Introduction
HIV testing in combination with appropriate counseling is an important tool in the public health response to AIDS. Counseling and testing programs designed to promote knowledge of serostatus can facilitate behavioral change, assist partners to negotiate safer sexual practices, and allow partner notification and referral to medical services for HIV-infected persons. Recent improvements in antiviral therapy, prophylaxis for opportunistic infections, and effective interventions to prevent perinatal transmission have given new urgency to identifying HIV-infected persons. However, many persons with HIV infection remain unaware of their serostatus, and frequently, those who have been tested learn that they are HIV-infected too late to benefit from these advances [1,2]. In publicly funded clinics, follow-up for HIV results remains poor, with 40% of clients failing to return for test results and post-test counseling (this ranges from 59% in STD clinics to 17% in HIV counseling and testing sites) [3,4]. New HIV testing technology has allowed provision of results and result-specific counseling on the day of initial visit, and has the potential to increase the efficiency of HIV counseling and testing.
The Single Use Diagnostic System (SUDS) HIV-1 test (Murex, Norcross, Georgia, USA) is the only rapid HIV test approved by the US Food and Drug Administration and available for use in the United States. SUDS uses enzyme immunoassay (EIA) technology to obtain results within 10 min, and is comparable to the standard EIA in diagnostic accuracy with a sensitivity of 99.9% and a specificity of 99.6% [5,6]. Similar to EIA, a negative test result on SUDS does not require further confirmation, and therefore negative results can be given out at the initial visit, eliminating the need for over 90% of clients to return for results. A reactive test requires confirmation by Western blot or immunofluorescence assay, so clients with a reactive test need to return [7,8]. Clients with a reactive rapid test can be told of their likelihood of being infected with HIV given a positive preliminary test, and can be counseled accordingly. This may increase the likelihood of an infected client returning for confirmatory results and further counseling. In a theoretical decision model that assumed an 80% return rate with rapid HIV testing, these procedures were more cost-effective than standard procedures in publicly funded clinics [9]. However, these rapid testing procedures have not been evaluated in real clinic settings.
Methods
From 6 July through 1 October 1993, on-site rapid HIV testing with provision of same-day results and counseling was implemented for all patients of the Dallas County Sexually Transmitted Disease (STD) Clinic and for all clients of the Anonymous Test Clinic (ATC). To evaluate HIV counseling and testing using the rapid assay, the rapid test procedures were compared with the standard pre-and post-test procedures during a baseline period of 26 January through 19 April 1993, consisting of the same number of clinic days (n = 60) as the implementation period. These procedures were compared with respect to (i) number of clients receiving results and post-test counseling, (ii) client satisfaction, (iii) counselor acceptance, (iv) cost, and (v) effectiveness. Changes in the clinic procedures are summarized in Fig. 1.
Testing procedures
Patients in the STD Clinic were asked about HIV risks and offered confidential HIV testing as part of their routine care. Clients of the ATC presented requesting an HIV test. At both sites, patients gave informed consent for HIV testing, received standard pre-test counseling, and peripheral blood was drawn. Blood was sent to the Texas state laboratory for HIV-1 EIA, and, if repeatedly reactive, Western blot analysis. Results were available within 2 weeks, and if patients returned for their scheduled appointments, they received their results and post-test counseling [10]. During the rapid testing period, on-site SUDS was performed on all specimens drawn for HIV testing in the STD Clinic and the ATC. Repeatedly reactive samples were sent to the State Laboratory for a confirmatory Western blot.
Counseling procedures
During standard pre-test counseling, the counselor asked for informed consent, conducted a risk assessment, provided prevention counseling and drew blood for testing. After about 2 weeks the client was given an appointment to return for the results and the counselor provided additional prevention counseling. If the test was positive, post-test counseling included emotional support, referral for medical services, and for social and psychological services if needed. Local health department staff attempted to locate STD Clinic patients with a positive test who did not return for results, and attempted to bring them back to the clinic, or to provide results and counseling in the field.
The counseling sequence and test result messages were modified for rapid testing. In the STD Clinic, patients provided their consent and risk information to a triage counselor. Patients then saw a clinician who obtained a blood sample, and provided clinical STD services while the rapid HIV test was being performed. After completion of the clinical visit, a counselor then notified them of their results and prevention counseling took place. Clients who had a reactive rapid test result were scheduled to return for Western blot results. Partner notification and referral to medical care occurred if the Western blot was positive. In the ATC, there was no triage counselor. Consent and risk assessment was performed by the counselor, as was the phlebotomy. Prevention counseling was carried out while the rapid HIV test was being performed. This led directly to notification of results and a continuous counseling session. For a client with a reactive result, a follow-up appointment was scheduled for Western blot results. These modifications to clinic procedures are shown in Fig. 1.
