What do South African adolescents want in a sexual health service? Evidence from the South African Studies on HIV in Adolescents (SASHA) project.

BACKGROUND: Young people in sub-Saharan Africa (SSA) are disproportionately affected by HIV, sexually transmitted infections and unplanned pregnancies. The provision of accessible sexual and reproductive health services (SRHS) for young people in SSA is vital to reduce this burden. OBJECTIVES: To examine the needs of South African (SA) adolescents with regard to differentiated, accessible and adolescent-responsive SRHS. METHODS: Data were drawn from a larger project examining the feasibility of conducting HIV vaccine trials in adolescents. Fifteen focus group discussions were conducted across five research sites in four SA provinces with 120 male and female adolescent human papillomavirus vaccine trial participants aged 12 - 19 years from low-income areas with a high incidence of HIV. Transcribed data were double-coded using framework analysis. RESULTS: Three main themes emerged on how best to improve SRHS for adolescents in resource-limited settings: adolescent-friendly services, availability of developmentally appropriate and tailored information, and improved relationships between healthcare workers and clinic attendees. Participants wanted more flexible opening hours at SRHS to account for travel time to clinics from school and home. They suggested that services include contraception, counselling, educational materials, links to adoption services, emergency vehicles, pre- and postnatal care, and improved service quality from clinic staff. CONCLUSIONS: While dedicated adolescent SRHS might best meet the needs of young people in SA, the study suggests that failing this, existing SRHS should be more responsive to adolescent use. Innovations such as mobile outreach services, self-testing and flexible hours will help SRHS respond to adolescents' needs.

Feasibility and acceptability of conducting HIV vaccine trials in adolescents in South Africa: Going beyond willingness to participate towards implementation.

BACKGROUND: HIV/AIDS remains a leading cause of death in adolescents (aged 15 - 25 years), and in sub-Saharan Africa HIV-related deaths continue to rise in this age group despite a decline in both adult and paediatric populations. This is attributable in part to high adolescent infection rates and supports the urgent need for more efficacious prevention strategies. In particular, an even partially effective HIV vaccine, given prior to sexual debut, is predicted to significantly curb adolescent infection rates. While adolescents have indicated willingness to participate in HIV vaccine trials, there are concerns around safety, uptake, adherence, and ethical and logistic issues. OBJECTIVES: To initiate a national, multisite project with the aim of identifying obstacles to conducting adolescent HIV vaccine trials in South Africa (SA). METHOD: A simulated HIV vaccine trial was conducted in adolescents aged 12 - 17 years across five SA research sites, using the already licensed Merck human papillomavirus vaccine Gardasil as a proxy for an HIV vaccine. Adolescents were recruited at community venues and, following a vaccine discussion group, invited to participate in the trial. Consent for trial enrolment was obtained from a parent or legal guardian, and participants aged 16 - 17 years were eligible only if sexually active. Typical vaccine trial procedures were applied during the five study visits, including the administration of vaccination injections at study visits 2, 3 and 4. RESULTS: The median age of participants was 14 years (interquartile range 13 - 15), with 81% between the ages of 12 and 15 years at enrolment. Overall, 98% of screened participants opted to receive the vaccine, 588 participants enrolled, and 524 (89%) attended the final visit. CONCLUSIONS: This trial showed that adolescents can be recruited, enrolled and retained in clinical prevention trials with parental support. While promising, these results were tempered by the coupling of sexual-risk eligibility criteria and the requirement for parental/guardian consent, which was probably a barrier to the enrolment of high-risk older adolescents. Further debate around appropriate consent approaches for such adolescents in HIV prevention studies is required.

Feasibility and acceptability of conducting HIV vaccine trials in adolescents in South Africa: Going beyond willingness to participate towards implementation

Background. HIV/AIDS remains a leading cause of death in adolescents (aged 15 - 25 years), and in sub-Saharan Africa HIV-related deaths continue to rise in this age group despite a decline in both adult and paediatric populations. This is attributable in part to high adolescent infection rates and supports the urgent need for more efficacious prevention strategies. In particular, an even partially effective HIV vaccine, given prior to sexual debut, is predicted to significantly curb adolescent infection rates. While adolescents have indicated willingness to participate in HIV vaccine trials, there are concerns around safety, uptake, adherence, and ethical and logistic issues.Objectives. To initiate a national, multisite project with the aim of identifying obstacles to conducting adolescent HIV vaccine trials in South Africa (SA).Method. A simulated HIV vaccine trial was conducted in adolescents aged 12 - 17 years across five SA research sites, using the already licensed Merck human papillomavirus vaccine Gardasil as a proxy for an HIV vaccine. Adolescents were recruited at community venues and, following a vaccine discussion group, invited to participate in the trial. Consent for trial enrolment was obtained from a parent or legal guardian, and participants aged 16 - 17 years were eligible only if sexually active. Typical vaccine trial procedures were applied during the five study visits, including the administration of vaccination injections at study visits 2, 3 and 4.Results. The median age of participants was 14 years (interquartile range 13 - 15), with 81% between the ages of 12 and 15 years at enrolment. Overall, 98% of screened participants opted to receive the vaccine, 588 participants enrolled, and 524 (89%) attended the final visit.Conclusions. This trial showed that adolescents can be recruited, enrolled and retained in clinical prevention trials with parental support. While promising, these results were tempered by the coupling of sexual-risk eligibility criteria and the requirement for parental/guardian consent, which was probably a barrier to the enrolment of high-risk older adolescents. Further debate around appropriate consent approaches for such adolescents in HIV prevention studies is required

Longitudinal determinants of consistent condom use by partner type among young injection drug users: the role of personal and partner characteristics.

