Approximately 25% of the estimated 850,000 to 950,000 people infected with the HIV in the United States are unaware of their infection.1 The Centers for Disease Control and Prevention (CDC) recommends that persons with ongoing risks be retested for HIV periodically2 (eg, at least annually for sexually active men who have sex with men [MSM]).3 Many people at risk are not tested for HIV,4,5 however, or are not tested regularly despite ongoing risks.6 Qualitative7 and survey8 studies have shown that clients perceive several barriers to traditional HIV counseling and testing: the need for a clinic appointment, venipuncture, face-to-face counseling, a wait for test results, and a return visit to the clinic. Those interviewed preferred oral fluid testing, rapid testing, and written materials instead of face-to-face counseling as well as access to test results by telephone. Although the CDC's revised guidelines for HIV counseling, testing, and referral support these alternatives and encourage practitioners to offer HIV testing services in nontraditional settings,2 alternative approaches have met resistance from testing providers because of concerns about decreased prevention effectiveness and client well-being.9
Few data have been available to establish the relative benefits, costs, and feasibility of alternatives for HIV counseling and testing in outreach settings. We conducted a randomized trial comparing 4 HIV counseling and testing alternatives and determined the net effect on how many clients received test results for each alternative.
The selection of alternative HIV counseling and testing approaches was based on the results of focus groups and interviews7as well as on a preference survey8 administered to clients at the 3 study sites in Seattle: a needle exchange program and 2 bathhouses frequented by MSM. Four strategies, each of which was offered by randomly determined day, were offered at the 3 sites during 1999 through 2000: (1) traditional serum testing (venipuncture and a return visit for test results) with standard counseling (face-to-face counseling before testing), (2) rapid serum testing (venipuncture and testing with the Single Use Diagnostic System for HIV-1 [SUDS; Murex/Abbott, Norcross, GA]) with same-day test results and standard single-session counseling, (3) oral fluid testing (oral fluid collection with OraSure [OraSure Technologies, Bethlehem, PA]) with standard counseling, and (4) traditional serum testing with the choice of pretest written materials or standard counseling. Consistent with the current practice in the Seattle-King County Public Health testing programs, the choice of receiving results by telephone or in person was offered to all participants. Sample sizes for each study site were adjusted for randomization by cluster.10 Offering each strategy on a minimum of 20 testing days to 20 eligible clients would detect 10% to 15% differences (power = 0.8, α = 0.05, 2-sided) in acceptance of testing and receipt of test results, depending on the study site. During the study, 1 staff member offered each client entering the site free HIV testing (whichever strategy had been randomly selected for the day). Prospective participants were asked whether they had been tested and, if so, the date of their most recent HIV test and whether they had been offered free HIV testing at the site during the past 3 months. English-speaking clients 14 years of age or older who responded to the interviewer's questions, were not known to be HIV-positive, and who had not been tested for HIV within the past 3 months were eligible. No incentives were offered for HIV testing or for study participation. Demographics were estimated for all clients approached. The interviewer also gave clients who accepted testing an estimate of how long they might need to wait for their HIV testing, which was recorded in the recruitment log. For 4 hours during each session, a second staff member counseled participants and obtained specimens or performed rapid testing. Specimens that tested positive with the rapid test were retested with Western blot analysis. This work was approved by the University of Washington Human Subjects Committee (No. 27-0287-C/B) and by the CDC Institutional Review Board (No. 2209).