In a previous study, the performance of the SUDS assay in the Dallas County Laboratory was validated, and was found to be diagnostically equivalent to the EIA [6]. Thus, negative results were reported as HIV-negative. Although a reactive result requires confirmation, these were reported as 'preliminary positive' during the initial visit, and the clients were counseled on the basis of these results. The likelihood of a positive Western blot following a positive SUDS test (positive predictive value) had been previously determined to be 81% in the ATC and 88% in the STD Clinic. Accordingly, on the basis of previously reported evaluations of the quantitative interpretation of qualitative probabilistic expressions in common clinical use, a series of phrases was adopted to communicate to patients the likelihood of being infected with HIV given a preliminary positive result [11,12]. In the ATC the following phrases were recommended: 'likely to be infected'; 'a good chance of being infected'; '…usually means you are infected'. In the STD Clinic, with a greater positive predictive value and a greater likelihood that a SUDS positive was a true positive, these phrases were modified by recommending the following: 'probably infected'; 'very likely (or highly likely) infected'; 'a very good chance of being infected'. In practice, based on their individual assessment of the client's risks during counseling, the counselors either strengthened or qualified the phrases used to communicate the probability of infection given a preliminary positive result.
Patient and client satisfaction
To assess the clients' response to rapid testing, 114 successive patients in the STD Clinic and 111 successive clients of the ATC were surveyed after they received counseling. The response rate was 100%. The survey was conducted in September 1993 when rapid testing was in place at both sites. Respondents were asked to reply 'yes', 'no', or 'not sure' to questions about basic services. Respondents were then asked to rank their responses on a five-point Likert-type scale from 1 to 5 for questions about their preferences regarding same-day results, their level of stress with the procedure, and how well they understood the meaning of the preliminary result that they received.
Counselor acceptance
Counselors in the STD Clinic and ATC were involved from the beginning in developing the counseling procedures. Before implementation of rapid testing, and 1 month after implementation, focus groups of counselors assessed and compared counselor reactions to the process, and identified potential problems with the newly modified counseling procedures.
Effectiveness of counseling procedures
Since the basic content and quality of the counseling was unchanged as a result of the study, changes in risk reduction associated with same-day test results could be measured and compared with the standard 2-week approach to providing results. To measure the differences in effectiveness of these two approaches in promoting HIV risk reduction, the rate of return to the clinic with an incident STD was examined for 1 year after HIV testing for all of the STD Clinic patients who received the rapid testing procedures. Incident STD for 1 year after HIV testing were compared with a control group consisting of all patients who received standard pre-and post-test counseling and testing during the baseline period. An incident STD was defined as a new laboratory-confirmed diagnosis of HIV infection, gonorrhea, chlamydia, trichomoniasis, primary, secondary or latent syphilis.
Economic analysis
The analysis was performed from both a societal perspective, which included all costs to both patients and providers, and from the clinic perspective, which included only the costs of service delivery. Cost data were collected for counseling and testing services at the STD clinic and ATC during the time when the standard and the rapid test procedures were in use, and thus were measured in 1993 US dollars. Fixed costs included rent, janitorial services, utilities, equipment, administrative costs such as clinic management, and laboratory personnel. Variable costs included provider time, and materials used for testing. Provider costs for counselling, testing and administrative functions were calculated by multiplying the hourly wages, including fringe benefits, for each personnel category by the average time it took to perform a specific task. The time spent by providers was estimated using time and motion studies, and supplemented with counselor time diaries from 100 standard counseling sessions, and 100 rapid test sessions. Participant costs were measured with a survey of 102 clients from the ATC and 100 patients of the STD Clinic, asking respondents about the money and time they spent traveling to the clinic. The time patients spent attending the clinic was assessed in the time and motion studies. The value of patient time, in terms of wages forgone, was calculated using the 1993 Texas Bureau of Labor Statistics average hourly wage rate of US$ 13.15 for the Dallas area.
Results
Knowledge of serostatus
At the ATC during the 60 days of clinic operation when rapid testing was implemented, 984 clients were counseled and tested, a 7% increase from the comparable 60-day baseline period of standard counseling and testing (Table 1). During the baseline period, 95% of HIV-negative and 86% of HIV-positive clients returned for their results and post-test counseling. During rapid testing, 99% of HIV-negative clients received their results and post-test counseling on the same day as testing (a 4% increase over baseline), and 100% of HIV-positive clients received their preliminary results and counseling, and all of them returned for confirmatory results and further post-test counseling (a 16% increase over baseline).
At the STD Clinic, 1493 patients were counseled and tested, a 29% increase from the baseline period (Table 1).
