Historically, prevention efforts have primarily focused on HIV-seronegative or untested people. Working effectively with people living with HIV (PLWH) has become increasingly important in the era of expanded treatment access [1–7]. Without prevention to help PLWH to adopt and maintain healthier and safer behaviours, the growing number of people living longer with HIV forms a potential source of infection [1,2] and may facilitate the evolution and spread of drug resistance [8,9]. While many PLWH eliminate or reduce behaviours that may expose others to HIV [10–14], a considerable percentage (ranging from 10 to 60% depending on type of behaviours, recall period, and partner's serostatus) do not consistently practice safer behaviours [15–23], thus placing others at risk for HIV infection and themselves at risk for sexually transmitted infections (STI) and possible superinfection with other strains of HIV . Although interventions designed primarily for uninfected persons have been shown to reduce self-reported HIV risk behaviours [25–28], it is not known whether interventions for PLWH are similarly efficacious.
This systematic review synthesizes the available literature on prevention interventions for reducing risky sex and needle-sharing behaviours in PLWH. The goal was to locate and describe controlled trials that rigorously evaluated the effects of interventions for PLWH. The present study expands the scope of earlier qualitative reviews [5,29,30] by conducting meta-analyses to assess quantitatively the overall efficacy of interventions in reducing HIV risk behaviours of PLWH and to identify intervention characteristics associated with efficacy.
Database and search strategy
As part of the HIV/AIDS Prevention Research Synthesis project at the US Centers for Disease Control and Prevention (CDC), five search strategies were implemented to identify published or unpublished interventions reported in English or non-English languages. First, a comprehensive search of electronic bibliographic databases was conducted including AIDSLINE (1988 to discontinuation in December 2000), EMBASE, Medline, PsycINFO, CINHAL and SocioFile from 1988 through 2004. Standardized search terms were cross-referenced (keywords and medical subject heading terms) reflecting four constructs: (a) HIV, AIDS, STI, (b) prevention intervention evaluation, (c) target population (i.e., PLWH) and (d) behavioural or biological outcome. Second, a hand search was conducted of 30 key journals from June 2002 to December 2004, using the same constructs as in the electronic database searches. Third, searches were made in on-line trial registry databases (i.e., Cochrane Controlled Trials Register, EPPI-Centre Education Database, Current Controlled Trials Register, and the US National Institutes of Health CRISP database) and the websites of international organizations that fund HIV/AIDS prevention programs worldwide (i.e., World Health Organization, UNAIDS, USAID, Family Health International, Kaiser Family Foundation). Fourth, HIV researchers were contacted to obtain unpublished manuscripts and upcoming publications. Finally, the references of all pertinent reports were reviewed for additional citations.
Interventions were included if they met all of the following criteria: (a) HIV/AIDS or STI behavioural interventions specifically designed for PLWH; (b) data collected on at least one HIV-related behavioural outcome (i.e., unprotected insertive or receptive anal intercourse, unprotected vaginal intercourse, condom use, needle or works sharing) or biological outcome (i.e., acquisition of STI or hepatitis B or C); (c) at least one follow-up assessment at 3 months (or longer) after the completion of the intervention, a practice recommended for demonstrating sustainable effects of an intervention ; (d) randomized controlled trials (RCT)  or controlled trials with designs that minimized systematic bias associated with non-randomization [33–35] (trials were excluded if they explicitly allowed participants to self-select into the intervention or if they provided intervention to only one group and assessed effect on the basis of behavioural change of the participants from before to after the intervention); and (e) data were necessary and sufficient for calculation of effect sizes.
Authors were contacted to obtain additional information before trials were excluded [36,37].
Information from eligible reports was independently abstracted by pairs of trained reviewers. Linkages among reports were identified to ensure that multiple reports describing an intervention were included in the coding. Using standardized coding forms, each intervention was coded for trial information (e.g., intervention dates, city/country), participant characteristics (e.g., sexual orientation, injection drug use, disease stage, age, gender, race/ethnicity), intervention features (e.g., theory, content, delivery method, duration, setting), and outcomes (e.g., type of behaviour, behaviour recall period, follow-up time). The methodological quality was assessed by coding: assignment method, type of control group, participation rate, overall and differential retention, and intent-to-treat . There was 91% agreement between reviewers across all variables. Discrepancies were reconciled through discussion.
Effect size calculation
Effect sizes were estimated with odds ratios (OR) because the majority of the trials compared two groups on a dichotomous outcome. For trials reporting means and SD values on continuous outcomes, standardized mean differences were calculated to be converted into OR values . An OR < 1 indicates a greater reduction in odds of a risky behaviour in the intervention group relative to the comparison group.