Data were analyzed by using the Statistical Package for the Social Sciences, version 9 (SPSS, Chicago, IL) and Stata, version 6.0 (Stata Corporation, College Station, TX). Multiple logistic regression was used to tabulate and analyze the outcomes, including acceptance of testing, completion of testing, receipt of test results, choice of written or face-to-face pretest counseling, and choice of telephone or face-to-face posttest counseling. Specific follow-up contrasts compared the effects on each outcome of all levels of multilevel factors such as the type of test offered against all other levels rather than merely comparing each with a single reference category, with Wald tests on the logistic regression coefficients. The needle exchange was modeled separately, but the 2 bathhouses were modeled as a single site. Univariate logistic models were computed for each candidate predictor, and all significant univariate factors for the needle exchange and bathhouses were adjusted for each other in a multiple regression model. Predictors examined for association with acceptance were estimated age, sex, and race; previous offers of testing at the site; quartile of the study; and identity of the recruiter. Association with completion of specimen collection was examined for the same predictors as well as previous testing within 6 months, identity of testing staff, and estimated wait time. Association with receipt of results was examined for the same predictors as acceptance as well as for previous testing within 6 months, education, income, insurance status, and numerous substance use and sexual behavior measures. Values from the univariate regression models are reported here as odds ratios (ORs); the multiple regression models were not substantively different. Cross-site comparisons were made by using Wald tests for a 3-level site factor, or its interaction with other factors, in a multiple logistic regression model. χ2 tests of association were used to determine whether there was any association between factors that were not randomized (eg, whether the racial profile of the participants at a site differed during specific periods of the study). No correction was made for multiple comparisons; we report raw P values.
Cost analysis is presented separately.11
During 221 days, 17,010 clients were offered testing (10,058 at the needle exchange on 122 days and 6952 at the bathhouses on 99 days). At the needle exchange (Fig. 1), 2675 (27%) people declined to respond to the staff member's approach and 3874 (53%) of the 7383 who responded were eligible for testing. Of the 3509 who were ineligible, 3075 (88%) had been tested during the past 3 months, 121 (4%) were HIV-positive, 72 (2%) did not understand English, and 241 (7%) were ineligible for other reasons (eg, being too high to give informed consent). At the bathhouses (Fig. 2), 1490 (21%) declined to respond to the staff member's approach and 3140 (58%) of the 5462 who responded were eligible for testing. Of the 2322 who were ineligible, 1913 (82%) had been tested during the past 3 months, 174 (8%) were HIV-positive, 100 (4%) did not understand English, and 135 (6%) were ineligible for other reasons. A mean of 39 clients at the needle exchange and 33 clients at the bathhouses was approached and determined to be eligible during a 4-hour session. Of eligible persons at the needle exchange, 20% were estimated to be in their 20s or younger, 69% were male, 62% were white, 21% were black, and 6% had never been tested. Of eligible persons at the bathhouses, 22% were estimated to be in their 20s or younger, 80% were white, 6% were black, and 6% had never been tested.
At the needle exchange compared with the bathhouses, clients who were tested had a lower level of education, more were homeless, and fewer had health insurance or a personal health care provider (Table 1). Although most clients had been tested for HIV before, more than 60% had not been tested during the past 6 months, and most had engaged in high-risk behaviors such as needle sharing or unprotected intercourse since their most recent test.
At each site, the proportion of eligible clients agreeing to be tested differed significantly by strategy (Table 2). At the needle exchange, acceptance of each of the 3 alternative strategies was significantly higher than acceptance of traditional testing: 8.7% agreed to traditional testing with standard counseling (reference), 11.9% agreed to traditional testing when counseling was optional, 14.1% agreed to rapid testing, and 19.5% accepted oral fluid testing. The relative acceptance for each of the other strategies is compared in Table 3. For example, acceptance of oral fluid testing was also significantly greater than acceptance of rapid testing (OR = 1.4; P = 0.006) or traditional testing with optional pretest counseling (OR = 1.7; P < 0.001). Blacks were less likely than whites to accept testing (OR = 0.47; P < 0.01), but the data for blacks and whites revealed a similar pattern of increased acceptance of the alternative strategies. Of those offered the options for pretest counseling at the needle exchange, 56 (77%) chose written materials instead of face-to-face counseling.
