In 2005, HIV screening and diagnostic testing was implemented at the Highland Hospital emergency department (ED) using the OraQuick Advance Rapid HIV 1/2 Antibody Test (OraSure Technologies, Bethlehem, PA) on oral fluid (OF) specimens.1 Although the accuracy of the OraQuick test with whole blood (obtained by fingerstick or venipuncture) is slightly better than with OF (sensitivity 99.6% vs. 99.3% and specificity 100% vs. 99.8%),2 we chose to test OF because of the ease of performance, speed of specimen collection, lack of the need for phlebotomy certification (required in California for persons who perform fingerstick or venipuncture blood collection), and reports of patient preference.3-7
Since starting our program, several HIV testing sites have reported unexplained clusters of an increased frequency of false-positive test results from the OraQuick test when used on OF specimens, but not with whole blood.8-10 The New York City Department of Health has suspended OF rapid HIV testing in sexually transmitted disease clinics due to an increase in the incidence of false-positive OF tests.10 In addition, the specificity of testing using the OraQuick test with OF specimens in an ED was recently found to be 96.9%, significantly lower than that reported by the manufacturer.11
In early 2007, we experienced a cluster of false-positive test results. Of the 1388 HIV tests performed between February and March 2007, 10 were false-positive. The specificity of the test during this period was 99.3%. These false-positive results created anxiety for patients and led to distrust of the rapid HIV test. Providers also became concerned about the accuracy of rapid HIV testing, which threatened their support for continued testing. Furthermore, significant resources were required to counsel patients, obtain specimens for confirmatory Western blot testing, and arrange follow-up appointments for patients whose results were false-positive.
In this study, we sought to determine whether patient acceptance of rapid HIV screening with fingerstick whole blood (FWB) differed from that with OF. If rates of rapid HIV screening using OF and FWB are similar, then screening using blood specimens may be preferable because of its greater sensitivity and specificity compared with OF testing and the fact that FWB testing has not been linked with clusters of excess false-positive test results.
This prospective study utilized an alternate day approach in which OF or FWB testing was performed on alternate weekdays. This approach was chosen instead of providing subjects with the option of choosing 1 testing method vs. the other in an attempt to mimic what one might expect to see in clinical practice (when either one or the other test modality is available but not both) and to minimize testing staff bias. The study was conducted at the Alameda County Medical Center Highland Hospital ED, an urban teaching hospital and regional trauma center in Oakland, CA, that serves a largely minority population with low socioeconomic status. The Institutional Review Board of the Alameda County Medical Center approved the study protocol with a waiver of written informed consent.
Triage nurses referred medically stable ambulatory patients older than 15 years and not known to be HIV infected to an HIV tester who was stationed in the triage area. Two full-time HIV testers, who were phlebotomy-certified medical assistants, performed HIV screening weekdays between 7:30 am and 8:30 pm. The HIV testers were trained by one of the study investigators (D.A.E.W.) to routinely offer HIV screening according to a scripted interaction. Before implementing the study, HIV testers demonstrated competency with the scripted interaction through a series of mock patient interactions. Testers would ask patients, “Would you like to have an HIV test today?” and explain that testing was free of charge and that results would be ready during their ED stay. To minimize bias, testers were instructed not to describe the specimen collection method. HIV testers determined the primary reason patients declined testing at 2 distinct stages of the screening process. The first time was at the level of offering HIV screening and the second time was at the level of performing the test. Patients who accepted screening initially but declined at the second stage were asked why they changed their mind. All patients who agreed to HIV screening signed separate written informed consent in accordance with state law.
HIV testers recorded the following information in specific HIV testing fields that were incorporated into the ED electronic medical record (Wellsoft Corporation, Somerset, NJ): whether patients were offered, accepted, and completed HIV screening, primary reasons for declining screening at the offering and testing stage, and test results. These data, as well as demographic data routinely collected during an ED visit (age, sex, race, and ethnicity) were exported to spreadsheets (Microsoft Excel 2003; Microsoft Corporation, Redmond, WA) from which identifying information was removed.
The primary outcome measure was the proportion of patients completing rapid HIV screening on days when OF or FWB HIV screening was available. Secondary outcome measures included the reasons why patients declined screening. Assuming a baseline screening acceptance rate of 60%, a priori sample size calculations estimated 538 patients per group were required to detect an absolute 10% difference in the proportion of patients completing HIV screening with an α of 0.05 with 90% power. Each patient visit was analyzed as a separate encounter. SPSS statistical software (version 11; SPSS, Inc., Chicago, IL) was used to calculate descriptive statistics (means, medians), bivariate statistical tests (χ2 and t test), and P values. Continuous data are reported as means with 95% confidence intervals (CIs) and categorical data are reported as percentages.