Standard meta-analytical methods [38–40] were used. The natural logarithm was used to obtain log OR and its corresponding weight (i.e., inverse variance) was calculated for each trial. To estimate the overall effect size, each log OR value was multiplied by its weight; the weighted log OR values were summed across trials and then divided by the sum of the weights. The magnitude of heterogeneity of the effect sizes was tested using the Q statistic. The final aggregation was based on a random-effects model, which provides a more conservative estimate of variance and generates more accurate inferences about a population of trials beyond the set of trials included in this review . The aggregated log OR was then converted back to OR by exponential function and a 95% confidence interval (CI) was derived.
The following rules were used to guide effect size abstraction for estimating the overall intervention effect.
Multiple intervention arms. To meet the independence of the effect size assumption, the contrast between the condition expected to have the greatest (e.g., enhanced intervention) and the least (e.g., standard of care) influence in producing intervention effects was selected, usually hypothesized by the author. If no hypothesis was indicated, the first group discussed in detail in the report was used as the intervention condition for analysis.
Multiple HIV risk behaviours. Separate analyses were conducted for sexual behaviours, needle sharing and biological outcomes. If multiple sex behaviours were reported, the behaviour representing the greatest risk of HIV transmission was selected (e.g., unprotected insertive anal intercourse). If a trial reported sexual behaviour data for two or more types of partner, the analysis focused on sex with at-risk partners (e.g., HIV-seronegative or serostatus-unknown partners) rather than HIV-seropositive or all partners.
Multiple follow-ups. In order to evaluate the long-term intervention effects on behavioural change, data collected at the longest follow-up were selected for calculating the overall effect size estimate.
Although the need for adjustment is much less in RCTs than non-RCTs, an adjusted analysis is more conservative, especially when one or more prognostic variables (e.g., baseline risk behaviours) may impact outcomes of interest [39,42]. Data were used from adjusted models reported by the authors for effect size calculation because baseline differences of potential confounding variables are typically controlled in these models. Otherwise, effect sizes were calculated for the follow-up outcome data by adjusting for baseline differences.
Sensitivity analyses were conducted to determine the robustness of intervention effects by assessing whether the overall results were sensitive to the aforementioned decision rules to guide effect size calculation. The overall effect size was recalculated using all available contrasts between study arms, different sexual behaviour outcomes (type of sex, type of partner) and data without adjustment for baseline behaviour. In addition, the combined effect size estimate among all trials was compared with the estimate obtained after excluding a trial (or set of trials) that might influence the overall estimate.
Stratified analyses, via random-effects models, determined whether methodological quality or particular characteristics of the interventions were associated with efficacy. Publication bias that may favour trials with significant findings was ascertained by inspection of a funnel plot of standard error estimates versus effect size estimates from individual samples  and also by a linear regression test . The results of both tests suggested no evidence of publication bias (not shown).
Twelve controlled trials that met the inclusion criteria (Fig. 1) are summarized in Table 1. All 12 trials were conducted in the United States and most trials were carried out after 1996 when HAART became available. One trial specifically targeted HIV-seropositive women only , while the remaining 11 trials consisted of 100% or majority male participants. Four trials [45–48] included samples that were more than two-thirds men who have sex with men and three trials [49–51] specifically targeted injection drug users or substance abusers. The median age across all study samples was 35 years (range, 20.7–41). Among the trials reporting participants' health and medical status, the median length of time diagnosed with HIV ranged from 2 to 6 years [44–47,52,53]; the percentage of AIDS cases ranged from 9 to 50% [46,48–50,53], and the percentage on HIV treatment ranged from 50 to 100% [47–50,54].
Ten trials [44–51,54,55] were RCTs while two trials [52,53] were non-RCTs with minimal assignment bias. The median participation rate was 83.5% of eligible persons (range, 38–100) and the median retention rate was 72% (range, 50–90) at the longest follow-up assessment (median, 6.5 months; range, 3–12). All trials used intent-to-treat for data analysis.
Regarding intervention characteristics, most trials were guided by behavioural theories including social cognitive theory/social learning theory , cognitive–behavioural coping , information–motivation–behavioural skills model  and theory of planned behaviour . All interventions provided information to increase HIV knowledge and motivated participants to eliminate or reduce risky drug use and/or sexual practices. Nine interventions [44–46,48–50,53–55] provided skill building through multiple activities, including live demonstrations, role plays or practice. The skills included technical (e.g., correct use of condoms or needle cleaning), personal (e.g., coping, problem solving) or interpersonal skills (e.g., communication about safer sex, harm reduction negotiation, disclosing one's seropositive status).
