HIV counseling and testing is an integral part of the United States HIV prevention program. The Centers for Disease Control and Prevention (CDC) recommended that individuals at risk for HIV receive counseling and testing, and has made this recommendation an important strategy for limiting the spread of HIV disease and reducing morbidity in individuals (1,2). In recent years, HIV counseling and testing programs have expanded dramatically (3). However, despite increased emphasis on HIV counseling and testing, little is known about the HIV testing histories of persons at risk for HIV. Although some data suggest that individuals at greater risk for HIV are more likely to be tested (4,5), questions remain concerning the factors that influence whether an individual receives an HIV test, whether s/he obtains the results of such testing, and the extent to which persons receive multiple tests.
The impact of HIV counseling and testing programs on women is particularly important. Women constitute the fastest growing group of persons with AIDS. In addition, AIDS is now the third leading cause of death of women between the ages of 25 and 44 and the leading cause of death among black women among this age group (6,7). Strategies to implement zidovudine prophylaxis for HIV-positive pregnant women (ACTG 076 protocol) will place even greater importance on counseling and testing programs (8). However, despite the goals of public health programs to encourage women to participate in counseling and testing, to our knowledge there are no data that specifically explore the testing histories of women.
To examine differences between women who have varying testing histories, we administered a questionnaire to women volunteering for a study of HIV disease in women. The purpose of the analysis reported here was to describe the prior HIV testing histories of these women, to compare sociodemographic and risk characteristics of women who had and had not been previously tested, to describe the characteristics of women who had been previously tested but did not return for HIV test results, and to determine the characteristics of women who had multiple HIV tests.
Between April 1993 and January 1994, women ages 17-55 were recruited for a screening protocol to determine eligibility for a longitudinal epidemiologic study of HIV infection in women. Women were recruited from locations likely to have contact with women at high risk of HIV infection, including drug treatment facilities, sexually transmitted disease clinics, HIV care providers, community groups involved with HIV education, and pediatric AIDS clinics. The screening facility, which was also the site of an epidemiological study of HIV infection in injection drug users, was located in a predominantly minority, inner-city neighborhood in Baltimore with a high prevalence of injection drug use and AIDS.
Women who volunteered for the screening component of the study were told they could be selected for an additional study of the health effects of HIV in women. Study volunteers were informed that both HIV-positive and HIV-negative women could participate. Study volunteers were given a questionnaire concerning demographic characteristics; injection drug use, including type of drug injected and frequency of injection; number of sexual partners in the previous 5 years; sexual contact with non-monogamous partners; sexual contact with anonymous partners (unknown partner or one-night stand); self-perceived risk for HIV infection; and prior HIV testing. Women participating in the screening study consented to HIV counseling and testing and were reimbursed $10.00 for participation. The analysis presented here includes self-reported history of HIV testing prior to entering the screening study. This protocol was approved by the Johns Hopkins Institutional Review Board.
The demographic and behavioral characteristics of women who reported ever having been tested for HIV were compared with those who had not had a previous HIV test. A second set of analyses compared women who had been tested and returned for test results with women who had not returned for results, using chi-square statistics. Finally, comparisons were performed between women who reported having multiple tests (defined as one or more tests) and those with only one test. Prior testing status (ever, never, once or more than once) were cross-tabulated with demographic and behavioral variables using chi-square statistics. To simultaneously control for multiple potential confounders, logistic regression procedures were used (9). Variables were included in the logistic regression model if they were found to be statistically significant in bivariate analysis.
Between April 1993 and January 1994, 705 consecutive women volunteered for the questionnaire and HIV screening. The majority of these women had been referred to the study by another participant or friend. Of the study participants, 93.9% self-identified as black and 7.1% self-identified as white, Hispanic, or Native American. The mean age was 33, and 7.7% were currently employed. More than half the women, 52.5%, reported a history of injection drug use; of these, 81.9% reported injection drug use within the past 6 months. Seventy-five percent, or 528 women reported having had a prior HIV test, not including the HIV test performed as part of the screening procedure. Of these, 12.5% with a prior HIV test had not returned for test results, 14.7% reported being HIV seropositive, and 72.8% reported being HIV seronegative. The mean number of prior HIV tests was 1.74 with a range from 0 to 20; the median number of prior tests was two.