During the baseline period, 30% of HIV-negative and 79% of HIV-positive clients received their results and post-test counseling. During rapid testing, 93% of HIV-negative clients received their results and post-test counseling on the same day as testing (a 210% increase over baseline), and 97% of HIV-positive clients received their preliminary results, and returned for confirmatory results and further post-test counseling (a 23% increase over baseline). During the baseline period, less than one-half (45%) of HIV-positive clients returned to the clinic on their own, and 34% of HIV-positive clients required Heath Department-initiated field follow-up to learn their results. During rapid testing, 94% of HIV-positive persons returned for confirmatory results and post-test counseling (a 109% increase), whereas only 3% of HIV-positive clients required Health Department follow-up.
Patient/client preferences
Overall, most of the clients preferred the rapid test (Table 2). The majority of clients surveyed liked receiving their HIV results on the same day that they were tested (92%) and understood the meaning of their test results (89%). Of those previously tested, 88% responded that they preferred the rapid test.
There were no significant differences between men and women in their responses. When stratified by race/ethnicity and by testing site, white respondents and those tested in the ATC were significantly (P < 0.05) more likely to report that they understood their rapid test result, and to report that they preferred receiving their results on the same day that they were tested.
Counselor acceptance
As assessed in focus groups before and after implementation, counselors were initially reluctant to adopt the rapid test. Their main concern was that rapid testing procedures would cause additional stress on the clients. Counselors expressed a lack of confidence in the validity of a positive rapid test result and were concerned about counseling a client on the basis of a preliminary test. There was concern that a preliminary positive result that was a false-positive would cause the client unnecessary psychological distress. There was also concern that the new procedures would be more stressful for the counselor. Specifically, giving results on the same day limited the time counselors had to prepare emotionally for the stress of telling a client they were HIV-infected.
After 1 month's experience with the new counseling and testing procedures, most of these concerns had been resolved. Counselors believed they became more efficient with their time. After adjusting to the new procedures, counselors did not report increased stress in themselves, nor did they report increased stress in their clients in response to the procedures. There were no reports of bad reactions from clients receiving preliminary HIV-positive results, either by clients who subsequently learned their confirmatory test was positive or by clients who learned their initial test was falsely positive. There were reports that some clients receiving preliminary positive results experienced reduced psychological distress, as this was a way of breaking the news gently.
Effectiveness
The review of clinic records for the cohorts of STD Clinic patients found no documented HIV seroconversions. At 6 months after initial testing, 4.7% (54 out of 1160) of clients who received the standard test and 4.0% (57 out of 1437) of clients who received the rapid test returned to the clinic with a new STD [odds ratio (OR), 0.85; 95% confidence interval (CI), 0.6-1.3]. At 1 year, 6.0% (70 out of 1160) of clients who received the standard test and 5.9% (84 out of 1437) of clients who received the rapid test returned with an incident STD (OR, 0.97; 95% CI, 0.7-1.4).
Economic analysis
Rapid testing was found to cost less than standard testing procedures (Table 3). The total cost from a societal perspective of the standard counseling and testing procedure at the ATC was US$ 141 per test and US$ 151 per person receiving their results. The total cost of the rapid counseling and testing procedure at the ATC was US$ 130 per test and US$ 131 per person receiving their results; this indicated a cost saving of US$ 11 per test and US$ 20 per person receiving results. In the STD Clinic, the total cost of the standard counseling and testing procedure was US$ 113 per test and US$ 375 per person receiving their results. The total cost of the rapid counseling and testing procedure was US$ 102 per test and US$ 95 per person receiving their results and post-test counseling; this indicated a cost saving of US$ 11 per test and US$ 280 per person receiving results and counseling.
From the clinic perspective (omitting the costs borne by the patients), rapid testing increased the cost at the ATC by US$ 2 per test, and decreased the cost by US$ 4 per person receiving results. Rapid testing resulted in a cost saving to the STD Clinic of US$ 6 per test and US$ 165 per person receiving results.
Mean patient costs, including travel time, travel costs, and time spent in the clinic were US$ 41 for the standard procedures. For patients who tested negative and did not need to return, these costs dropped to US$ 23. However, for those who tested positive, these costs rose to US$ 48. This was due to more time being spent in the STD Clinic counseling the HIV-positive clients.
The total cost to set up a similar rapid test laboratory and run it for 1 year was estimated to be US$ 99 500. This included the annual salaries of the technicians, the actual cost to the clinic to purchase the equipment used, and the cost to purchase laboratory supplies, test kits, and reagents based on a monthly volume of 760 tests. This estimate represented the incremental cost of establishing a laboratory in an existing clinic's unused space, and did not include the cost of renting the additional space, and the cost of additional utilities, and janitorial services used, which would have added a total of US$ 3905.