The interventions were commonly delivered by health-care providers/counsellors (75%) or peers (41%), delivered to small groups (67%) with 5–10 persons, and in settings where HIV-seropositive individuals receive services (e.g., outpatient clinics, community service centres, 58%). Two trials [47,51] provided on-going intervention activities to the participants (e.g., every time participants visited the clinic). For the remaining 10 trials, the number of intervention sessions ranged from 3 to 48 sessions; the amount of time to deliver the intervention ranged from 4.5 to 48 h; and, the total time span of the intervention ranged from 2 weeks to 1 year.
Overall efficacy of interventions
The overall weighted effect size estimate based on the 12 interventions indicated a significant reduction in odds of unprotected sexual behaviour assessed at the longest follow-up in the intervention group relative to the comparison group (OR, 0.57; 95% CI, 0.40–0.82; n = 2719). Although Fig. 2 shows the heterogeneity of trials (Q11 = 35.25; P < 0.001), sensitivity tests did not reveal any single trial that exerted influence on the overall effect size. The overall effect size estimate was robust as it remained significant when using data from multiple intervention contrasts, different sex outcomes, or without adjustment for baseline behaviour.
For the STI outcome (acquisition of chlamydia or gonorrhoea) [44,48], the effect size estimate was significant (OR, 0.20, 95% CI, 0.05–0.73; n = 1,177). For the needle-sharing outcome [49,51,52], the reduction was relatively large but the effect size estimate was not statistically significant (OR, 0.47; 95% CI, 0.13–1.71; n = 403). Stratified analyses were not conducted for these two outcomes because of the small number of trials.
Stratified analysis of trials reporting unprotected sex behaviours
Stratified analyses were implemented to assess efficacy based on trial and intervention characteristics (Table 2). With regarding to methodological quality, a significant effect was observed for trials that were RCT, had > 80% participation rate or had a comparison group that did not receive HIV behavioural intervention. Intervention effects were not affected by the retention rate. The significant intervention effect was also seen in the eight trials with the most rigorous methodological quality (i.e., RCTs with ≥ 70% retention and intent-to-treat: OR, 0.66; 95% CI, 0.45–0.97).
Interventions with the following characteristics were found to reduce unprotected sex significantly:
* guided by behavioural theory
* specifically focused on HIV transmission behaviours (more than two-thirds of sessions)
* provided skills building, such as demonstrating correct condom use, practicing coping or problem-solving skills, or role-playing safer sex communication with partners
* delivered to individuals on a one-to-one basis
* delivered by health-care providers or professional counsellors
* delivered in settings where people living with HIV receive services
* delivered in an intensive manner (> 10 intervention sessions, > 20 h)
* delivered over a longer duration (≥ 3 months)
* addressed a myriad of issues related to coping with one's serostatus, medication adherence, and HIV risk behaviours.
Three interventions [49,50,53] that included all the aforementioned intervention features, except individual-level delivery, showed a highly significant reduction in the OR unprotected sexual behaviour (OR, 0.25; 95% CI, 0.13–0.47). In addition, interventions delivered to groups or delivered by peers were marginally significant.
The length and number of follow-ups varied across trials, allowing analysis of whether the reduction in the OR of unprotected sex was consistent at different follow-up time points. Intervention effects were estimated at two follow-up intervals: 3–4 months and 6–12 months. Significant intervention effects were found at both times [3–4 months: OR, 0.55; 95% CI, 0.32–0.95 (seven interventions); 6–12 months: OR, 0.61; 95% CI, 0.43–0.87 (eight interventions)].
Our meta-analysis of 12 controlled trials showed that interventions, as a whole, successfully reduce self-reported unprotected sex among PLWH. This significant intervention effect is robust as it was not affected by the rules used to guide our meta-analyses. The reduced rates of unprotected sex were observed not only at 3–4 months but also at 6–12 months post-intervention. Most importantly, the overall intervention effect observed here for PLWH (OR, 0.57) was comparable to, or slightly stronger in magnitude, than the intervention effects observed in the literature for primarily uninfected or serostatus unknown men who have sex with men (OR, 0.77) , drug users (OR, 0.86) , sexually active youth (OR, 0.66)  or heterosexual adults (OR, 0.81) . When extrapolated to a population with 30% (range, 10–60) prevalence of sexual risk behaviour [15,16,18], we estimated a 43% (range, 21–63) relative risk reduction in unprotected sex. Risk reduction of this magnitude is well within the range considered to be cost-effective when translated into final health outcomes [60,61]. It is also encouraging to see evidence of efficacy in reducing chlamydia or gonorrhoea acquisition. A relatively large but non-significant intervention effect was observed for needle sharing. The findings on STI and needle sharing were based on a small set of trials and, therefore, the robustness of these findings needs to be reassessed when additional controlled trials are completed.
