sign in sign up
<<
Home , HIV

CHILD POLICY This PDF document was made available from www.rand.org as a public CIVIL JUSTICE service of the RAND Corporation. EDUCATION

ENERGY AND ENVIRONMENT Jump down to document AND HEALTH CARE 6

INTERNATIONAL AFFAIRS

NATIONAL SECURITY The RAND Corporation is a nonprofit research POPULATION AND AGING

PUBLIC SAFETY organization providing objective analysis and effective AND TECHNOLOGY solutions that address the challenges facing the public SUBSTANCE ABUSE and private sectors around the world. TERRORISM AND HOMELAND SECURITY

TRANSPORTATION AND INFRASTRUCTURE Support RAND Browse Books & Publications Make a charitable contribution

For More Information Visit RAND at www.rand.org Explore RAND Health View document details

Limited Electronic Distribution Rights This document and trademark(s) contained herein are protected by law as indicated in a notice appearing later in this work. This electronic representation of RAND intellectual property is provided for non-commercial use only. Permission is required from RAND to reproduce, or reuse in another form, any of our research documents for commercial use.  RESEARCH AND PRACTICE 

Sex Without Disclosure of Positive HIV in a US Probability Sample of Persons Receiving Medical Care for HIV

| Daniel H. Ciccarone, MD, MPH, David E. Kanouse, PhD, Rebecca L. Collins, PhD, Angela Miu, MS, James L. Chen, MPH, Sally C. Morton, PhD, and Ron Stall, PhD

It is difficult to identify a more charged Objectives. We estimated the proportion of HIV-positive adults who have any sexual issue in AIDS prevention than that of contact without disclosure and the proportion of their sexual partnerships that involve nondisclosure of positive HIV status to sex- unprotected sex without disclosure. ual partners. Seropositive individuals who Methods. We drew participants from the HIV Cost and Services Utilization Study do not disclose their status to sexual part- (n=1421). Interviews assessed disclosure and sexual activities with up to 5 recent ners are often portrayed as dangerous pari- partners. ahs.1 As of 1999, 31 states had statutes Results. Overall, 42% of the gay or bisexual men, 19% of the heterosexual men, and making sexual contact without disclosure a 17% of all the women reported any sex without disclosure, predominately within nonex- criminal offense.2 researchers, clusive partnerships (P<.001). Across all groups, 13% of serodiscordant partnerships involved unprotected anal or vaginal sex without disclosure, with no significant differ- responding to data reporting sexual contact ence between groups. without disclosure by HIV-positive persons, Conclusions. Risky sex without disclosure of serostatus is not uncommon among have called for interventions to promote in- people with HIV. (Am J Public Health. 2003;93:949–954) creased sexual responsibility and to in- crease rates of consistent safer sex among HIV-positive persons.3,4 als’ behavior varies according to their part- METHODS Public discourse on this topic is heavily ners’ HIV status and use. influenced by a few well-publicized cases in We report data from the Risk and Pre- Sample Description which individuals who knew they were HIV vention Study subset of the nationally rep- Respondents were participants in the positive reportedly infected a series of un- resentative probability sample of the HIV HCSUS, a national probability sample of suspecting partners. The public health sig- Cost and Services Utilization Study 2864 persons aged 18 years or older with nificance of nondisclosure, however, de- (HCSUS). We describe (1) the extent to known HIV infection who made at least 1 pends on both its and the which Americans in treatment for HIV visit to a nonmilitary, nonprison medical riskiness of the behaviors that occur with- have sex without disclosure, (2) the extent provider other than an emergency depart- out disclosure—that is, the extent to which to which this population has unprotected sex ment in the contiguous United States during sex without disclosure is unprotected sex without disclosure, and (3) the extent to the first 2 months of 1996. Full details of the between partners. Previous which unprotected sex without disclosure HCSUS design are available elsewhere.16 ,17 studies show that not disclosing positive occurs within serodiscordant sexual part- The Risk and Prevention Study subset used HIV serostatus to at least some sexual part- nerships. To the best of our knowledge, this for our study consisted of 1421 HCSUS par- ners is common,5–7 that HIV-positive per- is the first study to report data on the ex- ticipants. Eligible members of the HCSUS sons are more likely to disclose their status tent of sex without disclosure from a proba- sample were those who were interviewed in to steady partners than to nonsteady part- bility sample of HIV-positive adults in the English at the HCSUS baseline interview, ners,8–10 and that unprotected sex without United States and the first to provide com- whose sex was unambiguous on the basis of disclosure occurs within both types of part- parative data across risk groups. Because HCSUS data, and who participated in the sec- nerships.5,9 Much of the previous research data were collected on partnerships, we can ond follow-up HCSUS interview, conducted on this topic focused on specific subpopula- examine disclosure at both the dyad and in- from August 1997 through January 1998 tions, such as gay and bisexual men,9,11 dividual levels. We hope that this analysis (n=2205). We drew 1794 individuals from men in general,12 or women.7,13,14 Studies will help inform policy and intervention de- this group, sampling randomly after stratifying that included both sexes or multiple risk bates on the best ways to encourage disclo- by primary HCSUS sampling unit, type of groups8,10,15 generally did not use suffi- sure among HIV-positive persons and safer health care provider, age, ethnicity, and self- ciently large sample sizes to support de- sex practices among the broad range of described sexual orientation. We randomly tailed group comparisons. Nor did previous Americans infected with HIV or at risk of sampled one third of eligible White studies examine whether and how individu- infection. aged 40 years, 44% of eligible White gay