Figure 1 shows the results of screening. Between May 1 and June 30, 2007, OF and FWB HIV screening were each available for 22 days. A total of 2201 patients were offered HIV screening: 1089 during OF screening days and 1112 during FWB screening days. The demographic characteristics of patients offered OF and FWB were similar: 53% male, 44% Black, 37% Hispanic, and 15% White, with a mean age of 40 (SD ± 13) years (data not shown). Overall, 1330 of the 2201 patients (60.4%) completed HIV screening. Screening rates were similar on days that OF and FWB screening was available [674 of 1089 (61.9%) vs. 656 of 1112 (59.0%), P = 0.18].
At the offering stage, similar proportions of patients declined HIV screening on OF and FWB days [415 of 1089 (38.1%) vs. 430 of 1112 (38.7%), P = 0.87]. Reasons for declining screening at this stage were generally similar for the 2 groups (data not shown). The most common reasons were having recently been tested for HIV (49.9%) and lack of perceived HIV risk (31.4%). Additional reasons included other/unknown (11.4%), scared of result (3.7%), prefer to be tested elsewhere (1.4%), prefer to focus on main medical problem (1.4%), HIV positive (0.5%), and too sick (0.3%).
At the testing stage, 25 of the 682 patients (3.7%) that accepted screening on FWB days changed their minds and declined testing because they did not want a blood test (17) or they preferred to have an oral test (8). None of the 674 patients that accepted screening on OF days changed their minds or cited concerns about the collection method. The demographic characteristics (age, sex, and ethnicity) of the 25 patients who declined FWB for reasons related to the collection method were similar to those who declined for reasons unrelated to the FWB collection method (data not shown). Rapid tests were performed in all patients who accepted screening except one, who was too ill to complete FWB testing.
Six tests were preliminary positive and HIV infection was confirmed by Western blot in 4 of them. All confirmed positive HIV tests occurred with FWB testing. One OF and 1 FWB test were false-positive. The specificities of OF testing (99.85%; 95% CI: 99.56 to 100) and FWB testing (99.85%; 95% CI: 99.55 to 100) were similar.
The Centers for Disease Control and Prevention recommend HIV screening in health care settings when the HIV prevalence exceeds 0.1%.12 Rapid HIV testing is an important technological advancement that has enabled EDs to implement HIV screening.1,13-16 Unlike with conventional HIV testing, rapid HIV testing can be performed as point-of-care, allowing for results to be available during a patient's ED visit and to be used by physicians to guide clinical management.
In comparison to FWB testing, the more convenient OF HIV testing has 2 major disadvantages: an increased number of false-positive tests, attributed to the lower test specificity, and reports of clusters of excess numbers of false-positive tests.8-10 Several reports, on the other hand, have shown that patients prefer less invasive methods for HIV testing, which suggests that OF collection may improve uptake of testing and lead to more diagnoses of HIV.3-7,10,17 These studies, however, merely reflect the preferences of patients. And although patients may prefer OF testing to fingerstick or venipuncture testing when given a choice, these studies do not demonstrate whether this preference influences a patient's decision to accept or decline testing in clinical practice.
Our data suggests that, among ED patients, the preference for 1 rapid test collection modality over another has a minimal effect on actual screening rates. In fact, we demonstrate that in an ED where rapid HIV screening is routinely available, testing rates are similar regardless of whether OF or FWB tests are used. With only 25 of the patients declining FWB testing doing so for reasons related to the collection process, FWB testing does not seem to be a major barrier for HIV testing. Although the implications of this decreased acceptance with FWB testing may be clinically important, especially in high HIV prevalence settings, reasons for refusal other than collection modality are more frequent. For example, among both groups, the most common reasons for declining testing were patients reporting having been recently tested for HIV and patients perceiving themselves to be at low risk. Unfortunately, prior research has shown inconsistency in self-reports of HIV testing and the accuracy of patient reporting of recent HIV testing is unknown.18 Furthermore, a patient's self-perception of HIV risk may be incongruent with their true risk profile. Half of the patients diagnosed with HIV in an urban ED in Chicago reported no risk behaviors.15
Simple procedures may be implemented to influence the uptake of screening. For the majority of patients who decline screening, improved pretest education, including information about routes of HIV transmission, risk behaviors, and recommended repeat testing frequencies, might improve acceptance rates. For patients who decline FWB screening because of concerns about the collection modality, providing them with an explanation that blood testing is more accurate than OF testing may influence their decision. For the remaining patients who still decline FWB testing, offering OF testing as a backup might be implemented to improve overall acceptance.
The study has several limitations. Bias from testing staff may have influenced acceptance rates, either by increasing rates through the promotion of a preferred testing modality or by decreasing rates through the dissuasion of a disliked one. Although we attempted to minimize bias from testing staff by removing the description of the collection technique from the offering stage, this process was not monitored for compliance.