Aside from overall efficacy, intervention characteristics associated with efficacy were identified and summarized (see list on page 152). Interestingly, while interventions with a greater number of sessions over a longer duration and time span were associated with significant risk reduction, two interventions [47,51] with on-going activities failed to demonstrate efficacy. It is possible that the combination of the characteristics listed influenced intervention success as the three most intensive interventions, which were overall efficacious, were also delivered by professional counsellors in service settings, provided skills building and addressed a myriad of issues related to PLWH. Although the independent contributions of these characteristics cannot be disentangled within these data, the findings, taken as a whole, suggest that integrating prevention into settings where PLWH receive medical care or other services and addressing the health, behaviour and well-being of PLWH may be important for achieving success [15,48,62].
The findings of our meta-analytic review must be viewed within the context of the limitations of the available evidence and point to future research needs. Although not statistically significant, the magnitude of the intervention effect estimates for the subgroups of PLWH (i.e., men who have sex with men, injection drug users/substance abusers, other population) are comparable to the magnitude of the overall estimate in this review and in previous reviews [25–28]. The lack of statistical significance may be because of the small number of trials conducted within each subgroup. While subgroups of PLWH may share common experience in living with HIV, each subgroup may face unique challenges. For example, HIV-seropositive heterosexual women may wish to become pregnant and may struggle with the care of young children, while substance-using men who have sex with men must deal with homophobia and drug addiction. Consequently, it is important to examine further whether and to what extent the findings are applicable across different subgroups of PLWH when data from other controlled trials become available.
Several trials did not report important variables (e.g., medical status of participants, serostatus of partner, number of partners with whom unprotected intercourse occurred, cost data), which could have provided additional insights into the intervention effects, applicability of interventions, connection between unprotected sex and STI transmission and cost-based recommendations. Clear and transparent reporting of these key elements in intervention trials would greatly improve the quality of future meta-analytic work [63–65].
The majority of the trials were unblinded and relied on self-reported sexual behaviour, which may be open to socially desirable responding . Acknowledging this possible bias, many trials attempted to ensure the confidentiality of data, including asking participants to answer questionnaires without the presence of an interviewer. In addition, all the trials had a comparison group and the assignment method was either randomization or an unbiased method, which reduced the likelihood that individual characteristics (e.g., impression management) influenced the intervention effect. Given that blinded trials are not feasible in HIV behavioural prevention research, future intervention trials should consider complementing self-reported behavioural measures with biological endpoints to assess intervention efficacy . It should be noted, however, that an STI may be acquired from seroconcordant or serodiscordant partners and are not a specific measure of reduced HIV transmission risk to uninfected partners . New technologies (e.g., computer-assisted assessments) that are likely to improve the internal consistency of self-reported sexual behaviour and increase the reporting of stigmatized behaviours [68–70] are currently available and should also be widely utilized.
Few interventions in this review incorporated contextual, societal or structural factors. While addressing psychological processes, behavioural skills and communication within a relationship is important, socioecological models that identify multiple determinants of behaviours, including dyad/family, social/community (e.g., networks), structural and societal/cultural factors (e.g., stigma, social norms, social and economic status) can also be essential because individual behavioural change does not occur in a vacuum [15,29]. Future research should consider comprehensive efforts that intervene on HIV risk behaviours at the individual, group, structural and societal levels. In addition, it is also important to incorporate newly identified predictors of risky behaviours, including HIV treatment optimism , use of the Internet to meet sex partners [72,73] and serosorting  into the next generation of prevention interventions for PLWH.
Our findings showed that significant intervention effects were observed at 3–4 months and 6–12 months post-intervention. However, it is not known whether the intervention effect sustains beyond 12 months after the intervention. The on-going NIMH Healthy Living Project , which follows participants for 25 months, may provide some insights into longer-term intervention effects. Adopting and maintaining healthier and safer behaviours over a lifetime can be challenging for some PLWH and may require on-going behavioural reinforcement to prevent relapse. In addition, given that not all PLWH require on-going prevention programmes [15,76], it may be more cost-effective to identify specific groups of individuals who require on-going behavioural reinforcement and determine the types of intervention that work best for them. It is likely that those who are struggling with multiple health and social problems, such as substance abuse, severe and persistent mental illness, relationship abuse or poverty, may benefit most from intensive prevention interventions . Further research is recommended to examine the efficacy of interventions designed for PLWH who are struggling with multiple problems that increase their risk of transmitting HIV to others.