At the bathhouses (see Table 2), acceptance of oral fluid testing (22.8%) and rapid testing (21.2%) was significantly higher than acceptance of traditional testing with standard counseling (15.8%, reference). Although 82 (81%) of bathhouse clients offered the option for counseling chose written materials instead of face-to-face counseling, acceptance of testing with optional pretest counseling was not significantly different from acceptance of traditional testing with standard counseling (see Table 3). Acceptance of oral fluid testing was also similar to that of rapid testing.
Only 61% of clients at the needle exchange and 78% at the bathhouses who agreed to HIV testing were actually tested. Completion of testing correlated well with shorter estimated waiting times (at the needle exchange, r = 0.42; at the bath-houses, r = 0.58; P < 0.01 for both). Waiting time depended on the testing strategy and the number of clients who accepted testing during a given session. For example, time per client was longer for rapid testing because the counselor also performed the rapid test, and time per client was shorter when the client chose the option of written pretest materials because the counselor did not have to provide face-to-face counseling. At the needle exchange, 79% of clients who had agreed to testing were tested when there was no wait compared with 53% when clients were told the estimated waiting time was up to 15 minutes and 32% when the estimated waiting time was up to 3 hours. The trends at the bathhouses were similar to those at the needle exchange, although clients at the bathhouses were willing to wait relatively longer and were more likely to complete testing (OR = 2.4; P < 0.001) even when adjusting for waiting times. At the bathhouses, 88% of clients completed testing when there was no wait and 91% completed testing when estimated waiting times were up to 15 minutes. Even when the waiting time was expected to be 2 hours, 55% completed testing.
Overall effectiveness (defined as the proportion of eligible clients receiving test results) was highest for oral fluid testing at the needle exchange and highest for rapid testing at the bathhouses, although the differences between oral fluid and rapid testing did not reach significance (see Table 3). More clients who completed testing at both sites received their test results after rapid testing compared with traditional testing: at the needle exchange, 66 (83%) of 80 versus 27 (56%) of 48 (OR = 3.7; P = 0.002), and at the bathhouses, 102 (99%) of 103 versus 82 (74%) of 111 (OR = 36.1; P < 0.001). The proportion of those clients tested receiving results for the other testing alternatives was similar to that for traditional testing. At the time of testing, all clients except those tested with rapid tests were given the option of receiving their test results via telephone; 31% of needle exchange clients and 93% of bathhouse clients chose to do so.
The test results for 17 study participants were HIV-positive. At the bathhouses, all 3 clients with positive rapid test results returned for their confirmatory test results: by Western blot analysis, 2 were confirmed to be HIV-positive and 1 was HIV-negative. Those who were HIV-positive received coordinated early treatment and partner notification counseling when they received their confirmatory results. Of the 13 HIV-positive persons tested at the bathhouses with nonrapid alternatives, 4 received their test results. At the needle exchange, 2 persons were HIV-positive (1 tested on an oral fluid day and 1 on a day when counseling options were offered), both of whom received their test results.
Despite ongoing testing in these outreach settings, the number of eligible clients at the needle exchange and the bathhouses did not decrease. During 10-week periods at the beginning and end of the study, the proportion of clients who responded to staff and were eligible for testing increased from 28% to 46% at the needle exchange and from 42% to 48% at the bathhouses (P < 0.001). Although the proportion of clients who reported that they had been tested for HIV within the past 3 months increased from 31% to 46% (P < 0.001) at the needle exchange and from 35% to 39% (P < 0.05) at the bathhouses, the proportion of clients who declined to respond when approached by the study's outreach staff decreased from 45% to 17% at the needle exchange and from 25% to 16% at the bathhouses (P < 0.001). Thus, the overall number of clients eligible for testing increased.
