have a negative effect on people. Yet that disadvantage is offset by the benefit of so many more HIV-positive people receiving their test results after rapid HIV testing than when they had to return for a second visit for results."

Because it recognizes the potential emotional ramifications of positive HIV test results, Hope Action Care in San Antonio, a nonprofit HIV/AIDS social services agency, is reserving the use of OraQuick for people at high risk for HIV. "That includes people who inject drugs or who have multiple sexual partners," says project director Debbie Heil. Even then, Hope Action Care has found that some high-risk people are not emotionally ready to learn in the span of 20 minutes that they are HIV-positive. "They have too much fear and, therefore, need to get used to the idea that they could be positive," Heil says. In such cases, the person receives oral HIV testing, which takes about a week to get results, or the standard confirmatory testing, where results come back in about two weeks.

Sister Mary Joan Smith, RN, liaison nurse for Penn Community Infectious Diseases at Presbyterian Medical Center, Philadelphia, agrees. "It is very dangerous to provide test results to people without assessing their ability to cope and their available support systems," she says. When Smith used the rapid HIV test in a recently concluded pilot program in the inpatient and outpatient drug-treatment programs at Presbyterian Medical Center, she wouldn’t even begin the testing process before asking the patient: "Who in your circle of friends or family might you talk to about this if the test is positive? Who is there to help you with this?"

Yet Smith also encountered unexpected benefits to the rapid testing for people in recovery from substance abuse and addiction. "The CDC says about one-third of people don’t return for test results; I’ve worked with some of those people and they are very high risk for HIV," she says. "As a result, many of them have had that cloud hanging over their head—the fear of their HIV status—for many years. And testing negative gives them a whole new lease on life in terms of their recovery and their perspective on the meaning of their lives.

"Some of the people I tested didn’t consider themselves high risk," she says, "when in fact they were. One of them was a woman who said she wouldn’t have been tested for HIV if it hadn’t been offered at the drug and alcohol inpatient unit. She was HIV-positive and was linked to care at our infectious disease clinic and put on antiretroviral medication. Her viral load has dropped from 35,000 to less than 75 in three months."

Time will tell to what degree waived rapid HIV testing affects patient outcomes and public health. "In a few years, after there have been some studies, we may see this as a great advance-or it may prove to be a bust or not used quite the way it was envisioned," CLIAC’s Dr. Schwartz says. "But we must keep in mind not to attack the test. The test is a great tool. Our problem now is to figure out how to use it most effectively and to make sure no one is hurt by the way the tool is being used."

In Dr. Campbell’s view, "The key to getting maximum performance out of the new waived tests is to do them right, store them properly, perform quality control, and throw them away when they are outdated." And use them only in settings where rapid turnaround is truly required, he adds. "You use them only when the accessibility to the testing buys you something over central lab testing that is likely to be more accurate as well as significantly less expensive in materials and labor."

For example, he says, if you are caring for a woman in labor, you want to know her HIV status before she delivers the baby, when antiretroviral therapy offers the best chance of preventing maternal-infant transmission. "And if the results [using rapid HIV testing] are slightly less reliable," he says, "you have to live with that, because if you wait to get a perfectly accurate answer after delivery, you’ve lost your window for treating the baby most effectively."

Karen Lusky is a writer in Brentwood, Tenn.