Our study design did not give patients the option of choosing 1 test over the other because in clinical practice screening programs will most likely use a single collection method. Acceptance rates outside of a research setting, however, may differ because patients will probably be informed of the test type up front. In fact, programs using FWB collection will most likely emphasize the accuracy of blood testing, whereas programs using OF specimens will probably call attention to its ease, simplicity, and noninvasive nature, which may influence acceptance.
The results of this study are only applicable to point-of-care testing using FWB and OF collection. Programs performing rapid HIV tests in the laboratory, however, would most likely obtain venipuncture specimens.19 A comparison of the acceptance rates for venipuncture and OF testing was not performed.
Finally, this study only addressed one (testing rates) of the many variables that must be taken into consideration when deciding which rapid HIV test to use. Currently, there are at least 4 rapid HIV tests suitable for point-of-care testing that have different operating characteristics.20 Elements including cost, product shelf life, local and state phlebotomy requirements, staff preference, test processing times, and impact on departmental flow may also influence the decision to use one rapid test over another.
Compared with OF screening, rapid HIV screening with FWB does not seem to be a significant barrier. Because screening rates for the 2 collection modalities are similar, FWB testing, when feasible, may be preferable because of its greater accuracy.
The authors wish to acknowledge Harrison J. Alter, MS, MD, who provided statistical support.
1. White DA, Scribner AN, Schulden JD, et al. Results of a Rapid HIV screening and diagnostic testing program in an urban emergency department. Ann Emerg Med. 2009;54:56-64.
3. Liddicoat RV, Losina E, Kang M, et al. Refusing HIV testing in an urgent care setting: results from the “Think HIV” program. AIDS Patient Care STDS. 2006;20:84-92.
4. Peralta L, Constantine N, Griffin Deeds B, et al. Evaluation of youth preferences for rapid and innovative human immunodeficiency virus antibody tests. Arch Pediatr Adolesc Med. 2001;155:838-843.
5. Spielberg F, Branson BM, Goldbaum GM, et al. Overcoming barriers to HIV testing: preferences for new strategies among clients of a needle exchange, a sexually transmitted disease clinic, and sex venues for men who have sex with men. J Acquir Immune Defic Syndr. 2003;32:318-327.
6. Hutchinson AB, Corbie-Smith G, Thomas SB, et al. Understanding the patient's perspective on rapid and routine HIV testing in an inner-city urgent care center. AIDS Educ Prev. 2004;16:101-114.
7. Spielberg F, Kurth A, Gorbach PM, et al. Moving from apprehension to action: HIV counseling and testing preferences in three at-risk populations. AIDS Educ Prev. 2001;13:524-540.
8. Delaney KP, Branson BM, Uniyal A, et al. Performance of an oral fluid rapid HIV-1/2 test: experience from four CDC studies. AIDS. 2006;20:1655-1660.
9. Jafa K, Patel P, Mackellar DA, et al. Investigation of false positive results with an oral fluid rapid HIV-1/2 antibody test. PLoS ONE. 2007;2:e185.
10. Cummiskey J, Mavinkurve M, Paneth-Pollak R, et al. False-positive oral fluid rapid HIV tests-New York City, 2005-2008. MMWR Morb Mortal Wkly Rep. 2008;57:660-665.
11. Walensky RP, Arbelaez C, Reichmann WM, et al. Revising expectations from rapid HIV tests in the emergency department. Ann Intern Med. 2008;149:153-160.
12. Branson BM, Handsfield HH, Lampe MA, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep. 2006;55(RR-14):1-17.
13. Rapid HIV testing in emergency departments-three U.S. sites, January 2005-March 2006. MMWR Morb Mortal Wkly Rep. 2007;56:597-601.
14. Brown J, Shesser R, Simon G, et al. Routine HIV screening in the emergency department using the new US Centers for Disease Control and Prevention Guidelines: results from a high-prevalence area. J Acquir Immune Defic Syndr. 2007;46:395-401.
15. Lyss SB, Branson BM, Kroc KA, et al. Detecting unsuspected HIV infection with a rapid whole-blood HIV test in an urban emergency department. J Acquir Immune Defic Syndr. 2007;44:435-442.
16. Silva A, Glick NR, Lyss SB, et al. Implementing an HIV and sexually transmitted disease screening program in an emergency department. Ann Emerg Med. 2007;49:564-572.
17. Pant Pai N, Joshi R, Dogra S, et al. Evaluation of diagnostic accuracy, feasibility and client preference for rapid oral fluid-based diagnosis of HIV infection in rural India. PLoS ONE. 2007;2:e367.
18. Phillips KA, Catania JA. Consistency in self-reports of HIV testing: longitudinal findings from the National AIDS Behavioral Surveys. Public Health Rep. 1995;110:749-753.
19. Haukoos JS, Hopkins E, Eliopoulos VT, et al. Development and implementation of a model to improve identification of patients infected with HIV using diagnostic rapid testing in the emergency department. Acad Emerg Med. 2007;14:1149-1157.
© 2009 Lippincott Williams & Wilkins, Inc.