Monday, November 19, 2012

The case against universal HIV screening - Part 1 of 2

Today the U.S. Preventive Services Task Force released a draft statement that updates their previous (2005) recommendations on screening for HIV. Based on findings from a systematic review published in the Annals of Internal Medicine, the USPSTF now recommends routine HIV screening for all persons ages 15 to 65 years, regardless of risk status. Previously, the Task Force had only recommended screening in adolescents and adults at increased risk for HIV infection; this new statement, if finalized, makes the USPSTF's recommendations concordant with guidelines from the Centers for Disease Control and Prevention that have been in place since 2006.

I have had the privilege of caring for patients with HIV/AIDS since working in a Comprehensive Care Center during my family medicine residency. For the past eight years I have practiced  in and around Washington, DC, which has the highest prevalence of HIV infection (3 percent) of any city in the United States. It is clear to me that all adults and sexually active adolescents in the DC area are at increased risk for HIV and should be offered routine HIV screening. Further, I believe that screening for HIV should be provided to any patient who requests it. Nonetheless, I disagree with the USPSTF's new recommendation for universal HIV screening in adolescents and adults. My next two posts, drawn from an unpublished paper that I presented in part at a Family Medicine Grand Rounds in November, 2011, explain why.


Although the introduction of highly active antiretroviral therapy has led to substantial declines in AIDS diagnoses and deaths from their peaks in the mid-1990s, an estimated 50,000 persons in the U.S. are infected with HIV each year, and more than 236,000 persons living with HIV are unaware of their diagnoses (1). Noting that many HIV infections occur in persons without identifiable risk factors, in 2006 the Centers for Disease Control and Prevention recommended routine opt-out testing of all persons between the ages of 13 and 64 in order to provide treatment at earlier stages of disease and curb behaviors associated with a high risk of HIV transmission (2). These recommendations, which have been endorsed by the American College of Physicians (3), the American Academy of Pediatrics (4), and the Office of National AIDS Policy (5), stand in contrast to those from the U.S. Preventive Services Task Force (USPTSF) (6) and the American Academy of Family Physicians (7), which support individualizing screening decisions in adolescents and adults not at increased risk of HIV infection.

Absent the USPSTF’s endorsement, routine HIV screening is not included in the list of preventive services that the 2010 Affordable Care Act requires health plans to cover without cost-sharing (8). Disagreements about the added value of a routine opt-out rather than risk factor-based screening approach may arise from differing interpretations of the former’s cost-effectiveness and potential benefits and harms (9). In this paper, I argue that while routine opt-out screening for HIV infection may be effective in many U.S. clinical settings, there is insufficient data to support universal implementation.

Cost Effectiveness

Any screening approach that tests more people is likely to diagnose more infections than a targeted approach. However, to be preferred to less intensive approaches, universal screening should also be cost-effective (diagnosing and treating additional infections at reasonable societal cost) and lead to improved health outcomes for infected patients and uninfected partners. To support its 2006 recommendations, the CDC relied on a mathematical model (10) that estimated that one-time universal HIV screening met a traditional cost-effectiveness threshold ($50,000 per quality-adjusted life-year) in settings with a HIV prevalence as low as 0.2 percent. The CDC suggests that routine opt-out screening can be discontinued when HIV prevalence is documented to be lower than 0.1 percent (2). In settings of higher HIV prevalence, universal screening every 3 or 5 years appears to be cost-effective (10). A key variable in this model is the impact of increasing diagnoses on changing risky behaviors and reducing HIV transmission rates, which has not been studied directly (11). A subsequent analysis projected that while expanded HIV screening alone will prevent few new infections, earlier initiation of antiretroviral therapy and modifications in risky sexual behavior together could reduce new infections by up to 65 percent (12).

In contrast, others have argued that the programmatic expense of fully implementing the 2006 CDC recommendations offers poor value relative to spending additional funds on pre- and post-test counseling within the existing risk factor-based screening infrastructure. According to a 2007 analysis, for the $864 million cost of offering routine opt-out testing to every American between the ages of 13 and 64, targeted counseling could potentially diagnose and prevent as 3 times as many HIV infections (13). This analysis was limited by its assumption that a substantial minority of patients offered HIV testing would decline it; the CDC, in contrast, observed that pregnant women rarely opt out of routine HIV testing, and asserted that uptake of testing in non-pregnant persons could be similarly high (2).

The vast majority of persons with HIV infection who are not diagnosed via screening will ultimately be diagnosed when they present to care with symptomatic AIDS. Therefore, a major benefit of routine opt-out testing is tied to identifying such persons earlier in the course of their infections, when antiretroviral medications and prophylaxis against opportunistic infections will prevent morbidity and mortality. The CDC guideline (2) and mathematical models that have postulated large benefits of expanded screening (10, 12) assume that a large number of persons will be diagnosed at early stages of disease.

