Anne Paxton
July 2022—Thanks to more than two years’ experience with SARS-CoV-2, perhaps at no point in U.S. history has the general public been as aware of antigen and PCR testing, and the difference between them, as it is now. SARS-CoV-2 has also raised the profile of emergency departments as critical access points to get patients with infectious diseases diagnosed and in treatment. Yet much remains to be learned about exactly how screening using PCR testing for infectious diseases can improve public health and what role the ED should play in ordering that testing.
Clinical research on nontargeted screening for HIV and HCV in the ED suggests that real-time PCR testing, ordered earlier in the standard screening algorithm for certain patients, can help EDs quickly identify people with those diseases and link them to care.
From May 2016 to February 2022, the ED at Tampa General Hospital, where Jason Wilson, MD, MA, is director of the ED’s clinical decision unit, ran an HIV antigen/antibody test on 99,691 ED patients, of which 1,725 were reactive. That was followed by HIV-1/2 Ab confirmatory testing and, if the first antigen/antibody test was reactive, by an HIV RNA PCR viral load test. He and his team deviate from the currently recommended algorithm by ordering HIV RNA PCR on anyone who is reactive on the first screen, regardless of what the antibody test reveals. “The patient is in front of us, we can get blood if we need it, but also we know 13 percent of our [reactive] patients are equivocal, and of those, one-third are acute positives who have negative or undetectable antibody,” Dr. Wilson explained at a conference this spring on “Advancing HIV, STI, and Viral Hepatitis Testing.”
While the clinical encounter in the ED lasts on average four to six hours, the HIV RNA PCR test has a 72-hour turnaround time, which sometimes means lost opportunity for care. “We’ve really begun to recognize that there’s a difference between screening for HIV or hepatitis C and testing for HIV or hepatitis C,” said Dr. Wilson, associate professor of internal medicine at the University of South Florida Morsani College of Medicine.
In a recent interview and in his presentation at the conference, sponsored by the CDC, Association of Public Health Laboratories, American Sexual Health Association, and American Sexually Transmitted Diseases Association, Dr. Wilson related his hospital’s experience with having the ED play a more effective role in screening for HIV and HCV and getting infected patients into treatment. He and his colleagues have found that an alternative testing strategy, using real-time PCR testing and obtaining results in 91 minutes rather than 72 hours, could mean greater likelihood of success in meeting that goal.
The CDC in 2006 revised its algorithm for HIV screening to a public health model with a multistep approach. “That approach works just fine in the public health clinic or a public event where you’re seeing lots of people and trying to screen as many as possible for a disease state,” Dr. Wilson says. But there is a growing view that emergency departments should be doing public health HIV screening on their patients as well.
The CDC in 2007 funded a small group of 10 emergency departments to implement its algorithm in the ED setting. But as Dr. Wilson was trying to implement the 2006 guidelines in the ED at Tampa General Hospital, there were a few obstacles. “One was that we still had laws in Florida around written consent and pre- and post-test counseling that made it difficult to do this type of work in the ER.”
The main problem was Florida’s opt-in requirement for obtaining consent for HIV testing. When the ED uses a consent form requiring patients to opt in to HIV testing—rather than including HIV testing among the general medical consent tests unless patients opt out—it greatly affects the number of HIV screening tests performed, Dr. Wilson notes. “I would have had to go into the room with a consent form and say to the patient, ‘I would like to test you for HIV. Here are the risks. Here are the benefits. Can you sign this form?’”
“It’s not that we couldn’t do that. But it creates a stigma, potentially, and it’s time-consuming. It carves out HIV as different from testing your white blood cell count or testing your BUN and creatinine.”
In Florida, opt-in was the law until 2015 when the state passed a statute dropping the requirement for separate written consent. After that, “We were able to move to notifying patients of an opt-out test and that’s what we do with the signs that are in every patient room and in other areas that are publicly accessible in the emergency department.”
Moving to opt-out testing was a game changer, Dr. Wilson says. “It allowed us to test for HIV as a nontargeted screening practice. We could start putting that strategy into our routine workflow.”
“We went from doing zero HIV tests in the ER for screening to a handful—five or six a week—to now more than 1,100 HIV tests per month consistently since 2016.” The ED conducts hepatitis C screening during the ED encounter along with HIV screening.
The ED integrated clinical decision support into its screening processes through its Epic EHR. “The way we got to nontargeted high-volume screening is we took the CDC algorithm and built it into the computer. So as that information is gathered during the patient’s encounter and entered into Epic, once the patient meets the screening criteria and has not opted out, a test order will pop up for the provider and say, ‘Please order the test.’”
Over time, the decision support has become more nuanced and more detailed, Dr. Wilson says. “For example, for the patient who has a nonconfirmatory antibody test for HIV, the viral load test order will automatically pop up as well. For a patient we identify as having IV drug use, we will want to order a hepatitis C RNA test because we have a number of patients who may have just acquired HCV and have no HCV antibodies but they do have detectable, quantifiable RNA, so we move right to RNA testing for them.”
Within the first couple of weeks of broad high-volume HIV screening, however, ED clinicians encountered something surprising. “We had had a couple of patient scenarios where the HIV antigen/antibody test was reactive, but then in one of those patients early on there was no HIV virus in the RNA and in the other one there was. So we realized we’re dealing here with clinically equivocal results during the clinical encounter.”
“They can go two ways: Either the person is an acute seroconverter and it’s important we intervene right now, or the person doesn’t have HIV but has a potentially transient p24 reactivity.” Over the past three or four years, Dr. Wilson says, “we thought about that population of people with reactive HIV antigen/antibody tests. Doing 1,100 tests a month, we have 1.73 percent seropositivity and 13.4 percent of those are equivocal.” It is one in every eight patients with a reactive HIV Ag/Ab, he says. “These numbers add up pretty quickly.”
Equivocal results can present a clinical dilemma for emergency medicine providers who see a patient for generally only a few hours. “Some people get admitted but we’re not as worried about the people who are admitted because they’re going to be sticking around,” he says. For those not admitted, “we may have lost an opportunity if we don’t know the true HIV status of that patient during the encounter.” But the other 68 percent of patients—the majority—will be p24 antigen positive for some other reason. If I were to result those patients during the ED encounter, I would likely give them undue grief and worry and my colleagues may not be willing to do the testing.”
He described a recent patient with a reactive HIV antigen/antibody result but a history and lack of risk factors that suggested she was not seropositive. “I would tell her we’re at a coin toss as to whether you have HIV or not; we’ll run the PCR and I’ll get you the result in another two days. But we can easily lose that patient into the transition space as to whether or not she was acutely seroconverted. So it’s a big game changer to know what the specific true diagnosis clinically is for that patient.” For positive patients, Dr. Wilson says, “we have a very good linkage to treatment rate.” But “we worry about these equivocals because we’re not telling them much of anything” before the typical patient encounter is over.
For the past year, on an experimental basis “in the background,” Dr. Wilson’s ED started ordering a GeneXpert real-time RNA PCR test for HIV as soon as a reactive antigen/antibody test result is known. Then the research team measured concordance between the GeneXpert qualitative HIV assay (not FDA approved) and the standard-of-care Aptima HIV-1 Quant assay on the Panther; calculated and compared turnaround time on the GeneXpert to the standard of care; collected GeneXpert actual run time metrics; reviewed ED length of stay and laboratory TAT data for the standard-of-care test; and calculated differences between GeneXpert and the standard of care