Data from this randomized study indicate that in outreach settings, alternative HIV counseling and testing strategies help to maximize the number of clients who learn their HIV test results. Traditional HIV testing with standard counseling was least effective at providing clients with knowledge of their HIV status. At the needle exchange and bathhouse sites, oral fluid and rapid testing resulted in significantly more participants learning their HIV status. Each alternative evaluated in this study demonstrated different advantages: most clients accepted oral fluid testing, a higher proportion of those who accepted were tested with the use of the written pretest counseling materials because of shorter waiting times, and more of the clients tested with rapid tests received their results. Telephone results likely increased the proportion of clients who learned their status compared with face-to-face counseling at a return visit12 but not as much as rapid testing. Offering a combination of techniques such as rapid oral fluid testing with written pretest materials may be the best way to increase the number of persons who learn their HIV status.
Our results are subject to several limitations. Because more than 90% of study participants had been tested previously, these data may not be generalizable to populations with less testing experience. The differences in effectiveness between rapid testing and oral fluid testing at the bathhouses may have reached significance with a study of longer duration. This study was also conducted with the SUDS test, which required venipuncture and took more counselor time than recently available tests. More simple rapid tests and additional testing staff would have decreased waiting times and increased the testing rates for some strategies. We also found that the number of people who declined to speak with outreach workers diminished over time. This suggests that as testing programs in outreach settings become more routine, clients may be more comfortable talking with staff and the number of persons who accept testing may increase. Even in sites such as these with high levels of testing experience, the testing program continued to have a steady flow of clients who needed to be tested because of ongoing risk behaviors.
The recent approval by the Food and Drug Administration of a point-of-care rapid HIV test suitable for whole blood and oral fluid (OraQuick Rapid HIV-1 Antibody Test) presents another option that is likely to make outreach testing easier and more acceptable.13 Because this test requires little hands-on staff time, it is likely to decrease the waiting time associated with rapid testing (an important impediment in our study). The OraQuick Rapid HIV-1 Antibody Test performed on oral fluid is likely to have higher acceptability than the rapid test we evaluated, which required venipuncture.
It is clearly more difficult to recruit clients for testing in outreach settings, where their primary interest is not related to health care. In this study, we discovered a few techniques that improved acceptance.14 In the past, outreach workers offering HIV testing typically made an announcement and then waited for clients to come forward. In this study, acceptance improved when staff members approached each client, offered the free HIV test randomly selected for that day, and asked whether the person had been tested and, if so, the date of the most recent test. This interaction facilitated more personal interaction; over time, fewer and fewer clients declined the approach at the needle exchange and bathhouses. Before beginning this study, several testing providers and community members voiced concern about the acceptability to clients and bathhouse owners of providing rapid HIV testing and counseling in bathhouses. After clients demonstrated their interest, even eagerness, for onsite HIV testing, these concerns were allayed and additional bathhouse owners asked to participate in the project. Testing at the needle exchange continues to be well received by clients and volunteer staff as well. Currently, the new OraQuick Rapid HIV-1 Antibody Test is being offered successfully by the health department in both venues and in a community-based mobile testing program to reach people of color (People of Color Against AIDS Network).
Additional data on the effect on risk behaviors from rapid HIV testing at high-risk venues such as bathhouses are needed to understand the full effect on HIV prevention. The hope is that persons who receive positive test results are less likely to engage in unprotected sex with discordant partners. Clients who receive negative test results might choose to engage in protected sex to remain HIV-negative, but another possible outcome is that receiving negative test results might lower inhibitions and result in an increase in high-risk behaviors,15 especially if it is assumed that others at the bathhouse have also tested negative. The availability of rapid HIV testing in bathhouses might allow couples to make more informed choices about their sexual behaviors. If rapid testing becomes more widely available and disclosing test results becomes a norm, couples may choose to have unprotected sex only with others of concordant HIV status. Despite the false-negative test results from infected persons during the “window” period before antibodies develop, such partner selection could prevent HIV transmission.16
Our study suggests several ways in which testing programs in needle exchanges and bathhouses could help clients to learn their HIV status, for example, by including active recruiting, using rapid and oral fluid testing, offering pretest counseling options (at needle exchanges), and providing the option of receiving results by telephone. Each of these strategies is associated with different advantages. Alternative testing strategies should be implemented to make testing programs at needle exchanges and bathhouses more effective. The relative effect of counseling options on HIV risk behaviors is unknown but merits evaluation as a means of increasing the number of persons who accept testing and receive their results. Offering the option of receiving HIV test results by telephone after traditional blood testing or oral fluid testing is particularly important at bathhouses, where clients are less likely to want to return for results. During consideration of optimal strategies for specific venues, it is important to consider the overall effectiveness of each strategy as well as the cost per test result received.12 When resources are limited, these data can help to ensure selection of the most cost-effective testing and counseling options in outreach venues.