June 2003, Vol 93, No. 6 | American Journal of Public Health Ciccarone et al. | Peer Reviewed | Research and Practice | 949  RESEARCH AND PRACTICE 

men aged 39 years and younger, and 100% of engaging in oral, anal, and vaginal sex; to examine the prevalence of abstinence, sex of all other groups. We conducted interviews consistency of condom use with each activity; only with disclosure, any sex without disclo- from September through December 1998. disclosure; and timing of sexual activity rela- sure, and any unprotected anal or vaginal sex The completion rate was 79%, and the re- tive to disclosure. without disclosure by risk group. To adjust sponse rate after adjustment for known mor- standard errors and statistical tests for the dif- Measures tality was 84%. ferential weighting and complex sample de- Sexually active respondents were those re- The Risk and Prevention Study subsample sign, we used linearization methods18 , 2 0 avail- porting any oral, anal, or vaginal sex in the was weighted to represent a target population able in the statistical package Stata (Stata preceding 6 months; all others were consid- of 197063 HIV-positive adults receiving Corp, College Station, Tex). ered abstinent. Aggregating across partners, medical care in the 48 contiguous states of Analyses at the partnership level focused we classified each active respondent into 1 of the United States in 1996 and surviving until on 1273 serodiscordant partnerships, exam- 2 mutually exclusive categories reflecting be- 19 9 8. The analytic weights take into account ining the prevalence of unprotected anal havior in the past 6 months: (1) had sex only differential selection probabilities, nonre- and vaginal sex within these partnerships. after disclosure, and (2) had any sex without sponse, multiplicity, and attrition.18 Men who Results for the multiple partners of a re- disclosure. Respondents who refused to an- identified themselves as “gay,” “bisexual,” or spondent might be correlated. In a sensitiv- swer whether there had been any sex before “heterosexual” and all women were included ity analysis, we adjusted for this design ef- disclosure were classified as missing data in the analysis reported in this article. Men fect by treating each individual and his or (n=7); those who said they didn’t know or who specified “other” or who did not report her associated partners as a primary sam- who were otherwise missing data were their sexual orientation were excluded pling unit and by using the same analytic treated as having had sex only after disclo- (n=24). Tables show proportions weighted to weights. Because the sensitivity analysis sure (n=45). represent the population, as well as un- yielded the same conclusions as our original Within partnerships, we first combined weighted sample sizes. analysis, which did not account for the cor- the “sex without disclosure” variable with relation among partners, and because the consistency of condom use (“always” vs any Survey Instrument and Procedures sensitivity analysis reduced our statistical other response) for anal and for vaginal sex The Risk and Prevention Study covered power owing to the design effect, we pres- and then aggregated across partners to cre- sexual activities, attitudes, and beliefs related ent only our original analysis results in this ate the variable “any unprotected sex with- to HIV transmission. Interviews were con- article. out disclosure” in the past 6 months. Disclo- ducted in person. Interviewers asked ques- sure was not assessed for specific sexual tions and entered responses for most of the RESULTS activities. In addition, because timing specifi- interview using a laptop computer. The com- cally of unprotected sex was not assessed in puter was turned over to respondents for the The demographic characteristics of the relation to timing of disclosure (we asked section of the survey concerning sexual be- sample differed significantly according to sex- about timing of any sex, not timing of unpro- havior and disclosure. Studies have found the ual risk group (Table 1). Gay or bisexual men tected sex), it is possible that some respon- use of such techniques to result in more accu- more often were White, resided in the west- dents had unprotected sex only after disclos- rate reporting of sensitive and socially unde- ern region of the United States, and had ing their positive status. However, we sirable behaviors.19 higher levels of education and higher incomes consider this sequence unlikely with serodis- Sexual behavior was assessed for 6 months compared with the other 2 groups. Hetero- cordant partners, who are the focus of our before the interview date. The interview de- sexual men were by and large older, of Afri- analyses of this variable. fined oral, anal, and vaginal sex for respon- can American descent, and of lower educa- We classified respondents as having an ex- dents. Those who reported engaging in any of tion and income than the gay or bisexual clusive partnership if they reported having these categories of sexual activity during the men. Women, similar to heterosexual men, sex with only 1 partner in the past 6 months 6-month period answered a set of questions were often from the South and of lower in- and described this partner as a primary rela- regarding each (of up to 5) of their most re- come and education but tended to be youn- tionship partner or spouse. cent partners in the past 6 months. Some re- ger than both men’s groups, with one third We classified partnerships as serodiscor- spondents had a spouse or primary relation- being younger than age 35 years. A higher dant if a partner was reported to be HIV neg- ship partner with whom they were sexually proportion of heterosexual men than of ative (n=576), the partner’s HIV status was active over the 6-month interval but who was women or of gay or bisexual men reported unknown to the respondent (n=696), or the not among their 5 most recent partners. injecting drugs before their HIV diagnosis, respondent refused to answer (n=1). These participants described their behavior compared with women and gay or bisexual with their primary partner in a final set of Analysis men. Compared with the other risk groups, questions, which resulted in the assessment of We conducted analyses at both the respon- gay or bisexual men had been HIV positive 6 partners for these respondents. The ques- dent level and the partnership level. We used significantly longer and were more likely to tions covered partner HIV status; frequency the full analysis sample of 1397 respondents have progressed to AIDS.

