Hightow, Lisa B. MD, MPH*†; Miller, William C. MD, MPH, PhD*†; Leone, Peter A. MD*‡; Wohl, David A. MD*; Smurzynski, Marlene MSPH†; Kaplan, Andrew H. MD*§‖¶
HIV COUNSELING AND TESTING is currently the largest and most costly HIV prevention effort in the United States. Posttest counseling gives high-risk HIV-uninfected persons the opportunity to change their high-risk behavior and potentially, the behavior of their sex and drug-using partners. For HIV-infected persons, this session is a starting point for referral into medical evaluation and treatment.1 Approximately 27% to 66% of HIV-antibody tests are being performed on those who have previously tested negative.2–8 With limited resources, understanding the complex factors associated with repeat testing is critical for HIV prevention efforts.
Those who have tested previously appear to be at higher risk for HIV infection than people who come for a first HIV test.3–7,9,10 The factors associated with repeat testing, however, have not been well characterized. Several studies suggest that those who have tested previously could be more likely to practice unprotected oral sex,4 to have ever been diagnosed with a sexually transmitted disease (STD), and to know people who are HIV-infected.5 Furthermore, a survey of STD clinic subjects found that repeat testing was associated with continued unsafe sexual and/or drug use behavior.6 Understanding the burden of and factors associated with repeat testing among users of STD clinic services has important implications. A randomized, controlled trial targeting those who have tested previously showed significant reductions in future high-risk behaviors among HIV-negative men after a specific, single-session counseling intervention.11
In this study, we assessed the extent of and characteristics associated with repeat HIV testing in a publicly funded STD clinic located in the southeastern United States. Through subgroup analysis we also evaluated the association between repeat testing and HIV seroconversion. Overall, our results suggest that repeat HIV testing is common among patients receiving services at an STD clinic. We also found that a significant percentage of subjects with previous negative tests underwent HIV seroconversion. Results from this study could be used to help identify and target those undergoing testing who are at highest risk for contracting HIV for risk-reduction interventions.
To assess the frequency and predictors of repeat HIV testing, we conducted a retrospective cohort study including persons undergoing an HIV test during their visit to the Wake County, North Carolina, STD clinic between January 1, 1995, and December 31, 2000. The study population included all 101 newly diagnosed HIV-infected subjects and 411 HIV-uninfected subjects randomly selected from persons undergoing testing in the 3-day periods surrounding the test date for each HIV-infected subject. The total number of people who had an HIV-negative test on the day before, the day of, or the day after the date that the cases tested positive was 4113. The medical charts of all persons undergoing testing contain a standardized client record that has detailed information; including posttest return dates and client demographics (self-reported risk exposure[s], sex, age, marital status, and race or ethnicity) and contain visit information dating from January 1, 1989, through the date of entry into the study. Subjects consenting to HIV testing underwent standard pretest counseling and blood tests for HIV, which was performed by the state laboratory located in Raleigh, North Carolina. After each testing episode, clients were scheduled for a posttest counseling session 2 weeks later. The medical record of each subject was reviewed and prior testing episodes were recorded. All patient identifiers were removed before analysis and each subject was linked to a confidential study number. This study was approved by the University of North Carolina at Chapel Hill, Institutional Review Board.
Statistical analyses were conducted using Stata, version 7.0.12 Bivariate relationships between repeat HIV testing and client demographics, self-reported risk exposure behavior(s), principal reason for visit, STD diagnosis, HIV serostatus, and failure to return for posttest counseling appointment at the Wake County Health Department were assessed using chi-squared tests. Variables that had a P value <0.20 in bivariate analyses or that met a priori expectations of relevance were entered into a multivariate logistic regression model to determine variables independently associated with repeat HIV testing. Weighted analysis was done to account for the sampling of the HIV-negative group. A subgroup analysis was performed on those who had tested previously to examine relationships between HIV seroconversion and client demographics, self-reported risk exposure behavior, principal reason for the visit, STD diagnosis, and failure to return for the scheduled posttest counseling appointment. Variables that had a P value <0.05 in bivariate analyses or that met a priori expectations of relevance were entered into a multivariate logistic regression model to determine variables independently associated with HIV seroconversion. The lower P value of <0.05 was chosen to limit the number of variables included in the model given the small sample size within the subgroup. For the purposes of this study, subjects who had tested previously were defined as having undergone >1 previous HIV test at the Wake County Health Department over the period of time available for review (6–10 years).
