South Africa is currently experiencing one of the most severe human immunodeficiency virus (HIV) epidemics in the world, with 18.8% of the population (5.5 million people) infected with HIV, nearly 1000 AIDS deaths daily, and 1.2 million children orphaned due to AIDS.1-3 There is growing hope, however, that the antiretroviral (ARV) therapy rollout under way in South Africa will help to abate the epidemic by providing increased access to effective treatment.4-6
Providing ARV medications to those who are infected will not be sufficient, however, if people continue to engage in risky sexual behavior. Studies indicate that a substantial number of people in South Africa have difficulty initiating and sustaining safer sexual practices, including people living with HIV/AIDS (PLWHA).3,7-11 The South African ARV rollout presents a unique opportunity to address PLWHA's risk reduction needs by linking HIV prevention interventions with HIV clinical care.5 This linkage between prevention and care provides repeated opportunities for risk reduction interventions to be delivered to PLWHA over the ongoing course of clinical care, in a safe and supportive environment.12-14
Despite the need for the development of HIV risk reduction interventions for South African PLWHA and the efficiency of situating such interventions in clinical settings, little research has focused on developing effective HIV prevention interventions for PLWHA in clinical care in South Africa or elsewhere.15-18 The US-based Options Project is one of the few clinic-based HIV risk reduction interventions with demonstrated effectiveness in reducing sexual risk behavior among PLWHA.16,18 This intervention is based on the information-motivation-behavioral skills (IMB) model of HIV-preventive behavior,19-23 and it uses motivational interviewing techniques24,25 to deliver HIV risk reduction information, motivation, and behavioral skills content to help PLWHA reduce their HIV transmission risk behavior. The intervention consists of collaborative, patient-centered discussions between the health care provider and the patient during routine clinical visits in which the provider assesses the patient's risk behaviors, identifies barriers to consistently practicing safer sex, elicits strategies from the patient for overcoming those barriers, and negotiates an individually tailored HIV risk reduction (or safer sex maintenance) goal with the patient.16,18
The US-based Options Project intervention was delivered by HIV care clinicians to HIV-infected patients in a US inner city clinic setting on an ongoing basis during routine clinical care visits.16,18 This brief risk reduction intervention was consistently implemented with reasonable fidelity,16 and it significantly reduced self-reported unprotected vaginal, anal, and insertive oral sexual events among PLWHA over a follow-up interval of up to 18 months.18
The present study consisted of a pilot randomized controlled trial that explored the feasibility of implementing an adapted version of the Options Project intervention with PLWHA in a clinical care setting in South Africa, the fidelity with which the intervention could be implemented, and the potential effectiveness of the intervention in reducing risky sexual behavior over a 6-month period.
HIV-infected participants were recruited at an urban HIV care clinic in KwaZulu-Natal, South Africa, that provided highly subsidized treatment for PLWHA. Patients paid 130 Rand/month (approximately US $20/mo) for all clinical visits and laboratory tests; ARV medications were provided free of charge. According to the Programmes Administrator at the clinic (Linda Lodge, BAdmin, personal communication, January 2008), those patients who had difficulty paying this monthly fee were assisted in accessing state grants to fund their treatment. No one was refused treatment because they could not afford to pay. The only other option at the time for accessing ARV medications was private facilities that charged 5-6 times more per visit.
Study inclusion criteria were documented HIV infection, receiving HIV care at the clinic, and being 18 years or older. Trained multilingual research assistants described the study to interested patients, indicated that participation was voluntary, and obtained informed consent. Of the 154 HIV-infected patients approached, only 2 (1.3%) declined to participate. Ethics committees at the University of Connecticut, Yale University School of Medicine, and McCord Hospital in Durban, South Africa, approved the study protocol.
