Table of Contents
Bernard M. Branson, M.D. is the associate director for laboratory diagnostics at the Centers for Disease Control and Prevention, Division of HIV/AIDS Prevention in Atlanta, Georgia.
Dr. Branson is the lead author of the CDC's recommendations for opt-out HIV screening for adults, adolescents, and pregnant women. He heads CDC activities on new technologies for HIV testing, including rapid HIV tests.
Mascolini: What is the CDC's primary rationale for recommending opt-out HIV testing (Table 1) for all adults and adolescents seeking medical care?
Branson: There are several key reasons for recommending opt-out testing.1 One is to make testing more routine, because consent for most laboratory tests is based on an opt-out approach. The second reason is to remove potential barriers to testing that might be created by more elaborate requirements for HIV testing. And the third is to reduce stigma for patients who are undergoing HIV testing. It has been shown that when people perceive that something is offered to everyone as opposed to having some individuals singled out for targeted testing, they feel there's less stigma associated with the testing.
Mascolini: Can the US healthcare system absorb the direct cost of universal opt-out testing, as well as the cost of treating the many additional people who will presumably test positive?
Branson: I can't comment directly on all the costs, but I think it's important to point out that identifying HIV-positive people through testing does not create the expense. In other words, regardless of whether an infected person is tested, their need for treatment already exists, and they will still receive treatment when they are eventually diagnosed.
Usually, if a person presents later in the course of the disease, they're sicker and it's more expensive to take care of them than if they had been tested and diagnosed earlier, when they could have received optimal benefit from treatment. Some studies have shown that overall costs are lower when people are diagnosed earlier because they avoid hospitalizations and costs for more serious clinical events.2-4
Mascolini: AIDS Drug Assistance Programs (ADAPs) that help poor people pay for antiretrovirals are cutting back in several states. What will happen when more young, poor people who will need ADAP get diagnosed with HIV?
Branson: Again, I can't comment specifically on how wider HIV testing might influence ADAP support. We're in a state of change right now with healthcare reorganization in this country, and potential changes could affect eligibility for other programs like Medicaid. So I think it's difficult to predict exactly how care will be paid for if more people are diagnosed with HIV.
We do anticipate that identifying HIV-positive people earlier and giving them effective treatment that is likely to make them less infectious will in the long run decrease the number of people who will need treatment. That is the philosophy of the test-and-treat strategy that has been getting so much attention.5,6
One of the goals of the National AIDS Strategy7 is to increase the number of people who are diagnosed and to increase the number of people who are in treatment. I think opt-out testing recommendations are designed to address both of those goals.
Mascolini: Some cost-benefit analyses figure that universal opt-out testing is a good value in various scenarios.8-10 But one cost-benefit study determined that risk-based testing would diagnose more HIV cases than opt-out testing and prevent more HIV infections at a lower gross cost per infection averted.11 Can you summarize the arguments you made against this analysis in PLoS Medicine?12
Branson: First of all, I don't think HIV testing is an either/or question. In other words, there's not just one approach that should be applied. Nor is that the CDC's position. We believe expanded opt-out screening in healthcare settings will help detect HIV in people who would otherwise go undiagnosed. We're not proposing abandoning targeted testing for people at high risk and in particular for some people at very high risk who need to be retested periodically. Both strategies are necessary.
The model David Holtgrave developed11 incorporates some assumptions that can be questioned -- for example, that you can identify for testing, at no additional cost, a high-risk population with an HIV prevalence of 10%. Our studies show that seeking out a high-prevalence population for targeted testing is costly in terms of cost per test: You have to spend money to find those people, whereas there is a much smaller incremental cost for general HIV screening in healthcare settings because people are already coming in and seeking care in those settings.
Another important question involved in Holtgrave's analysis is the assumed prevention benefit of the counseling associated with HIV testing. The US Preventive Services Task Force reviewed behavioral interventions for sexually transmitted diseases including HIV and found no evidence that brief interventions similar to the kind that would be offered with an HIV rapid test have long-term benefits in changing behavior of people who test negative. More intensive interventions are necessary to have that kind of behavioral benefit.
I think those two factors -- the cost associated with targeted HIV testing and the question of exactly how much benefit there is from the brief counseling offered with an HIV test -- can explain the differences between Holtgrave's conclusions and the conclusions of several other models that focused more on detecting people with HIV infection and the benefits of getting them into treatment.