The authors thank Ken Hanes; Kevin Henderson; the HIV Alternative Testing Study staff; and the staff of the Downtown Needle Exchange, the bathhouses, and the sexually transmitted disease clinic for their hard work and support. We also thank Maria Witthans for superlative administrative and editorial support.
1. Fleming P, Byers RH, Sweeney PA, et al. HIV
prevalence in the United States, 2000 [abstract]. In: Program and Abstracts of the 9th Conference on Retroviruses and Opportunistic Infections, Seattle, February 2002. Page 56. Alexandria, VA: Foundation for Retrovirology and Human Health.
2. Centers for Disease Control and Prevention. Revised guidelines for HIV
counseling, testing, and referral. MMWR Recomm Rep
3. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines 2002. Centers for Disease Control and Prevention. MMWR Recomm Rep
4. Anderson JE, Carey JW, Taveras S. HIV
testing among the general US population and persons at increased risk: information from national surveys, 1987-1996. Am J Public Health
5. Kellerman SE, Lehman JS, Lansky A, et al. HIV
testing within at-risk populations in the United States and the reasons for seeking or avoiding HIV
testing. J Acquir Immune Defic Syndr
6. MacKellar DA, Valleroy LA, Secura GM, et al. Repeat HIV
testing, risk behaviors, and HIV
seroconversion among young men who have sex with men: a call to monitor and improve the practice of prevention. J Acquir Immune Defic Syndr
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
8. 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
9. Spielberg F, Critchlow C, Vittinghoff E, et al. Slow diffusion of home HIV
-specimen collection: provider concerns at odds with client preferences. Sex Transm Dis
10. Donner A, Birkett N, Buck C. Randomization by cluster. Sample size requirements and analysis. Am J Epidemiol
11. Spielberg F, Jackson S, Varghese B, et al. HIV
testing with oral fluids and rapid tests is more effective and less costly [abstract P63]. Presented at the 2002 National STD Prevention Conference, Atlanta, Georgia. March 2004.
12. Schluter WW, Judson FN, Baro'n AE, et al. Usefulness of human immu-nodeficiency virus post-test counseling by telephone for low-risk clients of an urban sexually transmitted diseases clinic. Sex Transm Dis
13. Centers for Disease Control and Prevention. Approval of a new rapid test for HIV
antibody. MMWR Morb Mortal Wkly Rep
14. Spielberg F, Branson BM, Goldbaum GM, et al. Designing an HIV
counseling and testing program for bathhouses: the Seattle experience with strategies to improve acceptability
. J Homosex
15. Weinhardt LS, Carey MP, Johnson BT, et al. Effects of HIV
counseling and testing on sexual risk behavior: a meta-analytic review of published research, 1985-1987. Am J Public Health
16. Varghese B, Maher JE, Peterman TA, et al. Reducing the risk of sexual HIV
transmission: quantifying the per-act risk for HIV
on the basis of choice of partner, sex act, and condom use. Sex Transm Dis
Keywords:© 2005 Lippincott Williams & Wilkins, Inc.
HIV; HIV testing barriers; acceptability; HIV prevention; alternative strategies; oral fluid HIV testing; rapid HIV testing