This prospective study was a small, randomized clinical trial in which 5 ARV adherence counselors from an HIV care clinic in Durban were randomly assigned to the intervention (3 counselors) or standard-of-care control (2 counselors) condition. Intervention counselors received training in the adapted version of the Options Project intervention (entitled Izindlela Zokuphila/Options for Health); standard-of-care counselors did not. At study enrollment, participants were randomly assigned to one of these counselors (with 2:1 over assignment to the intervention condition). The decision to use counselors to deliver this intervention was made after conducting focus groups and semistructured interviews with clinic staff (n = 21) and HIV-infected patients (n = 32), all of whom indicated that doctors did not have the time to implement the intervention and that patients would be most comfortable talking about sensitive topics, such as sex and HIV disclosure, with counselors. Because counselors already were meeting with HIV-infected patients at each clinic visit to discuss issues related to HIV, the intervention capitalized on existing interactions and fit neatly into the clinic ecology.
Patients in the intervention and standard-of-care control conditions met with their counselors approximately every 3 months during regularly scheduled clinic visits. Participants in the standard-of-care control condition received standard HIV counseling about HIV, ARVs, medication adherence, and nutrition. Standard HIV counseling did not include systematic discussion of HIV prevention, but such discussions were not prohibited and occurred on an ad hoc basis. Patients in the intervention condition received the Options risk reduction intervention in addition to standard HIV counseling.
Outcome assessments were conducted at baseline and 6-month follow-up. Using a structured questionnaire, research assistants assessed participants' demographics and sexual behavior. Clinical data (ie, opportunistic infections, CD4+ cell count, and HIV-1 viral load) were extracted from participants' medical charts by hospital staff. Bilingual interviews were conducted using either isi-Zulu or English questionnaires. Participants were reimbursed 50 Rand to cover their transportation costs each time they completed a questionnaire but not for participating in the intervention sessions. Study participants were informed that their counselors, HIV care providers, and other clinic staff would not have access to the data provided during the interviews.
The Izindlela Zokuphila/Options for Health intervention consisted of brief (15-minute) patient-centered discussions between a counselor and a patient during routine clinical visits, and these discussions were repeated at each visit over an interval of approximately 6 months. The intervention consisted of an 8-step framework used by the counselor to tailor the discussions to a specific patient's HIV risk reduction (or maintenance of safer behavior) needs. The counselor used motivational interviewing techniques24,25 to (1) introduce the discussion of safer sex, (2) assess the patient's risk behavior, (3) determine how important it is to the patient to change his/her risk behavior (or maintain his/her safer behavior), (4) determine how confident the patient is that he/she can change his/her risk behavior (or maintain his/her safer behavior), (5) identify information, motivation, behavioral skills, and other barriers to consistently practicing safer behavior, (6) discuss specific strategies for overcoming these barriers, and (7) negotiate a risk reduction (or safer sex maintenance) goal or action plan with the patient. The final intervention step was to document the agreed upon goal on the “Action Plan” form, which was handed to the patient. Upon completion of the Options for Health visit, the counselor documented what transpired during the discussion (intervention steps completed, risk behaviors identified, agreed upon goal, etc.) on the “Patient Record Form,” which was then stored in the patient's medical file.
Both the intervention and standard-of-care counselors were certified as lay counselors, which means that they had received, at a minimum, 10-15 hours of training in Department of Health-sponsored voluntary counseling and testing. In addition, to work as adherence counselors at the clinic, they had to complete 300 hours of training in adherence counseling provided by the clinic staff, which included a minimal amount of training in HIV risk reduction counseling. Whereas the standard-of-care counselors received no additional training for this study, counselors in the intervention condition were trained to criterion in the Izindlela Zokuphila/Options for Health protocol over a period of 3 days. This included learning culturally appropriate strategies that their HIV-positive patients could use to minimize sexual risk behavior. Before intervention implementation, counselors were observed in roleplay demonstrations until performance criteria were met. In addition, during implementation, counselors were observed “in vivo” for adherence to the protocol. Weekly supervision was provided by a trained social worker who was a member of the clinic staff.
