'Opt-Out' HIV Tests Have Modest Success: The study "demonstrates mostly the limitations of opt-out testing rather than its benefits," according to Michael Saag
Published: July 18, 2010
VIENNA -- The CDC's suggested routine "opt-out" testing for HIV modestly increased the number of people identified with the virus when it was used in a large emergency department.
But more than three-quarters of eligible patients, in fact, opted out of the testing, probably because they did not think they were at risk for HIV, according to Jason Haukoos, MD, of Denver Health Medical Center, and colleagues.
And most of the new cases identified by the system involved relatively advanced disease, they reported in the July 21 issue of the Journal of the American Medical Association and at a press conference held here in conjunction with the International AIDS Conference.
* Explain to interested patients that although the CDC recommends HIV screening in healthcare settings, patients can opt out of it.
* Note that this study suggests that the standard opt-out procedure may not be picking up people in the early stages of HIV infection.
Those results suggest that other methods may be needed to find people in the early stages of HIV infection, Haukoos and colleagues wrote.
In 2006, alarmed by estimates that as many as one-quarter of all HIV infections remain undiagnosed in the U.S., the CDC recommended routine nontargeted HIV screening in healthcare settings, including emergency departments, in places where the prevalence of undiagnosed infection is 0.1% or greater. (See CDC Urges HIV Tests as Routine in Health Care)
That recommendation changed the method of consent, so that patients would have to refuse the test, rather than accept it.
But the recommendation -- which generated some controversy (See Physicians Reluctant to Follow CDC Call for Routine HIV Testing) -- had not been formally tested in a large emergency department, Haukoos and colleagues said.
So they designed a "quasi-experimental" study that compared sequential four-month periods in which opt-out testing was alternated with "diagnostic testing" suggested by the treating clinician. During the opt-out periods, the treating physician could also order testing with a patient's specific consent.
Eligible patients coming to the emergency department were told during registration that rapid HIV testing would be conducted unless they refused it, which required them to check a box on the consent form and provide a signature or initial to show they were opting out.
Over the study period -- April 15, 2007 through April 15, 2009 -- 28,043 patients met eligibility criteria during the opt-out phases and 29,925 were eligible in the diagnostic phases, the researchers reported.
Of those, they found:
* During the opt-out phases, only 6,702 patients -- or 24% -- did not opt out of testing. Of those, 10 (0.15%) had new HIV diagnoses.
* 231 patients were diagnostically tested during the opt-out phases; five of them (or 2.2%) had new HIV diagnoses.
* During the diagnostic phases, 243 of the 29,925 eligible patients (0.8%) were tested and four (1.6%) had new diagnoses.
* The prevalence of new HIV diagnoses was 15 in 28,043 (0.05%) in the opt-out phases (including those diagnostically tested) and four in 29,925 (0.01%) in the diagnostic phases.
* Nontargeted opt-out HIV screening was associated with new HIV diagnoses, with a risk ratio of 3.6 (95% CI 1.2 to 10.8) after adjustment for patient demographics, insurance status, and whether diagnostic testing was performed in the opt-out phase.
* Median count of CD4-positive T cells for those with new HIV diagnoses in the opt-out phases (including those diagnostically tested) was 69 per microliter of blood. That was significantly higher (at P=0.02) than the 13 cells per microliter observed in new cases during the diagnostic phases, but still indicated advanced disease.
The researchers cautioned that the study was limited to a single institution and noted that the design of the study made randomization impossible, although the primary outcome -- HIV infection -- was not subject to ascertainment bias.
The low uptake of screening in the study "seems to challenge the hope for the opt-out approach," according to Roland Merchant, MD, ScD, and Michael Waxman, MD, both of Brown University in Providence, R.I.
Writing in an accompanying editorial, they argued that the key issue may be the way testing is presented to the patient. In the study, a registration clerk initiated HIV screening, they noted, without discussion of reasons for or merits of screening.
"As a result, patients could easily dismiss the need for screening and not be asked to provide reasons for declining," Merchant and Waxman wrote.
The study "demonstrates mostly the limitations of opt-out testing rather than its benefits," according to Michael Saag, MD, of the University of Alabama at Birmingham, who was not part of the study.
Saag said in an e-mail to ABC News and MedPage Today that a key question is why most patients refused the test. "In my experience," he said, "most individuals approached agree to be tested, but that may be because of the time I spend explaining why it is important."
Saag said the risk of death for people diagnosed late in the course of the disease remains high despite modern therapies, adding universal opt-out testing is "absolutely needed."
"Everyone who is sexually active should be screened at least once for HIV, regardless of their perceived risk factors," he said. "This should be a primary care initiative, similar to mammography and colonoscopy."
Meanwhile, Lewis Goldfrank, MD, of NYU Langone Medical Center in New York City, said his institution is still planning an opt-out strategy. In the meantime, counselors ask emergency patients to have the test, he said in an e-mail to ABC News and MedPage Today, although there are not enough to speak with all 100,000 patients every year.
But the numbers for the last 26,000 patients who agreed, he said, are similar to those seen in Denver -- 31 HIV-positive and six new diagnoses.
"Even with opt-in, we do not have enough staff to do all people in (the emergency department)," he said. "With opt-out we would have a similar problem." Goldfrank was not part of the Denver study.
The study had support from the CDC, the AHRQ, and the Colorado Center for AIDS Research.
Haukoos reported financial links with Abbott Laboratories and another author reported such links with ViiV Healthcare and Gilead Sciences. No other authors reported financial links with industry.
Merchant reported financial links with the Gilead Foundation.
Waxman reported that he had no financial links with industry.
This article was developed in collaboration with ABC News.
Primary source: Journal of the American Medical Association
Haukoos JS, et al "Routine opt-out rapid HIV screening and detection of HIV infection in emergency department patients" JAMA 2010; 304(3): 284-292.
Additional source: Journal of the American Medical Association
Merchant RC, Waxman MJ "HIV screening in health care settings: Some progress, even more questions" JAMA 2010; 304(3): 348-349.