The generalizability of our meta-analytic findings warrants comment. All 12 controlled trials evaluated in this review were conducted in the United States. Although we identified two published reports [37,78] and a few on-going prevention programmes with seropositive persons outside the United States, none of these trials met our stringent inclusion criteria, which had been designed to derive valid conclusions about intervention efficacy. This research gap is not surprising as many countries allocate the limited resources that are available to providing medical treatment for those who are in need of care, preventing mother-to-child transmission and identifying individuals with undiagnosed HIV to reduce the HIV infection rate, rather than for conducting controlled trials to evaluate behavioural interventions specifically designed for PLWH [3,79]. However, the US funding agencies have increased support for developing and evaluating interventions to help PLWH to reduce HIV transmission risk behaviours because the importance of primary prevention with PLWH has been recognized and given priority as a key component of the CDC's national HIV prevention agenda [1,2,5]. Although it is unclear the extent to which our meta-analytic findings (based on the experience in the United States) can be generalized to resource-poor settings and other populations, the lessons learned may provide insights into which strategies have the best chance of success in the global prevention efforts. As antiretroviral therapy programmes are expanded worldwide [3,8,79], effective prevention strategies should be integrated within routine medical care and services provided for PLWH.
Moving evidence-based research for people living with HIV into practice is an important step in making a greater impact on the HIV epidemic. As the effective strategies are woven into prevention efforts, both in the United States and internationally, the focus must also include assessing the effectiveness and the deployment of these strategies in real-world settings. Continually evaluating and maximizing the effectiveness of these activities in the field is critical for a sustainable impact on the HIV epidemic.
In conclusion, the behavioural interventions included in this review, as a whole, are efficacious in reducing HIV sexual risk taking across a wide range of PLWH. When selecting interventions for PLWH, prevention providers should consider those with the following characteristics: integrating theory-based prevention within routine medical care and services; addressing aspects of mental health and medical adherence in addition to HIV risk behaviour; and providing PLWH with the necessary skills for successful risk reduction. Given that the potential of an intervention delivered to a single HIV-seropositive person may prevent multiple HIV infections, prevention with PLWH offers an unprecedented opportunity for significantly reducing HIV transmission.
Other members of the HIV/AIDS Prevention Research Synthesis Team who contributed to this review are (listed alphabetically): Julia DeLuca, Tanesha Griffin, Linda Kay, Angela Kim, Mary Mullins, R. Tom Sherba and Sekhar Thadiparthi. We would like to thank Thomas Patterson, James Sorensen and Gina Wingood for providing additional analyses for our review; and Gary Marks, Markus Crepaz, Ann O'Leary and Linda Koenig for providing helpful comments on an earlier version of this paper.
Sponsorship: This work was supported by the Prevention Research Branch, Division of HIV/AIDS Prevention, US Centers for Disease Control and Prevention and was not funded by any other organization.
Note: The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
1. Centers for Disease Control and Prevention. Advancing HIV Prevention: Interim Technical Guidance for Selected Interventions
. Atlanta, GA: US Department of Health and Human Services; 2003:1–75.
2. Centers for Disease Control and Prevention. Advancing HIV prevention: new strategies for a changing epidemic, United States, 2003. MMWR Morb Mortal Wkly Rep
3. Global HIV Prevention Working Group. HIV Prevention in the Era of Expanded Treatment
. Washington, DC: Kaiser Family Foundation; 2004.
4. Gordon CM, Stall R, Cheever LW. Prevention interventions with persons living with HIV/AIDS: challenges, progress, and research priorities. J Acquir Immune Defic Syndr 2004; 37:S53–S57.
5. Janssen RS, Valdiserri RO. HIV prevention in the United States: increasing emphasis on working with those living with HIV. J Acquir Immune Defic Syndr 2004; 37:S119–S121.
6. Kalichman SC. Positive Prevention: Reducing HIV Transmission Among People Living with HIV/AIDS. New York: Plenum; 2004.
7. Consensus Development Conference Statement. Interventions to Prevent HIV Risk Behaviors
. Bethesda, MD: National Institutes of Health; 1997:1–41.
8. World Health Organization. Scaling up Antiretroviral Therapy in Resource-limited Settings: Treatment Guidelines for a Pubic Health Approach
. Geneva: World Health Organization; 2004:1–67.
9. Valdiserri RO. International scale-up for antiretroviral treatment: where does prevention fit? J Acquir Immune Defic Syndr 2004; 37:S138–S141.
10. Marks G, Crepaz N, Senterfitt JW, Janssen RS. Meta-analysis of high-risk sexual behavior in persons aware and unaware they are infected with HIV in the United States: implications for HIV prevention programs. J Acquir Immune Defic Syndr
2005; in press.