On the baseline and 6-month assessments, participants reported the number of vaginal, anal, and oral sex events they engaged in during the preceding 3 months; the number of times condoms were used for each type of sex; and the number and perceived HIV serostatus of their partners. The primary outcome measure was the total number of unprotected sex events, defined as the combined number of vaginal and anal sexual events in which condoms were not used. Although oral sex was assessed, it was not included in the primary outcome measure because of its low frequency in the present sample (11% of the sample) and its relative inefficiency for HIV transmission.26
An intent-to-treat analysis was used in which analysis of intervention effects was based on the condition to which participants were randomized rather than on actual intervention received. Analyses of variance for continuous measures and χ2 tests for categorical measures were conducted to examine whether there were pretest differences between participants in the intervention versus standard-of-care control conditions and whether there was differential attrition between the conditions.
To assess intervention outcomes, the primary outcome measure of unprotected sexual events (vaginal and anal) was modeled as a function of condition (intervention or control), time (baseline and follow-up), gender, and condition × time interaction. Preliminary analyses supported this approach as individual counselors seemed to have no independent effects on outcomes, and there were no “dosing” effects. Random coefficient regression modeling using HLM627 was utilized to account for the correlated nature of the longitudinal data (ie, repeated observations across subjects)28 and the Poisson distribution which is appropriate for count outcomes.28,29 This type of model yields an outcome variable that is the logarithm of the odds (ηij), which in the primary analyses is the number of unprotected sex events. The predicted probability and event rate ratio (ERR) can then be calculated from model estimates. Using the same approach, we also estimated a model examining the effect of condition, time, gender, and condition × time interaction on the total number of unprotected sexual events with perceived HIV-negative and unknown status partners and a model examining effects on the total number of sexual events (protected and unprotected).
Between October 2004 and August 2005, 152 HIV-infected patients (103 intervention, 49 standard-of-care control) participated in this study. Mean age of the participants was 34 years (range 18-58); 65 (43%) were male; and 139 (91%) were Zulu with the remaining participants of other ethnicities. A total of 67 participants (44.1%) had less than a high school education, 66 (43.4%) participants had completed high school, and 19 (12.5%) reported having education beyond high school. The majority (70%) were unemployed, and over half of them (56.7%) reported not having enough money for food. Only 27 participants (18%) in the study were married.
The majority of the participants (82%) had learned of their diagnosis within the past 3 years, with 43% diagnosed in the past year. The primary mode of HIV acquisition, according to participants, was heterosexual sex (93%), with 11 reporting that they acquired HIV through other means. Most patients (67%) had a history of opportunistic infections: 43% (66) of them had contracted tuberculosis once, 21% (33) had contracted tuberculosis 2 or 3 times, 1 patient had had tuberculosis and pneumocystis pneumonia, and 2 patients had had candidiasis. Mean CD4+ cell count at baseline was 122.24 cells/mm3 (SD = 92.27), which increased to 221.86 cells/mm3 (SD = 143.78) at 6-month follow-up. Over half (66%) of the patients were on ARVs at baseline, which increased to 78% by the 6-month follow-up. At baseline, patients had been on ARVs an average of 185.74 days (SD = 239.48, range 4-1791). At 6-month follow-up, patients were on 1 of 5 regimens: (1) lamivudine/stavudine/efavirenz (83%), (2) Combivir/efavirenz (6%), (3) didanosine/stavudine/efavirenz (5%), (4) lamivudine/stavudine/nevirapine (5%), or (5) Combivir/nevirapine (1%). Of the 8 participants who had HIV-1 viral load data at baseline, the mean viral load was 347,040 copies/mL (SD = 389,019); none had an undetectable viral load (<50 copies/mL). Of the 95 participants who had viral load data at 6-month follow-up, 89.5% were undetectable. Of the 10 participants who had detectable virus, the mean viral load was 371,903 copies/mL (SD = 1,020,215, range 209-3,268,304). Demographic characteristics, CD4+ cell counts at baseline, and history of opportunistic infections are presented in Table 1.
