To the Editors:
Since the Centers for Disease Control and Prevention1 and the American College of Physicians2 have recommended broader HIV screening in all healthcare settings, rapid HIV tests, particularly those accepting oral specimens, have been instrumental in increasing test acceptance rates. Overall, oral rapid tests have made HIV testing possible in many venues where obtaining blood from phlebotomy or fingerstick may be impractical.3 In one Boston emergency room testing program, Walensky et al found that only five of 31 (16%) initial reactive oral test results were confirmed to have HIV infection and concluded that their specificity of 96.9% and positive predictive value of 16.1% using the OraQuick Advance Rapid HIV-1/2 antibody test (OraSure Technologies, Inc, Bethlehem, PA) on 849 oral fluid tests was lower than anticipated.4 Given our known HIV rate within the VA Medical Center in Washington, DC (VAMC-DC), located in the metropolitan area with the highest HIV prevalence within the United States,5 we evaluated the test performance characteristics of our rapid HIV Point-of-Care (HIV POC) test.
This project was approved by the Human Studies Subcommittee and the Research & Development Committee at the VAMC-DC and granted waiver of Health Insurance Portability and Accountability Act authorization. At the VAMC-DC, OraQuick Advance Rapid HIV-1/2 antibody test kits were used in our HIV POC program on oral mucosal transudate and whole blood samples. All testers were trained and annually certified in the performance of this rapid HIV test according to the manufacturer's instructions. All reactive screening results were sent for confirmation by Western blot. Negative screening tests were not sent for confirmation as a result of volume of tests, cost, and the negative effect that requiring a needlestick would have on patient acceptance of the test. Therefore, we used the sensitivity provided by the manufacturer's package insert6 for our calculations.
Among the 3717 tests performed, 3716 were deemed valid and only a single invalid test resulted from an oral swab. This single test revealed the internal control line just outside of the result window, which, according to the package insert, is considered an invalid test and should be repeated. The manufacturer of the test determined that the invalid was caused by a slight shift in the assay strip in that particular device (OraSure Technologies, Inc, personal communication).
Of 3716 valid tests, 3007 were from oral mucosal transudate specimens, and the remaining 709 were from whole blood, as shown on Table 1 with our test performance characteristics. Of 42 initial reactive screening results, 31 were confirmed as true-positive and 11 were false-positive as determined by a negative Western blot assay and undetectable HIV RNA. Our test specificity was lower for oral (99.6%) than for whole blood (100%) specimens but remained within the manufacturer's specificity range (95% confidence interval, 99.6-99.9% for oral mucosal transudate and 99.7-100% for whole blood from the kit package insert). With our known HIV rate of 2.3%, the positive predictive value and likelihood ratio for HIV infection in a patient with a reactive rapid HIV POC test were high at VAMC-DC.
Our HIV POC testing program demonstrated an overall specificity of 99.7% in accordance with the manufacturer's stated range with the specificity being higher for whole blood compared with oral mucosal transudate specimens. Using our known HIV rate of 2.3%, our calculated positive predictive value at the VAMC-DC was fivefold higher than that reported by Walensky et al. Our positive likelihood ratio was considerably greater as our known HIV rate of 2.3% is nearly 2.6- to 3.8-fold higher than the HIV prevalence reported in Boston of 0.9%7 and the estimated 0.6% prevalence in a Boston testing program,4 respectively. The differences in positive predictive value and positive likelihood ratio between our study and that reported by Walensky4 are largely the result of the HIV prevalence rates at our respective sites. Our results are further supported by the high specificity of 99.9% and positive predictive value of 84.6% reported from the cumulative oral swab HIV screening program in an active academic emergency department in the District of Columbia.8
During this evaluation period, there was a single invalid test, representing a rate of 0.03%. A comparative performance evaluation of OraQuick rapid HIV antibody tests by nonlaboratorians revealed an invalid rate of five of 297 (1.6%) overall. Four of these errors were attributed to testing errors by untrained testers, and an invalid rate of one of 144 (0.7%) among testers who received a visual demonstration was related to “true device failure (control line outside prescribed area).”9 There are no other published reports of rates of invalid tests in the medical literature related to this specific product or other commercially available rapid HIV screening kits, suggesting extremely low or no reporting of invalid tests by testing sites to the manufacturers of these rapid test kits.
In summary, sites performing rapid HIV testing should evaluate their own test performance characteristics with their own known rates of HIV infection as a result of concerns for reduced specificity and false-positive screening results from oral mucosal transudates. In high HIV prevalence areas such as VAMC-DC, rapid HIV tests using whole blood and oral mucosal transudate specimens have had high specificity and positive predictive value.
We thank Mariza Alignay, BSN, Linda Allen, MSW, Debra Benator, MD, Rene Brooks, BSN, Hannah Cohen-Blair, BA, Pamela Crutchfield, BSN, Shirley Cummins, Maggie Czarnogorski, MD, Sharon Doughty, BSN, Judith DuBois, BSN, Irvin Hiralall, BA, Leigh Kennedy, DO, Margaret Lankford, CNP, Karen Lee-Barton, BSN, Christopher May, Angela McKnight, APRN-BC, Shrimant Mishra, MD, Sara Most, BSN, Laura Nesmith, BSN, Carrie Pineda, MSN, Marc Siegel, MD, Lynette Taitt, BSN, Melissa Turner, MSW, and Tzipora Wagner for their efforts in our HIV point-of-care rapid screening program.
Karen Rexroth, BS
Katherine Hare, BS
Virginia L. Kan, MD
Infectious Diseases Laboratory, VA Medical Center, Washington, DC
1. Branson BM, Handsfield HH, Lampe MA, et al; Centers for Disease Control and Prevention (CDC). Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep
2. Qaseem A, Snow V, Shekelle P, et al; Clinical Efficacy Assessment Subcommittee, American College of Physicians. Screening for HIV in health care settings: a guidance statement from the American College of Physicians and HIV Medicine Association. Ann Intern Med
3. Centers for Disease Control and Prevention. False-positive oral fluid rapid HIV tests-New York City, 2005-2008. MMWR Morb Mortal Wkly Rep
4. Walensky RP, Arbelaez C, Reichmann WM, et al. Revising expectations from rapid HIV tests in the emergency department. Ann Intern Med
6. OraQuick Advance pakage insert. Bethlehem, PA: OraSure Technologies, Inc; October 2007.
8. Brown J, Magnus M, Czarnogorski M, et al. Another look at emergency department HIV screening in practice: no need to revise expectations. AIDS Res Ther
9. Granade TC, Parekh BS, Phillips SK, et al. Performance of the OraQuick and Hema-Strip rapid HIV antibody detection assays by non-laboratorians. J Clin Virol