950 | Research and Practice | Peer Reviewed | Ciccarone et al. American Journal of Public Health | June 2003, Vol 93, No. 6  RESEARCH AND PRACTICE 

TABLE 1—Demographic Characteristics by Risk Groupa insertive sex to without disclosure in the past 6 months. Characteristic All Groups Gay/Bisexual Men Heterosexual Men Women P Sexually active members of our sample re- ported sex within 1273 serodiscordant part- Sample size 1397 606 287 504 nerships in the past 6 months. Approximately Target population 197 063 109 132 42 920 45 011 half of the sexually active gay or bisexual Age, % <.0001 men (58%), heterosexual men (46%), and 20–39 y 47 49 30 57 women (47%) had any serodiscordant sexual ≥40 y 53 51 70 43 partners during the 6 months before the in- Ethnicity, % <.0001 terview. Almost one third of the gay or bisex- African American 32 15 53 55 ual men (30%) had 2 or more serodiscordant Hispanic 13 10 15 16 partners, whereas fewer women (10%) and White 51 70 31 26 heterosexual men (9%) had 2 or more se- Other 3 5 1 2 rodiscordant partners (P<.001). Education, % <.0001 Table 3 shows the percentage of serodis- <12 y 24 10 38 43 cordant partnerships involving unprotected 12 y 28 25 32 30 sex without disclosure. There were no statis- 13–15 y 27 30 23 22 tically significant differences between risk ≥16 y 21 34 7 4 groups overall. Among gay or bisexual men Income ($), % <.0001 who had serodiscordant partnerships, nearly 0–10000 44 32 56 62 all of the unprotected anal and vaginal sex 10000–25000 25 24 26 27 without disclosure occurred in nonexclusive ≥25000 31 44 19 11 partnerships. Gay or bisexual men’s relation- History of injection drug 20 11 41 24 <.0001 ships were significantly more likely to in- use prior to HIV volve unprotected sex without disclosure diagnosis, % with a nonexclusive partner of unknown or AIDS indicator condition, % 42 46 38 34 .0143 negative serostatus than were women’s rela- Mean days since tested 2668 (2573, 2764) 2929 (2797, 3061) 2376 (2266, 2487) 2311 (2194, 2427) .0514 tionships (P <.001). Five percent of women HIV positive (95% reported not disclosing their HIV-positive confidence interval) status in serodiscordant exclusive partner- aPercentages are weighted to represent the population of HIV-positive individuals who received medical care in the United ships, compared with 1% to 2% of all men States in 1996 and survived to 1998. (P =.065). Across risk groups, most unprotected sex without disclosure in serodiscordant part- Abstinence rates for all 3 risk groups (P <.001). Among those who had engaged nerships appeared to involve mutual nondis- were fairly high and were significantly in sex without disclosure, gay or bisexual closure (i.e., with a partner of unknown HIV higher among heterosexual men than men were much more likely to be involved serostatus). Among gay or bisexual men among gay or bisexual men (P =.007 for in a nonexclusive rather than an exclusive who had serodiscordant partnerships, most heterosexual men vs gay or bisexual men) partnership, whereas heterosexual men and of the unprotected sex without disclosure (Table 2). Rates of any sex without disclo- women were equally likely to be in either occurred in nonexclusive partnerships with sure differed according to sexual risk group. type of partnership. partners of unknown HIV status. Of the Forty-two percent of gay or bisexual men re- The risk that sexual activity would lead to 13.8% of serodiscordant gay or bisexual ported any sex without disclosure, com- transmission varied by type of activity. partnerships in which there was unprotected pared with 19% of heterosexual men and Among gay or bisexual men who reported sex without disclosure, more than three 17% of women (P <.001). Among gay or bi- any sex without disclosure in a nonexclusive quarters (10.9% of all) involved nonexclu- sexual men, most sex without disclosure oc- partnership, 42% engaged only in oral or re- sive partners whose HIV status was un- curred within nonexclusive partnerships, ceptive (not shown). Gay or bisexual known, as opposed to partners known to be whereas the rates of nondisclosure within men were more likely than the other 2 HIV negative. exclusive partnerships were relatively low. groups to report unprotected anal or vaginal The percentage of partnerships in which Thirty-five percent of gay or bisexual men sex without disclosure than the other 2 unprotected sex occurred without disclosure reported any sex without disclosure in a groups (P<.001). However, although 16% of was similar in seroconcordant and serodiscor- nonexclusive partnership, compared with gay or bisexual men reported such behavior, dant partnerships. We saw no differences 9% of heterosexual men and 9% of women far fewer—3.2%—reported unprotected anal across risk groups.