To examine the differences between women who had been tested and those who had not, sociodemographic characteristics and risk practices were compared (Table 1). Women who had received HIV testing were significantly more likely to be non-black (p < 0.05), to report having a regular physician (p < 0.01), and to have been referred to the study from a health clinic (p < 0.05). In addition, women who had been tested were also significantly more likely to report having a sex partner who was an injecting drug user (p < 0.05), having a partner who has HIV positive (p < 0.01), and having four or more sex partners in the previous 10 years (p < 0.05). Women who had been tested were also more likely to report a history of injection drug use in the previous 10 years and to report injection use greater than once daily (p < 0.05).
Women who did not return for prior test results were compared with women who had received HIV test results (Table 2). No differences were noted in race, employment status, perceived health status, perceived risk for HIV, or drug use in the previous 6 months. Women who did not return for results were significantly more likely to report anonymous sex. Among women who reported injection drug use in the previous 6 months, women who injected more than once a day were significantly less likely to have returned for HIV test results (p < 0.05).
Table 3 presents selected sociodemographic and risk characteristics of study women who had had more than one HIV test compared with those with only one test (does not include HIV test conducted at study entry). Women with multiple tests were significantly more likely to have a history of drug use and to be current injection drug users. No differences were noted among women with multiple and single tests, by race, perceived health status, presence of a regular physician, perceived risk, or sexual practices.
Table 4 presents the results of the final multiple logistic regression model for having more than one prior HIV test. Variables were included in the model if they were found to be statistically significant in the bivariate analysis. Factors found to be significantly associated with repeated tests included injection drug use within the past 10 years (OR 1.95, 1.01-3.75), injection drug use one to five times a week (OR 1.92, 1.15-3.17), and injection drug use greater than once daily (OR 2.69, 1.69-4.31). The odds ratios for multiple tests increased with heavier drug use. A history of four or more sex partners in the past 10 years was also significantly associated with multiple tests (OR 1.49, 1.00-2.23).
In the United States, HIV testing is a prominent component of HIV prevention efforts. Individuals at high risk for HIV, particularly those involved in injection drug use or high-risk sexual practices, are encouraged to participate in HIV counseling and testing programs. The importance of HIV testing includes the benefit of early medical intervention for those who are HIV seropositive, and identifying and mitigating risk practices for HIV-negative individuals and their sex and drug-using partners. This study found that 75% of the women reported a prior HIV test. In particular, women who report particularly high-risk practices, such as current injection drug use, having a partner who is an injecting drug user, and having an HIV-positive partner, are significantly more likely to have received prior HIV testing. This is consistent with earlier literature that has documented that individuals participating in particularly high risk activities are more likely to have received HIV counseling and testing (8). Our results also demonstrate that many women involved in high-risk drug and sex practices are receiving HIV counseling and testing. However, it is note-worthy that 25% of the women who reported high-risk drug or sexual practices reported never obtaining an HIV test. Although untested women had somewhat fewer HIV risks than those who had been tested, many nonetheless reported engaging in high-risk activities. It is troubling to note that black women, who constitute the fastest growing group with AIDS, are less likely to report being tested and, as a consequence, may not have the opportunity to take advantage of early HIV clinical are.
A critical component of HIV testing is finding out one's test result and obtaining posttest counseling, which can reinforce risk reduction messages and help one assess individual risk. Twelve percent of the women who received prior HIV testing did not return for their test results. The rate of nonreturn is not substantially different than in other studies of men and women, in which rates of 6.3% (10) and 11.5% were observed (11). Although it was not statistically significant, there was an important finding suggesting that women who failed to return for test results were somewhat more likely than women who received test results to have engaged in particularly high-risk practices, including anonymous sex, sex with nonmonogamous partners, increased number of sex partners, and increased frequency of injection drug use. Our disturbing findings that women participating in particularly high-risk activities may be the least likely to return for results have been noted elsewhere (12). While the type of testing facility (i.e., anonymous or confidential) may influence the rate of return, such data were not collected in this study. Additional research on factors that contribute to women not returning for the test results, including factors such as who initiated the request for the test, perceived barriers to obtaining test results, characteristics of the test site, and the characteristics and training of the counselor, should be initiated.