Eighty-four participants (55%; 41 women and 43 men) reported having vaginal or anal sex during the 3 months preceding the baseline or follow-up assessment. Sixteen participants (11%) who reported having no vaginal or anal sex at baseline became sexually active (vaginally or anally) at follow-up, and 6 participants (4%) who reported having vaginal or anal sex at baseline reported none at follow-up. Of the participants reporting vaginal or anal sex, 23 (27%) reported one or more unprotected vaginal or anal sex events at baseline or follow-up. Nine participants who reported unprotected anal or vaginal sex at baseline (11% of the sexually active participants) reported no unprotected sex at follow-up; 3 participants who reported no unprotected sex at baseline (4% of the sexually active participants) reported at least one unprotected sex event at follow-up.
At baseline, a total of 676 vaginal and 12 anal sex events in the preceding 3 months were reported, which increased dramatically at follow-up to 1186 vaginal and 27 anal sex events. Whereas sexual activity increased over the course of the study, the number of unprotected vaginal and anal sex events across the conditions did not; 171 unprotected sex events were reported at baseline versus 165 unprotected sex events at follow-up. Sixty-seven percent of the unprotected sex events at baseline were reported to have occurred with 9 partners believed to be HIV-negative or of unknown HIV status. At 6-month follow-up, 40% of the unprotected sex events reportedly occurred with 4 HIV-negative or unknown status partners. Only 4 participants at baseline and 1 participant at follow-up reported having sex with more than one sexual partner during the assessment period.
CD4+ cell count (at baseline or follow-up), length of time on ARVs, and undetectable viral load (at follow-up) were not significant predictors of unprotected vaginal and anal sex. There were no statistically significant differences between intervention and control condition participants at baseline on any of the demographic variables, the clinical variables (taking ARVs, CD4+ cell counts, and history of opportunistic infections), the primary outcome variable (unprotected vaginal and anal sexual events), or on the total number of vaginal and anal sexual events (Table 1).
Only 17 participants, or 11% of the sample, reported any oral sex events on the baseline or follow-up assessment. Twelve of the 13 participants who reported having oral sex at baseline also reported having vaginal or anal sex. All the participants who reported having oral sex at follow-up also reported having vaginal or anal sex. In the control condition, 2.04% (n = 1) and 0% of participants reported oral sex at baseline and follow-up, respectively. In the intervention condition, 11.65% (n = 12) and 6.85% (n = 5) of participants reported oral sex at baseline and follow-up, respectively. The proportion of participants who engaged in oral sex did not significantly differ between conditions at either time point (baseline and follow-up) nor did it differ between baseline and follow-up as evaluated with χ2 tests.
Of 152 patients who completed the baseline assessment, 129 (85%) also completed the 6-month follow-up assessment. Of the 23 participants who did not complete the follow-up assessment, 9 (39%) had died and 14 (61%) were lost to follow-up. Reasons for loss to follow-up included seeking health care at another clinic and not returning to the clinic for unknown reasons.
Attrition analyses showed no differential attrition (P > 0.50) by study condition for the demographic factors, the clinical factors (taking ARVs, CD4+ count, and history of opportunistic infections), the primary outcome variable (number of unprotected vaginal and anal sex events) or for the total number of vaginal and anal sex events. Furthermore, those who attritted from the study did not differ from those who remained on any of these factors.
Feasibility, Fidelity, and Acceptability Analyses
Assessment of intervention feasibility (based on a review of “Patient Record Forms”) indicated that the intervention was delivered in 216 of 218 (99%) routine patient visits, with a mean of 2.5 intervention visits per patient over a 6-month interval. The 2 visits in which the intervention did not occur was due to other more pressing issues taking precedence. Assessment of intervention fidelity (via “Patient Records Forms” and in vivo observations) revealed that the intervention was delivered with considerable fidelity. In 78.8% of the intervention visits, the counselors implemented at least 7 of the 8 requisite intervention steps (mean = 7.2 steps, mode = 8.0, interquartile range = 1.0).