June 2003, Vol 93, No. 6 | American Journal of Public Health Ciccarone et al. | Peer Reviewed | Research and Practice | 951  RESEARCH AND PRACTICE 

TABLE 2—Respondent-Level Analyses: Disclosure of HIV-Seropositive Status to Sexual 1996 and does not include those who began Partnersa receiving HIV care between early 1996 and late 1998, these numbers should be consid- Gay/Bisexual Men (n=606), Heterosexual Men (n=287), Women (n=504), ered a lower-bound estimate. Weighted % (95% CI) Weighted % (95% CI) Weighted % (95% CI) When we focus more narrowly on the peo- Unweighted (N=1397) ple with the greatest risk of transmission—that Abstinent 28 (22, 33) 39 (34, 44) 34 (30, 38) is, those who report having unprotected anal Sex only with disclosure 29 (24, 35) 41 (36, 47) 48 (42, 54) or vaginal sex without disclosure—our results Any sex without disclosure 42 (34, 49) 19 (14, 23) 17 (13, 20) suggest that 17400 gay or bisexual men, Nonexclusive partnerships 35 (29, 41) 9 (6, 12) 9 (5, 12) 2000 heterosexual men, and 2900 women Exclusive partnerships 6 (3, 9) 10 (7, 13) 8 (5, 11) engaged in this behavior during the 6-month Unprotected anal or vaginal sex 16 (10, 21) 5 (2, 7) 7 (4, 9) reporting period. These estimates may be without disclosure high because our variable would misclassify anyone who engaged in protected sex before Note. CI=confidence interval. aSex is defined to include vaginal and insertive and receptive oral and anal sex unless otherwise stated. disclosure and unprotected sex after disclo- sure during the 6-month interval. Among those with seroconcordant partners (most often gay or bisexual men), this pattern is not unlikely. Nevertheless, these numbers are TABLE 3—Partnership-Level Analysis: Unprotected Anal or Vaginal Sex Without Disclosure large enough to suggest that substantial num- in Serodiscordant and Seroconcordant Partnerships Within the Past 6 Months bers of new HIV could occur Weighted % (95% Confidence Interval) among partners of HIV-positive persons who Partnerships of Partnerships of Partnerships of do not disclose their status. Gay/Bisexual Men Heterosexual Men Women The evidence that disclosure increases use of with serodiscordant partners is Serodiscordant Partnerships Analysis mixed.12 Still, prevention efforts designed to Unweighted number of serodiscordant partnerships n=796 n=167 n=310 promote disclosure in addition to reducing (n=1273), weighted % (95% confidence interval) unsafe sex among HIV-positive persons may Unprotected anal or vaginal sex without disclosure 13.8 (10.4, 17) 8.8 (1.6, 15.9) 9.5 (4.7, 14.3) yield important public health benefits. For ex- With HIV-negative partner 2.5 (1.1, 4.0) 0.9 (0, 2.6) 