11. Weinhardt LS, Carey MP, Johnson BT, Bickham NL. Effects of HIV counseling and testing on sexual risk behavior: a meta-analytic review of published research, 1985–1997. Am J Public Health 1999; 89:1397–1405.
12. Weinhardt LS. HIV diagnosis and risk behavior. In: Positive Prevention
. Edited by Kalichman SC. New York: Plenum; 2004:29–64.
13. Wolitski RJ, MacGowan RJ, Higgins DL, Jorgensen CM. HIV counseling and testing on risk-related practices and help-seeking behavior. AIDS Educ Prev 1997; 9:52–67.
14. Higgins DL, Galavotti C, O'Reilly KR, Schnell DJ, Moore M, Rugg DL, et al
. Evidence for the effects of HIV antibody testing on risk behaviors. JAMA 1991; 266:2419–2429.
15. Crepaz N, Marks G. Towards an understanding of sexual risk behavior in people living with HIV: a review of social, psychological, and medical findings. AIDS 2002; 16:135–149.
16. Kalichman SC. HIV transmission risk behaviors of men and women living with HIV-AIDS: prevalence, predictors, and emerging clinical interventions. Clin Psychol-Sci Pract 2000; 7:32–47.
17. Kok G. Targeted prevention for people with HIV/AIDS: feasible and desirable? Patient Educ Couns 1999; 36:239–246.
18. Marks G, Burris S, Peterman TA. Reducing sexual transmission of HIV from those who know they are infected: the need for personal and collective responsibility. AIDS 1999; 13:297–306.
19. McGowan JP, Shah SS, Ganea CE, Blum S, Ernst JA, Irwin KL, et al
. Risk behavior for transmission of human immunodeficiency virus (HIV) among HIV-seropositive individuals in an urban setting. Clin Infect Dis 2004; 38:122–127.
20. Centers for Disease Control and Prevention. High-risk sexual behavior by HIV-positive men who have sex with men: 16 sites, United States, 2000–2002.
MMWR Morbid Mortal Wkly Rep
21. Schiltz MA, Sandfort TG. HIV-positive people, risk and sexual behaviour. Soc Sci Med 2000; 50:1571–1588.
22. Elford J, Bucher HC, Rawstorne P, Fogarty A, van de Ven P, Kippax S, et al
. International perspectives.
In: Positive Prevention: Reducing HIV transmission among people living with HIV/AIDS
. Edited by Kalichman SC. New York: Plenum; 2004:245–278.
23. Parsons J T SE, Wolitski R J, Halkitis PN, Purcell DW, Hoff C, Gómez CA, for the SUMIT Study Group. Sexual harm reduction practices of HIV-seropositive gay and bisexual men: serosorting, strategic positioning, and withdrawal before ejaculation. AIDS
24. Smith DM, Richman DD, Little SJ. HIV superinfection. J Infect Dis 2005; 192:438–444.
25. Herbst JH, Sherba RT, Crepaz N, Britton J, Zohrabyan L, Stall RD, et al
. A meta-analytic review of HIV behavioral interventions for reducing sexual risk behavior of men who have sex with men (MSM). J Acquir Immune Defic Syndr
2005; in press.
26. Mullen PD, Ramirez G, Strouse D, Hedges LV, Sogolow E. Meta-analysis of the effects of behavioral HIV prevention interventions on the sexual risk behavior of sexually experienced adolescents in controlled studies in the United States. J Acquir Immune Defic Syndr 2002; 30:S94–S105.
27. Neumann MS, Johnson WD, Semaan S, Flores SA, Peersman G, Hedges LV, et al
. Review and meta-analysis of HIV prevention intervention research for heterosexual adult populations in the United States. J Acquir Immune Defic Syndr 2002; 30:S106–S117.
28. Semaan S, Des Jarlais DC, Sogolow E, Johnson WD, Hedges LV, Ramirez G, et al
. A meta-analysis of the effect of HIV prevention interventions on the sex behaviors of drug users in the United States. J Acquir Immune Defic Syndr 2002; 30:S73–S93.
29. Gordon CM, Forsyth AD, Stall R, Cheever LW. Prevention interventions with persons living with HIV/AIDS: state of the science and future direction. AIDS Educ Prev 2005; 17:6–20.
30. Wolitski RJ, Janssen RS, Onorato IM, Purcell DW, Crepaz N. An overview of prevention with people living with HIV.
In: Positive Prevention: Reducing HIV Transmission Among People Living with HIV/AIDS
. Edited by Kalichman SC. New York: Plenum; 2004:1–28.
31. Flores SA, Crepaz N. Quality of study methods in individual- and group-level HIV intervention research: critical reporting elements. AIDS Educ Prev 2004; 16:341–352.