Exit interviews with a modest number of patients (n = 20) indicated that the counselors implemented the intervention in a supportive, helpful, nonjudgmental fashion. Exit interview findings were consistent with the results from 2 focus group discussions conducted at the completion of the study with HIV-infected patients who had participated in the intervention (1 group with 7 women and 1 with 7 men). Men and women in the focus groups reported that they felt “comfortable” talking with their counselors about their sexual behavior and that it was “easy” to do so because they “trusted” their counselors. The consensus was that counselors seemed at ease talking about sex, seemed to be knowledgeable about the topic (eg, “She seems like she knows everything.” “They're well trained.”), and were not judgmental (eg, “She's accepting of me, whatever I say.” “He doesn't criticize or make you feel badly about yourself.”). Women reported that they found it “very helpful” to have open, honest discussions about their sexual behavior (eg, “It's good to talk about it, so that you know how to enjoy your life.”). Men also indicated that the conversations were beneficial and that they were “gaining a lot” from them. When asked for suggestions on how to improve these discussions, women indicated that they would like to have them “more often” and would like to see them focus more on “relationship” issues rather than just on sex and condoms. The men voiced satisfaction with the frequency and content of the discussions.
Intervention Outcome Analyses
Analysis of intervention impact on the primary outcome of number of unprotected vaginal and anal sexual events revealed a significant condition × time interaction [b = −2.24, se = 0.92, P = 0.016, ERR = 0.11, confidence interval (CI) = 0.01 to 0.87]. The mean number of reported unprotected vaginal and anal sexual events “decreased” significantly over time (from 2.64 to 0.40) among HIV-infected patients who received the intervention (b = −1.41, se = 0.65, P < 0.05, ERR = 0.24, CI = 0.07 to 0.89), whereas there was a marginally significant “increase” in the mean number of unprotected sexual events (from 2.26 to 3.85) over time for HIV-infected patients in the standard-of-care control condition (b = 0.67, se = 0.34, P = 0.05, ERR = 1.95, CI = 1.0 to 3.83) (Table 2). There were no overall gender differences nor were there gender differences in the effectiveness of the intervention. Demographic and clinical characteristics listed in Table 1 did not moderate any of the effects.
Changes over time and across conditions in the number of unprotected sex events with perceived HIV-negative or unknown status partners were also examined. There was a marginally significant condition × time interaction (b = −2.35, se = 0.10, P = 0.06, ERR = 0.096, CI = 0.008 to 1.11), such that the number of unprotected sex events with HIV-negative or unknown status partners decreased from baseline to follow-up for participants in the intervention condition but increased over time for participants in the control condition. Limitations of sample size, however, precluded post hoc comparisons. The interaction effect did not vary significantly by gender, and there were no independent effects of condition, time, or gender. Demographic and clinical characteristics listed in Table 1 did not moderate the interaction effect.
Finally, changes in the total number of vaginal and anal sex events (protected and unprotected) over time across conditions were examined. There was a statistically significant effect of time (b = 70, se = 0.16, P < 0.01, ERR = 2.01, CI = 1.48 to 2.74), with the number of sex events increasing from baseline to follow-up regardless of condition. This effect was moderated by gender (b = −0.31, se = 0.73, P < 0.05, ERR = 0.73, CI = 0.57 to 0.94) such that men displayed greater increases than did women (Table 2). There was no effect of condition on the total number of sex events. Also, there was no significant condition × time interaction, which indicates that the conditions did not differ in the extent to which the number of sex events increased over time. Demographic and clinical characteristics did not moderate any of the effects.
Intervention outcome results for sex, unprotected sex, and unprotected sex with HIV-negative and unknown serostatus partners are summarized in Figure 1.
The current findings from this pilot study are among the first to demonstrate that a brief counselor-delivered HIV risk reduction intervention for PLWHA in South Africa is feasible to implement during the course of routine clinical care, can be delivered with fidelity, is acceptable to HIV+ patients, and may be effective at reducing HIV-infected patients' unprotected vaginal and anal sexual behavior. Specifically, whereas patients in the standard-of-care control condition reported an increase over a 6-month period of time in the mean number of unprotected sex events, those in the intervention condition reported a significant reduction in those events. This reduction in risky behavior occurred despite the fact that PLWHA overall increased their sexual activity over the course of the study. The increase across conditions in sexual activity is not surprising considering that most patients were newly prescribed ARV medications and experienced improvements in objective health (eg, increases in CD4+ count and decreases in viral load) over the course of the study; this likely led to improvements in subjective health, thereby increasing patients' interest in and ability to have sex. The fact that those in the intervention condition became more sexually active but reduced their risky behavior strongly supports the potential effectiveness of this intervention approach.