3.8 (0.1, 7.5) ample, the increasing availability of postexpo- With HIV-unknown partner 11.3 (8.4, 14.2) 7.9 (0.9, 14.9) 5.7 (3.3, 8.1) sure prophylaxis makes it possible for HIV- Nonexclusive partnerships 12.5 (9.4, 15.6) 6.7 (0.2, 13.2) 4.5 (2.8, 6.2) negative partners who know they are at risk With HIV-negative partner 1.6 (0.5, 2.6) 0 (0, 0) 1.3 (0, 2.9) to obtain treatment in the case of condom With HIV-unknown partner 10.9 (8.0, 13.9) 6.7 (0.2, 13.2) 3.2 (1.1, 5.2) failure during sex. Exclusive partnerships 1.3 (0, 2.6) 2.1 (0, 5.3) 5.0 (0.6, 9.3) Our results reveal substantially higher rates With HIV-negative partner 1.0 (0, 2.2) 0.9 (0, 2.6) 2.4 (0.1, 4.8) of sex (including protected and ) with- With HIV-unknown partner 0.4 (0, 0.9) 1.2 (0, 3.8) 2.5 (0.1, 4.9) out disclosure among gay or bisexual men Seroconcordant Partnerships Analysis than among heterosexual men or women. Unweighted number of seroconcordant partnerships n=303 n=64 n=101 This suggests that the norms regarding disclo- (n=468), weighted % (95% confidence interval) sure may be quite different among gay or bi- Unprotected anal or vaginal sex without disclosure 13.0 (6.5, 19.6) 6.1 (0.1, 12.1) 5.1 (1.3, 9.0) sexual men than they are among the other Nonexclusive partnerships 10.8 (4.4, 17.3) 2.8 (0, 6.5) 2.2 (0, 4.9) risk groups. Within the gay community, the Exclusive partnerships 2.2 (0.3, 4.1) 3.3 (0, 6.6) 2.9 (0, 5.8) prevalence of HIV infection is substantially higher than it is among heterosexuals, provid- ing a basis for HIV-positive gay or bisexual DISCUSSION erosexual men, and 7500 women—all HIV- men to assume that their partners are aware infected—engaging in sex without disclosure of HIV transmission risk even if they do not The results of this study indicate that sex in our reference population of individuals disclose their serostatus. Moreover, public without disclosure of HIV status is relatively who were in care for HIV in early 1996 and health messages urging gay men to “act as if common among persons living with HIV. The who survived until follow-up and were eligi- every partner is HIV positive” may have con- rates of sex without disclosure found in our ble to be interviewed in the fall of 1998. Be- tributed to norms that make disclosure op- sample of HIV-positive individuals translate cause the reference population is smaller than tional. Among heterosexual men and women, into 45300 gay or bisexual men, 8000 het- the entire population receiving care in early the perceived and actual risk that a partner is

952 | Research and Practice | Peer Reviewed | Ciccarone et al. American Journal of Public Health | June 2003, Vol 93, No. 6  RESEARCH AND PRACTICE 