Of the 508 subjects (99%) with available records (Table 1), 160 (32%) had tested previously. Among those who had tested previously, 46% had been tested twice, 34% had been tested 3 or 4 times, 17% had been tested 5 and 10 times, and 3% had been tested more than 10 times.
Subjects who had tested previously were older (Table 2; unadjusted odds ratio [OR], 1.96; 95% confidence interval [CI], 1.00–3.85 for subjects 40 years or older; referent = subjects aged 30–39 years) and those age 18 years of age or younger were the least likely to have tested previously (OR, 0.92; 95% CI, 0.39–2.13). Persons who had tested previously were less likely to be white or non-white, non-black when compared with blacks (unadjusted OR [whites], 0.29; 95% CI, 0.18–0.49 and OR [non-white, non-black], 0.33; 95% CI, 0.13–0.82).
Repeat testing rates varied significantly by the client’s stated reason for coming to the clinic. Subjects presenting to the clinic for STD-related reasons (STD symptoms or requesting an STD check) or those who presented as a result of a contact referral (sex with a contact who was either diagnosed with an STD or complained of symptoms of an STD) were more likely to have tested previously compared with those subjects presenting to the clinic for the sole purpose of receiving an HIV test (unadjusted OR [STD-related reasons], 3.75; 95% CI, 1.11–12.67 and unadjusted OR [contact referral], 1.39; 95% CI, 0.38–5.11).
We also assessed the association of STD diagnoses with repeat testing. At each clinic visit, clients were diagnosed with 1) no STD, 2) mucosal STD (gonorrhea, chlamydia infection, nongonococcal urethritis, trichomoniasis, mucopurulent cervicitis), 3) ulcerative STD (initial presentation of herpes simplex virus infection or syphilis), or 4) initial presentation of human papillomavirus (HPV). There was no association between having an STD diagnosis and repeat testing. Subjects found to be HIV-infected were not significantly more likely to have had >1 HIV test than those subjects testing HIV-negative (unadjusted OR, 0.72; 95% CI, 0.44–1.18 for HIV-infected compared with those who were HIV-uninfected). Subjects who failed to return for their 2-week posttest counseling return appointment were more likely to have tested previously compared with persons who returned at 2 weeks (unadjusted OR, 1.96; 95% CI, 1.28–2.99). A lack of association was found among marital status, gender, and all self-reported HIV risk behaviors and repeat testing.
In multivariate analysis (Table 2), persons who had tested previously were less likely to be white or non-white, non-black when compared with blacks (adjusted OR [whites], 0.28; 95% CI, 0.16–0.51 and OR [non-white, non-black], 0.34; 95% CI, 0.13–0.88). Subjects presenting to the clinic for STD-related reasons (STD symptoms or requesting an STD check) were more likely to have tested previously compared with those subjects presenting to the clinic for the sole purpose of receiving an HIV test (adjusted OR, 4.93; 95% CI, 1.15–21.02).
We examined the associations between repeat testing and HIV seroconversion among the subgroup of persons who had tested previously (n = 160). In multivariate analysis (Table 3) of those subjects who had tested previously, self-identifying as a man who has sex with men (MSM) or having a history of incarceration remained strongly associated with HIV seroconversion (adjusted OR, 51.82; 95% CI, 9.10–295.13; adjusted OR, 83.98; 95% CI, 17.26–408.69, respectively). Presenting for STD-related reasons (STD symptoms or requesting an STD check) had a negative association with HIV seroconversion (adjusted OR, 0.07; 95% CI, 0.01–0.90) compared with presenting for the sole purpose of requesting an HIV test.
For the 26 subjects who had tested previously and who had HIV seroconversion, the median number of days between the last negative test and first positive test was 1094 days (interquartile range, 235–1762 days). Four subjects had indeterminate results at their previous testing visit. All 4 of these became HIV-seropositive; the median number of days between the intermediate result and first positive test result was 39 days (interquartile range, 20.5–94.0 days).