The finding that HIV-infected patients in the standard-of-care condition increased their unprotected sexual behavior across time is consistent with the findings from several studies.18,30-33 Although the reasons for this increase in risky behavior are not clear, such findings underscore the urgency of linking HIV prevention with HIV care.
Limitations of the current study include the use of a single clinic, a limited number of counselors, a relatively small sample size, reliance on self-reports of unprotected sexual behavior, and a relatively short follow-up period. Because participants in the intervention condition were being routinely and systematically encouraged to practice safer sexual behavior and those in the control condition were not, there is also the possibility of desirability bias in the reporting of unprotected sex by the intervention condition participants. In other words, the intervention condition participants may have reported fewer unprotected sex events to please the researchers and clinical staff, not because they were actually engaging in less risky sex. In addition, the generalizability of these findings to public health care clinics throughout KwaZulu-Natal and South Africa is limited by the fact that these clients were required to pay $20 per month for clinic visits, and public health care clinics do not require their patients to pay for care; the participants in this study may therefore not be representative of the PLWHA who attend public health care clinics. A large-scale randomized controlled trial that includes a representative sample of counselors, patients, and clinical care sites; the use of biological outcomes; and a longer follow-up period is required to more fully evaluate this intervention. In addition, it would be valuable to compare the Options for Health intervention with another risk reduction intervention to determine which is more effective and to control for social desirability in reporting.
To help stop the spread of HIV in South Africa, it is essential that efficient and effective interventions be developed and deployed to reduce risky sexual behavior among PLWHA.14,34 With the rollout of ARV medications in South Africa, the clinical care setting provides the greatest possible access to HIV-infected persons and offers repeated opportunities for HIV risk reduction discussions.14,35 The Izindlela Zokuphila/Options for Health intervention capitalizes on those opportunities by having HIV adherence counselors conduct ongoing, tailored risk reduction discussions with PLWHA during routine clinical visits. Training counselors to criterion in the intervention takes a relatively short period of time and can be readily integrated into ongoing Department of Health trainings in which counselors are required to participate.36 Results from this study thus suggest that this intervention may be a promising first step toward integrating HIV prevention and care in South Africa and reducing HIV risk behavior among PLWHA.
The authors express their appreciation to the participants in the study and to the following individuals for their contribution to the project: Dr Helga Holst, Dr Janet Giddy, Sizakele Ndlovu, Sibongile Hadebe, Trueman Gambushe, Aletta Sabasaba, Muse Cele, Sbongile Maimane, and Patricia Masebe.
3. Shisana O, Rehle T, Simbayi LC, et al. South African National HIV Prevalence, HIV Incidence, Behaviour and Communication Survey, 2005
. Cape Town, South Africa: Human Sciences Research Council Publishers; 2006.
5. Mbewu A, Simelela N, Makubalo, L, et al. Operational Plan for Comprehensive HIV and AIDS Care, Management, and Treatment for South Africa
. Pretoria, South Africa: Republic of South Africa Department of Health; November 19, 2003. Available at: http://www.info.gov.za/otherdocs/2003/aidsplan/index.html
. Accessed October 27, 2006.