seropositive is quite low unless the person is quently, nonexclusive partnerships—which agnosed but were not receiving medical care an injection drug user. In view of these gener- often may be short-term relationships in for HIV. Moreover, the people we studied had ally low risks, it cannot fairly be assumed that which assumptions about the other person’s all been receiving care for at least 2.5 years. one’s partner is aware of and accepts HIV serostatus can easily be erroneous—may be Because many individuals do not seek treat- transmission risk unless there has been ex- those in which the greatest risk for HIV trans- ment until they begin having symptoms, this plicit disclosure. mission lies. Preventive approaches for gay or study represents those whose HIV disease is The proportions of people who engage in bisexual men that focus on more enduring re- more advanced, on average, than would be sex without disclosure differ not only by risk lationships as the source of HIV transmission found in the population of all diagnosed HIV- group but also according to the types of part- may miss the subpopulations at greatest risk positive persons. Patterns of disclosure and nerships they have. Gay or bisexual men who for HIV transmission. sexual behavior may well differ between peo- have sex without disclosure are much more Because we collected detailed information ple with early-stage HIV illness who have not likely to be in nonexclusive than in exclusive on sexual risk behavior and disclosure for yet sought care and the people represented in partnerships. Exclusive relationships are likely each individual partner, we were able to con- our study. to involve substantial commitment, and indi- duct analyses at both the dyad and the indi- In addition, as a study of people receiving viduals may feel a greater responsibility to vidual level. Across risk groups, 13% of medical care, our study underrepresents disclose to their partners in such relation- serodiscordant partnerships involved unpro- those with poor access to care, including the ships. Such relationships may align more with tected anal or vaginal sex without disclosure; uninsured, minorities, and persons with low the traditional heterosexual relationship no significant differences were found between incomes. Also, all data were self-reported. model than with casual relationships of gay or groups overall. The partners of HIV-positive Some HIV-positive persons may have been bisexual men. Together, these factors may ac- gay or bisexual men represent a much larger reluctant to admit that they engage, without count for the higher prevalence of unpro- portion of the population of partners of per- informing their partners, in sexual behaviors tected sex in nonexclusive partnerships in this sons with HIV. Thus, targeting interventions that may transmit HIV; accordingly, the be- risk group. Among women and heterosexual to this class of relationships should have a havioral estimates reported here may be con- men who have sex without disclosure, the greater impact. However, among the serodis- sidered lower-bound estimates. However, our proportions do not differ significantly by type cordant partners of persons with HIV, the sample had already been interviewed several of partnership. partners of gay or bisexual men do not ap- times, establishing a relationship with the In all 3 risk groups, most of those who re- pear to be at higher relative risk of unpro- larger HCSUS study, and we used computer- ported engaging in sex without disclosure also tected sex without disclosure. assisted self-interviewing methods, which reported having only protected sex or oral Although our study does not examine indi- have been shown to improve the accuracy of sex, both of which pose less risk of transmis- vidual reasons for nondisclosure, significant data obtained from self-report.33 These fac- sion than unprotected anal or vaginal sex. disincentives to disclosure exist at the socio- tors may have reduced underreporting to This suggests that people who do not disclose cultural level. As we enter the third decade of some extent. In addition, because the “timing their HIV status often take steps to reduce the AIDS epidemic, the stigma related to of disclosure” variable and the sexual activity HIV transmission risk to their partners—or HIV-positive status continues to influence the variables were ascertained separately, we do that they consider that disclosure is not neces- behavior of persons living with HIV and not know the specific (i.e., whether high- or sary, given that they have taken these steps. AIDS.2,27 Disclosure is undoubtedly compli- low-risk) sexual activities that occurred before However, because these lower-risk activities cated by perceived fears of rejection,28 dis- disclosure. Finally, we acknowledge that no still carry some risk of transmission (e.g., from crimination,29 and violence from partners and clear relationship exists between disclosure condom failure), they are of public health others.30,31 Women especially may fear retri- and unprotected sex.12 concern. Moreover, use of unilateral risk re- bution for disclosure of positive serosta- Many intriguing questions remain. What is duction strategies is ethically indefensible, in tus.13 ,14 , 3 0 , 3 2 Among gay men, HIV status has the relationship between partner selection, that such strategies do not allow one’s partner become a defining characteristic that creates on the basis of serostatus, and disclosure? To the opportunity of exercising informed choice social barriers between individuals of differ- what degree are HIV-positive persons making about what level of risk is acceptable.21–23 ing serostatus. Interventions to reduce stigma sexual risk decisions on the basis of their It is frequently reported in the prevention at the community or societal level deserve partners’ disclosure or lack thereof? Is it un- literature that the majority of unprotected sex further attention. ethical to have sex without disclosing one’s among male couples occurs within monoga- This study had several limitations that status when one’s partner also does not dis- mous relationships.24–26 However, when is- should be kept in mind when interpreting the close? Whose responsibility is it to disclose? sues of disclosure and HIV serodiscordance results. First, the study focused only on HIV- Is the ethical obligation to disclose greater are taken into account, most of the unpro- positive persons who were receiving medical for the HIV-positive person? The data re- tected sex without disclosure among these care. The reference population did not in- ported here suggest that these questions men occurs within relationships that do not clude HIV-positive persons who did not know should be in the forefront of HIV prevention meet our definition of exclusivity. Conse- they were infected or those who had been di- interventions for HIV-positive populations.