Of the women who reported obtaining any prior HIV test, 62% obtained multiple tests. Multiple testers were more likely to have higher levels of risky behavior. Among our population, it was not the “worried well” who were having multiple tests, but those at particularly high risk for HIV. It is disturbing that despite repeated HIV tests and the associated HIV counseling, high-risk practices continued in this group. This suggests that although the HIV risks associated with certain practices are known, individuals may not have the skills to avoid participation in the high-risk activity. The study findings also suggest that the HIV counseling and risk reduction education that are delivered within the context of HIV testing may not sufficiently influence the behavior of some women engaged in high-risk practices. Several recent reports have also noted the lack of decrease in high-risk sex practices after HIV counseling and testing (13-15), and one study reported an increase in high-risk behaviors following a negative HIV test result (16). Additional research on ways to enhance behavioral change within the testing environment should be undertaken.
Before firm conclusions are drawn, it should be acknowledged that this sample was composed of paid volunteers, who may be different from other women. Study participants may have joined the study in an effort to participate in an ongoing study of HIV infection in women. These women may regard themselves as particularly vulnerable to HIV or may have been motivated to join this study by a desire for HIV-related services. It is possible that this study population represents women who have had unusually frequent interactions with the health care system, and hence would be more likely to have had prior testing. However, only 10% of the participants were referred to this study by a health care provider; the majority of women reported finding out about the study from other women, not from health care settings.
Although this study is limited by the unknown representativeness of the study participants who volunteered for HIV testing as part of a recruitment protocol for a HIV study, and by the cross-sectional design of the study, it may still provide insights into the responses of women to testing. Although there are many possible and varied responses of individuals to the risk of HIV, including decisions not to be tested, not to get test results, or to have repeated HIV tests, the results suggest that individuals who have repeated HIV tests may represent a subset at particularly high risk for HIV. These individuals appear concerned about their HIV status but seem unable to change their behavior. Since this group seems at particularly high risk for infection, these individuals may be an especially important group for intensive public health and behavioral interventions.
Acknowledgment: This study was supported by cooperative agreement #U64/CCU3068 from the Centers for Disease Control and Prevention.
1. Centers for Disease Control. Additional recommendations to reduce sexual and drug abuse-related transmission of human T-lymphotrophic virus type III lymphadenopathy-associated virus. MMWR
2. Centers for Disease Control. Public Health Service guidelines for counseling and antibody testing to prevent HIV infection and AIDS. MMWR
3. Centers for Disease Control. Publicly funded HIV counseling and testing—United States, 1985-1989. MMWR
4. Battle R, Cummings G, Krasnovsky F, Saliba P, Di Clemente R, Siegel D. Is my mommy at risk? HIV testing among low income African American mothers [Abstract]. Int Conf AIDS
5. McFarland W, Fischer-Ponce L, Katz MH. Repeat negative HIV testing in San Francisco: magnitude and characteristics. Am J Epidemiol
6. Centers for Disease Control and Prevention. Update: mortality attributable to HIV infection among persons aged 25-44 years—United States, 1994. MMWR
7. Selik RM, Chu SY, Buehler JW. Human immunodeficiency virus infection as the leading cause of death among young adults in U.S. cities and states. JAMA
8. Centers for Disease Control. Zidovudine for the prevention of HIV transmission from mother to infant. MMWR
9. Brown BW, Hollander M. Statistics: a biomedical introduction
. New York: John Wiley and Sons, 1977.
10. Centers for Disease Control. Differences between anonymous and confidential registrants for HIV testing—Seattle, 1986-1992. MMWR
11. Centers for Disease Control. Self-reported HIV-antibody testing among persons with selected risk behaviors—Southern Los Angeles County, 1991-1992. MMWR
12. Behrendt C, Kendig N, Clark JP, Vlahov D. HIV infection and AIDS among U.S. prison inmates. J Crime Justice
13. Zenilman JM, Erickson B, Fox R, Reichart CA, Hook EW III. Effect of HIV posttest counseling on STD incidence. JAMA
14. Landis S, Earp JL, Koch GG. Impact of HIV testing and counseling on subsequent sexual behavior. AIDS Education Counseling
15. Ickovics JR, Morrill AC, Beren SE, Walsh U, Rodin J. Limited effects of HIV counseling and testing for women
: a prospective study of behavioral and psychological consequences. JAMA
16. Otten Jr MW, Zaidi AA, Wroten JE, Witte JJ, Peterman TA. Changes in sexually transmitted disease rates after HIV testing and posttest counseling, Miami, 1988 to 1989. Am J Public Health