7. Hendriksen ES, Pettifor A, Lee S-J, et al. Predictors of condom use among young adults in South Africa: the Reproductive Health and HIV Research Unit National Youth Survey. Am J Public Health
8. Kiene SM, Christie S, Cornman DH, et al. Sexual risk behaviour among HIV-positive individuals in clinical care in urban KwaZulu-Natal, South Africa. AIDS
9. Maharaj P, Cleland J. Condom use within marital and cohabiting partnerships in KwaZulu-Natal, South Africa. Stud Fam Plann
10. Olley BO, Seedat S, Gxamza F, et al. Determinants of unprotected sex among HIV-positive patients in South Africa. AIDS Care
11. Simbayi LC, Kalichman SE, Strebel A, et al. Disclosure of HIV status to sex partners and sexual risk behaviours among HIV positive men and women in Cape Town, South Africa. Sex Transm Infect
15. Kalichman SC, Rompa D, Cage M, et al. Effectiveness of an intervention to reduce HIV transmission risks in HIV-positive people. Am J Prev Med
16. Fisher JD, Cornman DH, Osborn CY, et al. Clinician-initiated HIV-risk reduction intervention for HIV+ persons: formative research, acceptability, and fidelity of the Options Project. J Acquir Immune Defic Syndr
17. Richardson JL, Milam J, McCutchan A, et al. Effect of brief provider safer-sex counseling of HIV-1 seropositive patients: a multi-clinic assessment. AIDS
18. Fisher JD, Fisher WA, Cornman DH, et al. Clinician-delivered intervention during routine clinical care reduces unprotected sexual behavior among HIV-infected patients. J Acquir Immune Defic Syndr
19. Fisher JD, Fisher WA. Changing AIDS risk behavior. Psychol Bull
20. Fisher JD, Fisher WA. Theoretical approaches to individual-level change in HIV risk behavior. In: Peterson JL, DiClemente R, eds. HIV Prevention Handbook
. New York: Kluwer Academic; 2000.
21. Fisher JD, Fisher WA. The information-motivation-behavioral skills model. In: DiClemente R, Crosby R, Kegler R, eds. Emerging Promotion Research and Practice
. San Francisco, CA: Josey Bass Publishers; 2002.
22. Fisher WA, Fisher JD. Understanding and promoting AIDS preventive behavior: a conceptual model and educational tools. Can J Hum Sex
23. Fisher WA, Fisher JD, Harman JJ. The information-motivation-behavioral skills model: a general social psychological approach to understanding and promoting health behavior. In: Suls J, Wallston KA, eds. Social Psychological Foundations of Health and Illness
. Malden, MA: Blackwell Publishing; 2003:82-106.
24. Miller WR, Rollnick S. Motivational Interviewing: Preparing People to Change Addictive Behavior
. New York: Guilford Press; 1991.
25. Rollnick S, Mason P, Butler C. Health Behavior Change: A Guide for Practitioners
. London, United Kingdom: Churchill Livingstone; 1999.
26. Rothenberg RB, Scarlett M, del Rio C, et al. Oral transmission of HIV: editorial review. AIDS
27. Bryk AS, Raudenbush SW, Congdon R. HLM 6.0
[computer program]. Chicago, IL: Scientific Software; 2005.
28. Cohen J, Cohen P, West SG, et al. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences
. 3rd ed. Mahwah, NJ: Lawrence Erlbaum Associates; 2003.
29. Edwards LJ. Modern statistical techniques for the analysis of longitudinal data in biomedical research. Pediatr Pulmonol
30. Van de Ven P, Prestage G, Crawford J, et al. Sexual risk behaviour increases and is associated with HIV optimism among HIV-negative and HIV-positive gay men in Sydney over the four-year period to February 2000. AIDS
31. Dukers NH, Goudsmit J, de Wit JB, et al. Sexual risk behaviour relates to the virological and immunological improvements during highly active antiretroviral therapy in HIV-1 infection. AIDS
32. Desquilbet L, Deveau C, Goujard C, et al. Increase in at-risk sexual behavior among HIV-1 infected patients followed in the French PRIMO cohort. AIDS
33. Elford J, Bolding G, Sherr L. High-risk sexual behaviour increases among London gay men between 1998 and 2001: what is the role of HIV optimism? AIDS
34. Valdiserri RO. International scale-up for antiretroviral treatment: where does prevention fit? J Acquir Immune Defic Syndr
. 2004;37(Suppl 2):S138-S141.
35. CDC/HRSA/NIH/HIVMA of the IDSA. Recommendations for incorporating HIV prevention into the medical care of HIV-infected persons. Infect Dis Clin Pract