Discussion
The use of on-site, rapid HIV testing offers several potential advantages to the current strategy. Among these advantages are the following: (i) Clients can receive counseling that is relevant to their HIV infection status on the same day they come in. (ii) Client satisfaction with the process can be improved because they receive their results sooner. (iii) More clients receive their test results, since most do not have to return for their results and post-test counseling. (iv) Clinics are able to provide negative results at less cost since only one visit is needed.
The compression of pre-and post-test counseling session into a single session, with the absence of a 2-week waiting period for HIV-negative results, raised concern that the counseling associated with rapid testing may not be as effective as the standard procedures in promoting HIV risk reduction. Using one indirect measure of HIV risk, acquisition of new STD following HIV testing, no difference was found between STD Clinic patients counseled using rapid test procedures and patients receiving standard pre-and post-test counseling. Although larger trials may be needed to definitively resolve some of these issues, these data indicate that program managers considering the use of rapid testing to improve service delivery can be reassured that counseling associated with rapid testing does not appear to be less effective.
Overall, the majority of those counseled and tested understood their rapid test results, and preferred these procedures. It is unclear why slightly more white clients of the ATC understood their results, and preferred receiving their results on the day of initial visit at slightly higher rates than the black clients of the STD Clinic. Because of the pronounced differences in the clientele at testing sites (82% of ATC attendees were white and 8% were black, whereas 17% of STD Clinic attendees were white and 68% were black), our survey could not assess whether these differences were due to the cultural or socioeconomic factors, or to differences in the testing milieu between the STD Clinic, where clients received the HIV test as part of the diagnosis and treatment of another STD, and the ATC, where clients presented requesting an HIV test.
This study has demonstrated that rapid, on-site HIV testing is feasible, is preferred by clients, and can result in significant improvement in the delivery of counseling and testing services. More than 97% of those tested in both clinics were uninfected. Thus, for the majority of clients, counseling and testing could be accomplished in a single clinic visit. This was cheaper and more convenient to the client, resulted in more people receiving their HIV-negative results, and was cost-saving to the clinic. Furthermore, this single session counseling did not appear to result in a greater amount of behavioral disinhibition or increased HIV risk. For HIV-infected clients, delivering preliminary positive results improved the number who returned to the clinic for confirmatory results and post-test counseling and reduced the need for Health Department follow-up. This resulted in considerable cost-savings in the STD Clinic where return rates were poor, and considerable increases in the number of infected clients who learn their serostatus, who could be referred for additional prevention services and could benefit from medical treatment.
Acknowledgements
The authors are grateful to assistance of R. Wilson, T. Schenk, D. Hutcheson, A. Freeman, C.G. Beeker, K. Stark, C. Orsini, A.R. Gerber, A. Haddix, D. Holtgrave and T. Peterman. Special thanks go to the HIV counselors and the disease intervention specialists in Dallas for their input, participation, and support throughout the project.
References
1. Wortley P, Chu SY, Diaz T, Ward J: HIV testing patterns: where, why and when persons with AIDS were tested for HIV? AIDS 1995, 9:487-492.
2. Berrios DG, Hearst N, Coates TJ, et al.: HIV antibody testing among those at risk for infection: the national AIDS behavioral surveys. JAMA 1993, 270:1576-1580.
3. Valdiserri RO, Moore M, Gerber AR, Campbell CH, Dillon B, West GR: A study of clients returning for counseling after HIV testing: implications for improving rates. Public Health Rep 1993, 108:12-18.
4. Centers for Disease Control and Prevention: HIV Counseling and Testing in Publicly Funded Sites: 1993-1994 Summary Report. Atlanta: US Public Health Service; 1996.
5. Murex: Summary Basis of Approval. Bethesda: Food and Drug Administration, US Department of Health and Human Services; 1992.
6. Kassler WJ, Haley C, Jones WK, Gerber AR, Kennedy EJ, George JR: Performance of a rapid, on-site human immunodeficiency virus antibody assay in a public health setting. J Clin Microbiol 1995, 33:2899-2902.
7. Centers for Disease Control: Public Health Service guidelines for counseling and antibody testing to prevent HIV infection and AIDS. MMWR 1987, 36:509-515.
8. Centers for Disease Control: Update: serologic testing for antibody to human immunodeficiency virus. MMWR 1988, 36:833-845.
9. Farnham PG, Gorsky RG, Holtgrave DR, Jones WK, Guinan ME: Counseling and testing for HIV prevention: costs, effects, and cost-effectiveness of more rapid screening tests. Public Health Rep 1996, 111:44-53.
10. Centers for Disease Control and Prevention: Technical guidance on HIV counseling. MMWR 1993, 42 (RR-2):11-16.
11. Bryant GD, Norman GR: Expressions of probability: words and numbers [letter]. N Engl J Med 1980, 302:411.
12. Mosteller F, Youtz C: Quantifying probabilistic expressions. Stat Sci 1990, 5:2-34.
© Lippincott-Raven Publishers.