June 2003, Vol 93, No. 6 | American Journal of Public Health Ciccarone et al. | Peer Reviewed | Research and Practice | 953  RESEARCH AND PRACTICE 

Such interventions should focus on specific References Baseline Methods Technical Report: Weighting, Imputa- relationships and contexts in which disclo- 1. Sobo E. Self-disclosure and self-construction tion, and Variance Estimation. Santa Monica, Calif: among HIV-positive people: the rhetorical uses of RAND; 1999. MR-1060-AHCPR. sure is most likely to affect behavior. Further stereotypes and sex. Anthropol Med. 19 97;4:67–87. analyses of sexual risk, HIV-positive status, 19.Metzger D, Koblin B, Turner C, et al. Randomized 2. Shriver MD, Everett C, Morin SF. Structural inter- controlled trial of audio computer-assisted self- and disclosure among HIV-positive persons ventions to encourage primary HIV prevention among interviewing: utility and acceptability in longitudinal have the potential to make such interventions people living with HIV. AIDS. 2000;14(suppl 1): studies. Am J Epidemiol. 2000;152:99–106. S57–S62. more effective. 20. Kish L, Frankel MR. Inference from complex sam- 3. R. AIDS prevention—sexual ethics and re- ples. J Royal Stat Soc B. 1974;36:1–37. sponsibility. N Engl J Med. 1996;334:1540–1542. 21. Bayer R, Stryker J. Ethical challenges posed by About the Authors 4. Wenger NS, Kusseling FS, Beck K, Shapiro MF. clinical progress in AIDS. Am J Public Health. 19 97;87: At the time of the study, Daniel H. Ciccarone was with the Sexual behavior of individuals infected with the human 159 9–1602. Urban Health Study, Department of Family and Commu- . The need for intervention. nity Medicine, University of California San Francisco and Arch Intern Med. 1994;154:1849–1854. 22. Bennett R, Draper H, Frith L. Ignorance is bliss? HIV and moral duties and legal duties to forewarn. the Center for AIDS Prevention Studies, University of Cali- 5. Marks G, Richardson JL, Maldonado N. Self- J Med Ethics. 2000;26:9–15. fornia San Francisco. David E. Kanouse, Rebecca L. disclosure of HIV infection to sexual partners. Am J Collins, Angela Miu, and Sally C. Morton are with RAND, Public Health. 19 91;81:1321–1322. 23. Ainslie D. AIDS and sex: is warning a moral obli- Santa Monica, Calif. James L. Chen is with the California 6. Mansergh G, Marks G, Simoni JM. Self-disclosure gation? Health Care Anal. 2002;10:49–66. Epidemiologic Investigation Service, Los Angeles. Ron of HIV infection among men who vary in time since Stall was with the Division of General Internal Medicine, 24.Crepaz N, Marks G, Mansergh G, Murphy S, Miller seropositive diagnosis and symptomatic status. AIDS. Center for AIDS Prevention Studies, University of Califor- LC, Appleby PR. Age-related risk for HIV infection in 19 95;9:639–644. nia San Francisco, during the time that this analysis was men who have sex with men: examination of behav- completed. 7. Simoni JM, Mason HR, Marks G, Ruiz MS, Reed ioral, relationship, and serostatus variables. AIDS Educ Requests for reprints should be sent to Daniel H. Cic- D, Richardson JL. Women’s self-disclosure of HIV in- Prev. 2000;12:405–415. carone, MD, MPH, Urban Health Study, Department fection: rates, reasons, and reactions. J Consult Clin 25. Hays RB, Kegeles SM, Coates TJ. Unprotected sex of Family and Community Medicine, 3180 18th St, Psychol. 19 95;63:474–478. and HIV risk taking among young gay men within Ste 302, San Francisco, CA 94110 (e-mail: ciccaron@ 8. Stein MD, Freedberg KA, Sullivan LM, et al. Sex- boyfriend relationships. AIDS Educ Prev. 19 97;9: fcm.ucsf.edu). ual ethics. Disclosure of HIV-positive status to partners. 314–329. This article was accepted October 8, 2002. Arch Intern Med. 1998;158:253–257. 