32. Alderson P, Green S, Higgins JPT (eds). Cochrane Reviewers' Handbook 4.2.2
. [updated March, 2004]. Chichester, UK: Cochrane Library; 2004.
33. Victora CG, Habicht JP, Bryce J. Evidence-based public health: moving beyond randomized trials. Am J Public Health 2004; 94:400–405.
34. Cook T, Campbell D. Quasi-experimentation: Design and Analysis Issues for Field Settings. Boston, MA: Houghton Mifflin; 1979.
35. Norris SL, Atkins D. Challenges in using nonrandomized studies in systematic reviews of treatment interventions. Ann Intern Med 2005; 142:1112–1119.
36. Fogarty LA, Heilig CM, Armstrong, Cabral R, Galavotti C, Gielen AC, et al
. Long-term effectiveness of a peer-based intervention to promote condom and contraceptive use among HIV-positive and at-risk women. Public Health Rep 2001; 116:103–119.
37. MacNeil JM, Mberesero F, Kilonzo G. Is care and support associated with preventive behaviour among people with HIV? AIDS Care 1999; 11:537–546.
38. Lipsey MW, Wilson DB. Practical Meta-analysis. Thousand Oaks: Sage; 2001.
39. Cooper H, Hedges LV. The Handbook of Research Synthesis. New York: Russell Sage Foundation; 1994.
40. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. Methods for Meta-analysis in Medical Research. Chichester, UK: Wiley; 2000.
41. Hedges LV, Vevea JL. Fixed and random effects models in meta-analysis. Psychol Meth 1998; 3:486–504.
42. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al
. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001; 134:663–694.
43. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a sample, graphical test. Br Med J 1997; 315:629–634.
44. Wingood GM, DiClemente RJ, Mikhail I, Lang DL, McCree DH, Davies SL, et al
. A randomized controlled trial to reduce HIV transmission risk behaviors and sexually transmitted diseases among women living with HIV: The WILLOW Program. J Acquir Immune Defic Syndr 2004; 37:S58–S67.
45. Kelly JA, Murphy DA, Bahr GR, Kalichman SC, Morgan MG, Stevenson Y, et al
. Outcome of cognitive-behavioral and support group brief therapies for depressed. HIV-infected persons. Am J Psychiatry 1993; 150:1679–1686.
46. Patterson TL, Shaw WS, Semple SJ. Reducing the sexual risk behaviors of HIV+ individuals: outcome of a randomized controlled trial. Ann Behav Med 2003; 25:137–145.
47. Richardson JL, Milam J, McCutchan A, Stoyanoff S, Bolan R, Weiss J, et al
. Effect of brief safer-sex counseling by medical providers to HIV-1 seropositive patients: a multi-clinic assessment. AIDS 2004; 18:1179–1186.
48. Wolitski RJ, Gómez CA, Parsons JT, for the SUMIT Study Team. Effects of a peer-led behavioral intervention to reduce HIV transmission and promote serostatus disclosure among HIV-seropositive gay and bisexual men. AIDS 2005; 19:S99–S110.
49. Margolin A, Avants SK, Warburton LA, Hawkins KA, Shi J. A randomized clinical trial of a manual-guided risk reduction intervention for HIV-positive injection drug users. Health Psychol 2003; 22:223–228.
50. Rotheram-Borus MJ, Swendeman D, Comulada S, Weiss RE, Lightfoot M. Prevention for substance using HIV positive young people: telephone and in-person delivery. J Acquir Immune Defic Syndr 2004; 37:S68–S77.
51. Sorensen JL, Dilley J, London J, Okin RL, Delucchi KL, Phibbs CS. Case management for substance abusers with HIV/AIDS: a randomized controlled trial. Am J Drug Alcohol Abuse 2003; 29:133–150.
52. Grinstead O, Zack B, Faigeles B. Reducing postrelease risk behavior among HIV seropositive prison inmates: the health promotion program. AIDS Educ Prev 2001; 13:109–119.
53. Rotheram-Borus MJ, Lee MB, Murphy DA, Futterman D, Duan N, Birnbaum JM, et al
. Efficacy of a preventive intervention for youths living with HIV. Am J Public Health 2001; 91:400–405.
54. Kalichman SC, Rompa D, Cage M, DiFonzo K, Simpson D, Austin J, et al
. Effectiveness of an intervention to reduce HIV transmission of risks in HIV-positive people. Am J Prev Med 2001; 21:84–92.
55. Cleary PD, van Devanter N, Steilen M, Stuart A, Shipton-Levy R, McMullen W, et al
. A randomized trial of an education and support program for HIV-infected individuals. AIDS 1995; 9:1271–1278.
56. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall; 1986.
57. Beck AT. Cognitive Therapy and the Emotional Disorders. New York: International University Press; 1976.
58. Fisher JD, Fisher WA, Williams SS, Malloy TE. Empirical tests of an information–motivation–behavioral skills model of AIDS-preventive behavior with gay men and heterosexual university students. Health Psychol 1994; 13:238–250.
59. Fishbein M, Ajzen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975.
60. Holtgrave DR. Estimating the effectiveness and efficiency of US HIV prevention efforts using scenario and cost-effectiveness analysis. AIDS 2002; 16:2347–2349.
61. Kahn JG, Sanstad KC. The role of cost-effectiveness analysis in assessing HIV-prevention interventions. AIDS Public Policy J 1997; 12:21–30.
62. Centers for Disease Control and Prevention. Incorporating HIV prevention into the medical care of persons living with HIV: recommendations of CDC, the Health Resources and Services Administration, the National Institutes of Health, and the HIV Medicine Association of the Infectious Disease Society of America. MMWR Morb Mortal Wkly Rep
63. Moher D, Schulz KF, Altman D, for the CONSORT Group (Consolidated Standards of Reporting Trials). The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA 2001; 285:1987–1991.
64. Des Jarlais DC, Lyles C, Crepaz N, for the TREND Group. Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: the TREND statement. Am J Public Health 2004; 94:361–366.
65. Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers of economic submissions to the Br Med J
. The Br Med J Economic Evaluation Working Party. Br Med J 1996; 313:275–283.
66. Weinhardt LS, Forsyth AD, Carey MP, Jaworski BC, Durant LE. Reliability and validity of self-report measures of HIV-related sexual behavior: progress since 1990 and recommendations for research and practice. Arch Sex Behav 1998; 27:155–180.
67. Fishbein M, Pequegnat W. Evaluating AIDS prevention interventions using behavioral and biological outcome measures. Sex Transm Dis 2000; 27:101–110.
68. Newman JC, Des Jarlais DC, Turner CF, Gribble J, Cooley P, Paone D. The differential effects of face-to-face and computer interview modes. Am J Public Health 2002; 92:294–297.
69. Kissinger P, Rice J, Farley T, Trim S, Jewitt K, Margavio V, et al
. Application of computer-assisted interviews to sexual behavior research. Am J Epidemiol 1999; 149:950–954.
70. Johnson AM, Copas AJ, Erens B, Mandalia S, Fenton K, Korovessis C, et al
. Effect of computer-assisted self-interviews on reporting of sexual HIV risk behaviours in a general population sample: a methodological experiment. AIDS 2002; 15:111–115.
71. Crepaz N, Hart TA, Marks G. Highly active antiretroviral therapy and sexual risk behavior: a meta-analytic review. JAMA 2004; 292:224–236.
72. Centers for Disease Control and Prevention. Internet use and early syphilis infection among men who have sex with men - San Franciso, California, 1999–2003. MMWR Morb Mortal Wkly Rep
73. Elford J, Bolding G, Davis M, Sherr L, Hart G. Web-based behavioral surveillance among men who have sex with men: a comparison of online and offline samples in London, UK. J Acquir Immune Defic Syndr 2004; 35:421–426.
74. Truong HM, McFarland W, Kellogg T, Dilley, K. Increases in ‘serosorting’ may prevent further expansion of the HIV epidemic among MSM in San Francisco. 11th Conference on Retroviruses and Opportunistic Infections
. February, San Francisco 2004 [abstract 833].
75. Gore-Felton C, Rotheram-Borus MJ, Weinhardt LS, Kelly JA, Lightfoot M, Kirshenbaum SB, et al
. The Healthy Living Project: an individually tailored, multidimensional intervention for HIV-infected persons. AIDS Educ Prev 2005; 17:21–39.
76. Holtgrave DR. Estimation of annual HIV transmission rates in the United States, 1978–2000. J Acquir Immune Defic Syndr 2004; 35:89–92.
77. Stall R, Mills TC, Williamson J, Hart T, Greenwood G, Paul J, et al
. Association of co-occurring psychosocial health problems and increased vulnerability to HIV/AIDS among urban men who have sex with men. Am J Public Health 2003; 93:939–942.
78. Rao AV, Swaminathan R, Baskaran S, Belinda C, Andal G, Saleem K. Behaviour change in HIV infected subjects following health education. Indian J Med Res 1991; 93:345–349.
79. World Health Organization and UNAIDS. Treating 3 Million by 2005: Making it Happen
. Geneva: World Health Organization; 2003:1–55.