Our results demonstrate that a significant percentage of those subjects undergoing HIV counseling and testing have been tested previously. Overall, almost one third of subjects undergoing HIV counseling and testing in this STD clinic had previously tested HIV-negative. In accordance with past work,6 we found that those subjects presenting to the clinic for STD-related reasons, including symptoms of an STD, were more likely to have been tested previously. The acquisition of new STDs reflects continued high-risk sexual behaviors that can lead to acquisition of both STDs and HIV. The high repeat testing rate is concerning in light of evidence that indicates an increased risk for HIV infection among persons who have other STDs.13 We found that repeat testing was more common among older subjects, but was less common among whites and nonwhite, nonblack subjects. Among those who had tested previously, HIV seroconversion was more common among MSM and those having a history of incarceration and less common among those presenting to the clinic for STD-related reasons.
Consistent with data presented by others,6,10 older persons (age ≥40 years) were significantly more likely to have been tested previously than subjects in all other age groups after controlling for other variables. This could represent a confounding between testing episodes and the years of sexual activity. Older subjects have likely been sexually active for a longer period of time than subjects in younger age groups and also have had more opportunities to have undergone testing. The design of this study did not allow us to determine true testing rates in units of person-time, which could have accounted for some of these associations. Black subjects were significantly more likely to have tested previously than white and nonblack, Hispanic, and Asian clients after controlling for other variables. Previous work has suggested that testing messages may not reach all racial and ethnic groups equally.10 However, this study was not designed to evaluate the factors responsible for the higher repeat testing rate found among blacks.
A recent study found a high failure to return rate among STD clinic patients undergoing HIV counseling and testing.14 We found that those subjects who failed to return for their 2-week posttest counseling return appointment were more likely to have tested previously compared with persons who returned at 2 weeks. HIV-uninfected clients who do not return for their test results do not receive additional counseling regarding risk reduction and behavioral change and might not be reaping the full benefits of HIV prevention messages.14
A unique feature of this study was our ability to perform a subgroup analysis on those who had tested previously to assess for associations between repeat testing and HIV seroconversion. A significant proportion of those who had tested previously identifying as MSM or with a history of incarceration were found to undergo seroconversion. Similarly, those who had previously tested who presented to the STD clinic for the sole purpose of obtaining an HIV test had a higher likelihood of being HIV-infected than those presenting for STD-related reasons or those presenting as a result of a contact referral. These results could be useful to providers in the STD clinic setting because they attempt to identify high-risk subjects and deliver client-specific HIV counseling and testing messages.
We were surprised to find that 4 of the 100 subjects with a positive test result during the study period had an indeterminate result at a previous testing visit. Thus, 4% (4 of 100) of the population testing HIV-positive were likely to have undergone testing during the period of acute HIV infection as defined by Hecht et al. It has been hypothesized that early antiretroviral treatment during the period of acute infection could augment host immunity, contain viremia, and improve long-term prognosis for treated patients,15 and thus current USPHS/DHHS guidelines recommend that urgent treatment be considered in the setting of acute infection. Moreover, it was found that HIV is readily transmitted by sexual intercourse during early primary HIV infection.16 As much as one half of all HIV transmission could occur during the primary infection interval.17 Of note, 2 of the 4 subjects had 1 or more previous negative tests before their indeterminate result. These findings suggest that future research should evaluate the prevalence and predictors for acute HIV infection in the STD clinic setting.
This investigation has a number of potential limitations. First, these results reflect the experiences of a single publicly funded STD clinic. HIV tests are routinely offered to all clinic attendees regardless of the reason for their clinic visit. Therefore, repeat testing for HIV observed in STD clinics could reflect the routine provision of testing for clients with repeated episodes of STDs and be more a measure of HIV test acceptance rather than of HIV test-seeking behavior. These clients could differ from those individuals who regularly return to testing sites with the sole purpose of getting an HIV test. This study does not specifically address repeat testers who undergo multiple HIV tests unrelated to STD clinic visits. It is important to note, however, that STD clinics performed the largest proportion of all HIV tests in 1998 and were second only to counseling and testing sites in the proportion of HIV-positive test results18; HIV testing during an STD clinic visit is a relatively common occurrence. From the point of view of the clinician, understanding HIV testing behaviors among high-risk STD clinic patients remains an important goal. Although any generalization beyond this study area should be made with care, we believe that our clinic population is similar to other Southeastern states. Secondly, we cannot be sure that subjects did not undergo HIV counseling and testing at other sites during the course of the study. Therefore, our data could represent an underestimate of the true rate of repeat testing in this population. Finally, as a result of the retrospective nature of this study, we were unable to verify clients’ self-reports of their risk behaviors. Some clients might not be open about discussing their sexual and drug use histories with providers, especially during a notoriously time-pressured STD visit. However, we do assume that all patients presenting to a STD clinic are at some risk for acquisition of HIV infection and thus these results could be relevant for any at-risk population.