Mascolini: Are there legal risks in making opt-out testing standard of care in a medical practice or hospital -- especially in regard to informed consent, counseling, and linkage to care?
Branson: I think the legal risks go in two directions. Certainly there could be legal risks when a person is infected with HIV but the infection goes undetected because that person was not offered testing.
When you consider issues related to informed consent, one reason opt-out testing is recommended for healthcare settings and not for nonclinical settings is that healthcare settings already operate under a doctrine of informed consent. You can't do any procedure for a person in a healthcare setting without their consent. So the question with opt-out HIV testing is whether you need a separate, more elaborate written consent documenting that a person consents to HIV testing. Evidence from many places shows that separate written consent does not confer additional legal protection.
The opt-out recommendations have now been in place for 5 years, and we do not have evidence or reports of people suffering adverse consequences or being tested without their permission. We are confident that opt-out screening has not resulted in that kind of problem. In areas or institutions that have a long history of a separate process for HIV testing, the attorneys involved may wish to perpetuate that separate process, but we are not seeing that this provides a legal benefit to the institution.
Mascolini: Are US healthcare professionals adopting this strategy?
Branson: What I can say certainly is that professional organizations like the American College of Physicians and the American College of Obstetricians and Gynecologists have issued very similar recommendations to their members for HIV screening.
Percentage of persons aged 18-64 years who reported ever being tested for HIV (United States, 2001-2009), and percentage of late HIV diagnoses (AIDS diagnosis within 12 months of initial HIV diagnosis) (33 states, 2001-2007). (Reproduced from: Centers for Disease Control and Prevention. Vital signs: HIV testing and diagnosis among adults -- United States, 2001-2009. MMWR Morb Mortal Wkly Rep. 2010;59(47):1550-1555.13)
A CDC analysis showed that the proportion of Americans who said they'd been tested for HIV remained stable at 40% from 2001 to 2006, but then rose to 45% in the 3-year period from 2006 to 2009 (Figure).13 In 2009 a record 82.9 million adults in the United States reported having ever been tested for HIV infection. That represents an increase of 11.4 million people since the recommendations were issued in 2006. And we have numerous reports from groups that have initiated opt-out HIV screening programs.
All of this suggests that at least some healthcare professionals are adopting these recommendations. We always expected adoption and implementation to be incremental. We didn't think that everyone would just begin screening universally the year after the recommendations came out. But I do think we have evidence that they are being adopted.
Mascolini: If an institution decides to implement opt-out testing, what steps should it take to put this policy in place?
Branson: The most important first step is to get buy-in from all the stakeholders involved, both people in the institution and in the community. Also, for each institution, questions related to nursing, standing orders, and the laboratory, for example, must be addressed and solutions designed that are specific to the needs of that institution.
The CDC recommends that people receive information about HIV at the time they are tested, so there has to be a mechanism to deliver that information, whether with a pamphlet, a handout, or a video (Table 2). Similarly, people need to be advised that they have an opportunity to decline testing. Many institutions have done this with a separate information sheet that might say an HIV test is recommended as part of your care, here are the reasons for it, and if you don't want this test you should sign this and give it to one of the health care providers. I think those are necessary steps in the process of implementing opt-out testing.
When CDC's recommendations came out, there was some confusion about whether CDC was recommending opt-out HIV testing or recommending rapid testing. Many places assumed that they have to do rapid point-of-care testing to implement these recommendations. Rapid testing is not part of the recommendation. Rapid testing may be important in settings where patients are unlikely to receive their results otherwise. But with the newer technologies available today, a conventional test can produce results in an hour or less, so it is not always necessary to use a point-of-care rapid test.
Mascolini: Are there other points you would like to add on any of these issues?
Branson: I think an important point is that programs that are implementing these recommendations are in fact identifying people who have undiagnosed HIV infection. The recommendations are having the desired outcome -- getting more people tested and getting more people linked into services where they can access treatment. And that treatment will have a substantial impact on improving their life expectancy and potentially in decreasing the likelihood that they will transmit HIV to others. I think the bottom line is that evidence accumulating since the recommendations came out indicates that they seem to be working: The number of persons who have been tested is going up, and the number whose HIV is diagnosed late is going down.