26.Wagner GJ, Remien RH, Carballo-Dieguez A. “Ex- 9. Wolitski RJ, Rietmeijer CA, Goldbaum GM, Wil- tramarital” sex: is there an increased risk for HIV Contributors son RM. HIV serostatus disclosure among gay and bi- transmission? A study of male couples of mixed HIV D.H. Ciccarone contributed to the interpretation of the sexual men in four American cities: general patterns status. AIDS Educ Prev. 1998;10:245–256. analyses and led the writing of the article. D.E. and relation to sexual practices. AIDS Care. 19 9 8;10: 27.van der Straten A, Vernon KA, Knight KR, Gómez Kanouse and R.L. Collins contributed to the conception 599–610. CA, Padian NS. Managing HIV among serodiscordant and design of the study and the interpretation of the 10. Niccolai LM, Dorst D, Myers L, Kissinger PJ. Dis- heterosexual couples: serostatus, stigma and sex. AIDS analyses and made substantial contributions to all sec- closure of HIV status to sexual partners: predictors and Care. 1998;10:533–548. tions of the article. A. Miu led the analyses, contributed temporal patterns. Sex Transm Dis. 1999;26:281–285. to the interpretation of the data, and made substantial 28. Sobo EJ. Human immunodeficiency virus seropos- contributions to the Methods and Results sections. J.L. 11. Marks G, Ruiz MS, Richardson JL, et al. Anal in- itivity self-disclosure to sexual partners: a qualitative Chen contributed to the interpretation of the analyses tercourse and disclosure of HIV infection among sero- study. Holist Nurs Pract. 19 95;10:18–28. and made substantial contributions to the Results and positive gay and bisexual men. J Acquir Immune Defic 29.Hays RB, McKusick L, Pollack L, Hilliard R, Hoff Discussion sections. S.C. Morton contributed to the Syndr. 1994;7:866–869. C, Coates TJ. Disclosing HIV seropositivity to signifi- study design (including development of the weighting 12. Marks G, Crepaz N. HIV-positive men’s sexual cant others. AIDS. 19 93;7:425–431. methods) and interpretation of the analyses and made practices in the context of self-disclosure of HIV status. substantial contributions to the Methods section. R. Stall J Acquir Immune Defic Syndr. 2001;27:79–85. 30. Gielen AC, O’Campo P, Faden RR, Eke A. contributed to the study design and interpretation of Women’s disclosure of HIV status: experiences of mis- 13. Gielen AC, Fogarty L, O’Campo P, Anderson J, the analyses and made substantial contributions to the treatment and violence in an urban setting. Women Keller J, Faden R. Women living with HIV: disclosure, Introduction, Results, and Discussion sections. Health. 19 97;25:19–31. violence, and . J Urban Health. 2000;77: 480–491. 31. Sowell R, Seals B, Moneyham L, Guillory J, Acknowledgments 14 . Moneyham L, Seals B, Demi A, Sowell R, Cohen Mizuno Y. Experiences of violence in HIV-seropositive This study was supported by the National Institute of L, Guillory J. Perceptions of stigma in women infected women in the south-eastern United States of America. Child Health and Human Development (grant 1 R01 with HIV. AIDS Patient Care STDs. 1996;10:162–167. JAdv Nurs. 1999;30:606–615. HD35040) and by the Agency for Healthcare Research 15. Kalichman SC, Nachimson D. Self-efficacy and dis- 32. Zierler S, Cunningham W, Andersen R, et al. Vio- and Quality (cooperative agreement U-01HS08578). closure of HIV-positive serostatus to sex partners. lence victimization after HIV infection in a US proba- We gratefully acknowledge the support of Martin F. Health Psychol. 1999;18:281–287. bility sample of adult patients in primary care. Am J Shapiro, Samuel A. Bozzette, and the HIV Cost and Public Health. 2000;90:208–215. Services Utilization Study consortium. 16.Shapiro MF, Berk ML, Berry SH, et al. National We thank those who participated in this study and probability samples in studies of low-prevalence dis- 33. Turner CF, Ku L, Rogers SM, Lindberg LD, Pleck those who helped in the study, including Sandra Berry eases, I: perspectives and lessons from the HIV cost JH, Sonenstein FL. Adolescent sexual behavior, drug for her contribution to project design, questionnaire and services utilization study. Health Serv Res. 1999; use, and violence: increased reporting with computer construction, and data collection; Shirley Nederend for 34:951–968. survey technology. Science. 1998;280:867–873. help with instrument design; and Robin Beckman for 17.Frankel MR, Shapiro MF, Duan N, et al. National file preparation. probability samples in studies of low-prevalence dis- eases, II: designing and implementing the HIV cost and Human Participant Protection services utilization study sample. Health Serv Res. The study protocol was approved by the RAND and 1999;34:969–992. local health care provider institutional review boards. 18. Duan N, McCaffrey DR, Frankel M, et al. HCSUS

954 | Research and Practice | Peer Reviewed | Ciccarone et al. American Journal of Public Health | June 2003, Vol 93, No. 6