Those subjects who have tested previously appear to represent a high-risk group on which to focus HIV prevention efforts. A subject’s previous testing history must be reviewed at each new testing visit to allow more targeted prevention messages to be delivered and to attempt to elicit behavior change. An important goal should be the development of a model of the high-risk repeat tester to be used by counselors in the STD clinic setting. Understanding the repeat tester’s motivations, beliefs, and responsiveness to standard HIV counseling efforts is a necessary component of any future research.
1. Solomon L, Landrigan J, Flynn C. Barriers to HIV testing and confidentiality: The concerns of HIV-positive and high-risk individuals. AIDS & Public Policy Journal 1999; 14:147–156.
2. Kalichman SC, Schaper PE, Belcher L, et al. It’s like a regular part of gay life: Repeat HIV antibody testing among gay and bisexual men. AIDS Educ Prev 1997; 9(suppl B):41–51.
3. Leaity S, Sherr L, Wells H, et al. Repeat HIV testing: High risk behaviour or risk reduction strategy? AIDS 2000; 14:547–552.
4. Phillips KA, Paul J, Kegeles S, Stall R, Hoff C, Coates TJ. Predictors of repeat HIV testing among gay and bisexual men. AIDS 1995; 9:769–775.
5. Norton J, Elford J, Sherr L, Miller R, Johnson MA. Repeat HIV testers at a London same-day testing clinic. AIDS 1997; 11:773–781.
6. Schwarcz SK, Spitters C, Ginsberg MM, Anderson L, Kellogg T, Katz MH. Predictors of human immunodeficiency virus counseling and testing among sexually transmitted disease clinic patients. Sex Transm Dis 1997; 24:347–352.
7. MacKellar DA, Valleroy LA, Secura GM, et al. Young Men’s Survey Study Group. 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 2002; 29:76–85.
8. HIV Counseling and Testing in Publicly Funded Sites: Annual Report, 1997 and 1998. Atlanta: Centers for Disease Control and Prevention, 2001:1–20.
9. Fischer JA, Spurlock-McLendon JL, DelGado BP, Melchreit RL. An assessment of the counseling and programmatic implications of repeat HIV testing. National HIV Prevention Conference; August 29–September 1, 1999; Abstract no. 532.
10. Fernyak SE, Page-Shafer K, Kellogg TA, McFarland W, Katz MH. Risk behaviors and HIV incidence among repeat testers at publicly funded HIV testing sites in San Francisco. J Acquir Immune Defic Syndr 2002; 31:63–70.
11. Dilley JW, Woods WJ, Sabatino J, et al. Changing sexual behavior among gay male repeat testers for HIV: A randomized, controlled trial of a single-session intervention. J Acquir Immune Defic Syndr 2002; 30:177–186.
12. STATA, version 7.0. College Station, TX: Stata Corp, 2001.
13. Centers for Disease Control and Prevention. HIV prevention through early detection and treatment of other sexually transmitted diseases—United States. Recommendations of the Advisory Committee for HIV and STD Prevention. MMWR Morb Mortal Wkly Rep 1998; 47(RR12):1–24.
14. Hightow LB, Miller W, Leone PA, Wohl DA, Smurzynski M, Kaplan AH. Failure to return for HIV posttest counseling in an STD clinic population. AIDS Educ Prev 2003; 15:282–290.
15. Rosenberg ES, Altfeld M, Poon SH, et al. Immune control of HIV-1 after early treatment of acute infection. Nature 2000; 407:523–526.
16. Pilcher CD, Eron JJ, Vernazza PL, et al. Sexual transmission during the incubation period of primary HIV infection. JAMA 2001; 286:1713–1714.
17. Cates W, Chesney MA, Cohen MS. Primary HIV infection—A public health opportunity. Am J Public Health 1997; 87:1928–1930.
18. Centers for Disease Control and Prevention. HIV Counseling and Testing in Publicly Funded Sites Annual Report, 1997 and 1998.