HIV Screening via Fourth-Generation Immunoassay or Nucleic Acid Amplification Test in the United States: A Cost-Effectiveness Analysis

Citation: Long EF (2011) HIV Screening via Fourth-Generation Immunoassay or Nucleic Acid Amplification Test in the United States: A Cost-Effectiveness Analysis. PLoS ONE 6(11): e27625. doi:10.1371/journal.pone.0027625 HIV Screening via Fourth-Generation Immunoassay or Nucleic Acid Amplification Test in the United States: A Cost-Effectiveness Analysis Elisa F. Long 0 Michael George Roberts, Massey University, New Zealand 0 School of Management, Yale University , New Haven, Connecticut , United States of America Background: At least 10% of the 56,000 annual new HIV infections in the United States are caused by individuals with acute HIV infection (AHI). It unknown whether the health benefits and costs of routine nucleic acid amplification testing (NAAT) are justified, given the availability of newer fourth-generation immunoassay tests. Methods: Using a dynamic HIV transmission model instantiated with U.S. epidemiologic, demographic, and behavioral data, I estimated the number of acute infections identified, HIV infections prevented, quality-adjusted life years (QALYs) gained, and the cost-effectiveness of alternative screening strategies. I varied the target population (everyone aged 15-64, injection drug users [IDUs] and men who have sex with men [MSM], or MSM only), screening frequency (annually, or every six months), and test(s) utilized (fourth-generation immunoassay only, or immunoassay followed by pooled NAAT). Results: Annual immunoassay testing of MSM reduces incidence by 9.5% and costs ,$10,000 per QALY gained. Adding pooled NAAT identifies 410 AHI per year, prevents 9.6% of new cases, costs $92,000 per QALY gained, and remains ,$100,000 per QALY gained in settings where undiagnosed HIV prevalence exceeds 4%. Screening IDUs and MSM annually with fourth-generation immunoassay reduces incidence by 13% with cost-effectiveness ,$10,000 per QALY gained. Increasing the screening frequency to every six months reduces incidence by 11% (MSM only) or 16% (MSM and IDUs) and costs ,$20,000 per QALY gained. Conclusions: Pooled NAAT testing every 12 months of MSM and IDUs in the United States prevents a modest number of infections, but may be cost-effective given sufficiently high HIV prevalence levels. However, testing via fourth-generation immunoassay every six months prevents a greater number of infections, is more economically efficient, and may obviate the benefits of acute HIV screening via NAAT. - Each year, more than 56,000 people in the United States acquire HIV, many of whom are infected by individuals with acute HIV infection (AHI), although the exact contribution of AHI is uncertain.[14] AHI typically lasts for two to three months after initial infection and individuals with AHI are exceptionally infectious during this period due to rapid viral replication,[2,5,6] because blood plasma viral loads are 100 times higher than during asymptomatic infection.[7] Moreover, individuals with AHI are likely status-unaware and may have had recent sexual contact with one or more partners. Prior studies indicate that individuals identified with AHI may reduce risky sexual behavior.[8,9] Successfully identifying such individuals during a short window may necessitate a frequent AHI screening program. Third-generation enzyme linked immunosorbent assays (ELISA) do not detect antibodies for at least three weeks after infection, and newer fourth-generation antigenantibody combination tests reduce this window by several days. Before third-or fourth-generation assays detect infection, plasma viral RNA may be detected with a nucleic acid amplification test (NAAT). Individual NAAT screening is cost-prohibitive in many settings, and several studies have developed and piloted pooled NAAT testing, with the optimal pooling algorithm depending on undetected AHI prevalence.[6,1016] Pooled NAAT has been shown to be cost-effective in a community clinic serving high-risk men who have sex with men (MSM), although the study did not compare testing with a fourth-generation immunoassay.[17] Another study found that fourth-generation tests detect 62% of samples classified as acute infection, suggesting that newer immunoassays may obviate the need for NAAT testing.[18] Recent guidelines recommend routine HIV screening of adults and adolescents aged 13 to 64,[19] but it is unknown to what extent concomitant efforts to increase AHI testing via NAAT will prevent new infections and whether such a strategy is cost-effective. Additionally, it is unclear whether NAAT testing should be utilized given that a fourth-generation immunoassay was approved by the U.S. Food and Drug Administration in June 2010. Identifying the optimal HIV screening strategy, including which test(s) to administer, screening frequency, and target population, could potentially prevent thousands of new HIV infections, adding millions of life years to the population. The present study is the first to compare the population-level health benefits and costs of universal or targeted HIV screening with a fourth-generation immunoassay, versus screening for acute infection with pooled NAAT. Study Design The authors previously published model [20,21] of HIV transmission and disease progression was modified to include acute HIV screening via NAAT. I instantiated the model using demographic, epidemiologic, and cost data for the United States. I then numerically simulated the epidemic over a 20-year time horizon and estimated population-level outcomes, including HIV incidence, AHI identified, quality-adjusted life years (QALYs), costs, and cost-effectiveness. Additional model details are provided as Supporting Information (Text S1). Population To account for variations in behavior and infection risk, the adult population aged 15 to 64 years was subdivided based on gender, risk behavior (MSM, injection drug users (IDU), MSM/ IDU, or low-risk), and male circumcision status (Table S1). By integrating data on population sizes, number of people living with HIV, and the distribution of infections by transmission mode, undiagnosed HIV prevalence in each risk group was estimated: 4.3% (MSM), 4.4% (male IDUs), 6.4% (MSM/IDUs), 5.9% (female IDUs), 0.03% (low-risk men), and 0.07% (low-risk women).[1,22 28] The HIV-infected population was further divided based on Proportion tested in past 12 months (status quo) Symptom-based case finding per year Window period of detection (days) 3rd generation ELISA 4th generation immunoassay NAAT pooling algorithm sensitivity NAAT pooling efficiency (tests/specimen) Proportion tested who receive NAAT test results Reduction in sexual partners if identified Cost of 3rd-generation ELISA Cost of Western Blot confirmatory test Cost of quantitative viral load assay Cost of HIV counseling disease stage, identification status, and antiretroviral treatment status. The model included population entry and exit, non-HIVrelated mortality, and IDU-related mortality (Table S2). HIV Transmission and Progression An important p (...truncated)