Studies of Opt-Out Screening Effectiveness

Unfortunately, studies of routine opt-out HIV screening in emergency department settings have challenged this assumption. In a quasi-experimental time-samples study at a large emergency department in Denver, Colorado, researchers compared the effectiveness of opt-out rapid HIV testing with physician-directed diagnostic rapid HIV testing (14). More than 75 percent of eligible patients were either not offered or declined screening during the opt-out phase of the study; of the 6702 patients who received HIV testing, 10 (0.15%) tested positive. In contrast, 5 of 231(2.2%) patients who received HIV testing during the physician-directed phase tested positive for HIV. Moreover, 6 of 10 patients who tested positive during the opt-out phase had established risk factors for HIV infection (and would have been eligible for targeted screening) and 6 of 10 patients had CD4 counts lower than 350/µL. Similarly, in a study of opt-out rapid HIV testing in 29 emergency departments in metropolitan Paris, France, only 27% of 78,000 eligible patients were offered testing and only 16% (12,754) were actually tested (15). Of the 18 patients (0.14%) who tested positive, all but 1 belonged to a high-risk group, and 8 already had symptoms suggestive of opportunistic infections or AIDS.

One could argue that despite the large sample sizes of these two studies, the effectiveness of universal HIV screening can only be evaluated on a national level. The CDC recently reported the results of its Expanded HIV Testing Initiative, a program that provided more than $100 million from 2007-2010 to health departments in 24 U.S. states and the District of Columbia to increase screening rates and linkages to care among populations disproportionately affected by HIV, such as African Americans (16). To qualify for funding, health departments had to have reported at least 140 AIDS diagnoses among African Americans in 2005. Of more than 2.7 million HIV tests performed during the initiative, nearly 30,000 (1.1%) were positive, and the majority of patients with positive tests had not previously been aware of their HIV diagnosis (16). As encouraging as these results are, they are more supportive of the effectiveness of targeted testing (mostly performed in communities with well above-average HIV prevalence) than universal screening.

In 2009, the Veterans Administration implemented a new HIV testing policy in its health system that is consistent with the CDC’s 2006 recommendations for routine opt-out testing. Between 2009 and 2010, among VA patients with at least one outpatient visit, an electronic health record analysis indicated significant increases in the percentages of patients who had ever been tested for HIV (9.2 to 13.5 percent) and those who had been tested during that calendar year (2.9 to 5.8 percent) (17). Positive results were reported in 0.7 to 1.2 percent of tested patients; data was not available on how many of these patients were previously undiagnosed, would have been eligible for testing in a risk factor-based approach, or were diagnosed at preclinical stages of HIV infection (17).

- continued in my next post -

References 1-17

1. Centers for Disease Control and Prevention. HIV surveillance - United States, 1981-2008. MMWR Morb Mortal Wkly Rep 2011;60:689-93.
2. Centers for Disease Control and Prevention. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep 2006;55(RR-14):1-17.
3. Qaseem A, Snow V, Shekelle P, et al. Screening for HIV in health care settings: a guidance statement from the American College of Physicians and HIV Medicine Association. Ann Intern Med 2009;150:125-31.
4. American Academy of Pediatrics. Adolescents and HIV infection: the pediatrician’s role in promoting routine testing. Pediatrics 2011;128:1023-29.
5. Office of National AIDS Policy. National HIV/AIDS Strategy. Washington, DC: Office of National AIDS Policy; 2010. Available at:
6. U.S. Preventive Services Task Force. Screening for HIV: recommendation statement. Ann Intern Med 2005;143:32-7.
7. Rodnick JE. The CDC and USPSTF recommendations for HIV testing. Am Fam Physician 2007;76:1456-59.
8. Bayer R, Oppenheimer GM. Routine HIV screening – what counts in evidence-based policy? N Engl J Med 2011;365:1265-8.
9. Chou R. Routine screening for chronic human immunodeficiency virus infection: why don’t the guidelines agree? Epidemiol Rev 2011;33:7–19.
10. Paltiel AD, Walensky RP, Schackman BR, et al. Expanded HIV screening in the United States: effect on clinical outcomes, HIV transmission, and costs. Ann Intern Med 2006;145:797-806.
11. Chou R, Huffman L. Screening for human immunodeficiency virus: focused update of a 2005 systematic evidence review for the U.S. Preventive Services Task Force. AHRQ Pub No. 07-0597-EF-1. Agency for Healthcare Research and Quality, April 2007. Available at:
12. Long EF, Brandeau ML, Owens DK. The cost-effectiveness and population outcomes of expanded HIV screening and antiretroviral treatment in the United States. Ann Intern Med 2010;153;778-89.
13. Holtgrave DR. Costs and consequences of the US Centers for Disease Control and Prevention’s recommendations for opt-out HIV testing. PLoS Med 2007;4:e194.
14. Haukoos JS, Hopkins E, Conroy AA. Routine opt-out rapid HIV screening and detection of HIV infection in Emergency Department patients. JAMA 2010;304:284-92.
15. Wilson d’Almeida K, Kierzek G, Truchis P, et al. Modest public health impact of nontargeted human immunodeficiency virus screening in 29 emergency departments. Arch Intern Med 2011 Oct 24 [Epub ahead of print] doi:10.1001/archinternmed.2011.535
16. Centers for Disease Control and Prevention. Results of the expanded HIV testing initiative – 25 jurisdictions, United States, 2007-2010. MMWR Morb Mortal Wkly Rep 2011;60:805-10.
17. Halloran J, Czarnogorski M, Dursa EK, et al. HIV testing in the Department of Veterans Affairs, 2009-2010. Arch Intern Med 2011 Oct 24 [Epub ahead of print] doi:10.1001/archinternmed.2011.510