We investigated the longitudinal influence of individual-, relationship- and social-level factors on condom use by partner type among young injections drug users (IDUs) enrolled in the Collaborative Injection Drug Users Study-III/Drug Users Intervention Trial (CIDUS-III/DUIT) from 2002 to 2004. Based on longitudinal analysis using generalized estimating equations (GEE), consistent condom use with main partners was more commonly reported among males and those with greater self-efficacy for condom use; main partner's desire for pregnancy and needle sharing were negatively associated with consistent condom use. Among those with casual partners, having fewer sex partners was associated with consistent condom use. Positive attitudes toward condom use and partner norms supporting condom use were associated with greater consistent condom use with both partner types. These findings suggest that intervention strategies targeting individual- and partner-level factors may provide avenues for intervening upon sexual risks among young IDUs.

Dynamics of electroencephalogram entropy and pitfalls of scaling detection.

In recent studies a number of research groups have determined that human electroencephalograms (EEG) have scaling properties. In particular, a crossover between two regions with different scaling exponents has been reported. Herein we study the time evolution of diffusion entropy to elucidate the scaling of EEG time series. For a cohort of 20 awake healthy volunteers with closed eyes, we find that the diffusion entropy of EEG increments (obtained from EEG waveforms by differencing) exhibits three features: short-time growth, an alpha wave related oscillation whose amplitude gradually decays in time, and asymptotic saturation which is achieved after approximately 1 s. This analysis suggests a linear, stochastic Ornstein-Uhlenbeck Langevin equation with a quasiperiodic forcing (whose frequency and/or amplitude may vary in time) as the model for the underlying dynamics. This model captures the salient properties of EEG dynamics. In particular, both the experimental and simulated EEG time series exhibit short-time scaling which is broken by a strong periodic component, such as alpha waves. The saturation of EEG diffusion entropy precludes the existence of asymptotic scaling. We find that the crossover between two scaling regions seen in detrended fluctuation analysis (DFA) of EEG increments does not originate from the underlying dynamics but is merely an artifact of the algorithm. This artifact is rooted in the failure of the "trend plus signal" paradigm of DFA.

Wavelet assessment of cerebrospinal compensatory reserve and cerebrovascular reactivity.

We introduce a wavelet transfer model to relate spontaneous arterial blood pressure (ABP) fluctuations to intracranial pressure (ICP) fluctuations. We employ a complex continuous wavelet transform to develop a consistent mathematical framework capable of parametrizing both cerebral compensatory reserve and cerebrovascular reactivity. The frequency-dependent gain and phase of the wavelet transfer function are introduced because of the non-stationary character of the ICP and ABP time series. The gain characterizes the dampening of spontaneous ABP fluctuations and is interpreted as a novel measure of cerebrospinal compensatory reserve. For a group of 12 patients who died as a result of cerebral lesions (Glasgow Outcome Scale (GOS) = 1) the average gain in the low-frequency (0.02- 0.07 Hz) range was 0.51 +/- 0.13 and significantly exceeded that of 17 patients with GOS = 2 having an average gain of 0.26 +/- 0.11 with p = 1x10(-4) (Kruskal-Wallis test). A time-averaged synchronization index (which may vary from 0 to 1) defined in terms of the wavelet transfer function phase yields information about the stability of the phase difference of the ABP and ICP signals and is used as a cerebrovascular reactivity index. A low value of synchronization index reflects a normally reactive vascular bed, while a high value indicates pathological entrainment of ABP and ICP fluctuations. Such entrainment is strongly pronounced in patients with fatal outcome (for this group the low-frequency synchronization index was 0.69 +/- 0.17). The gain and synchronization parameters define a cerebral hemodynamic state space (CHS) in which the patients with GOS = 1 are to large extent partitioned away from those with GOS = 2. The concept of CHS elucidates the interplay of vascular and compensatory mechanisms.

Correlates of consistent condom use with main partners by partnership patterns among young adult male injection drug users from five US cities.

This paper examined correlates of consistent condom use with a main partner among heterosexual male injection drug users (IDUs). Using data from a multi-site sample of young IDUs, we identified 1770 sexually active men of whom 24% (429/1770) reported an exclusive main female sex partner and 49% (862/1770) reported both main and casual female sex partners. Consistent condom use with a main partner was low among men with an exclusive main partner and those with multiple partners (12% and 17%, respectively). In multivariate analysis, consistent condom use with a main partner across partnership patterns was directly associated with anticipating a positive response to requests for condom use and by partner support of condom use; consistent condom use was inversely associated with a main partner's pregnancy desires. Among men with an exclusive main partner, consistent condom use was also inversely associated with needle sharing with a main partner. Among men with multiple partners, consistent condom use with a main partner was inversely associated with injecting with a used needle and intimate partner violence. The low prevalence of consistent condom use with main female partners among heterosexually active male IDUs indicates an increased risk for HIV transmission between men and their primary sex partners. Interventions for heterosexual males that are geared toward increasing condom use in primary relationships are warranted.

The influence of acute progressive hypoxia on bioelectrical activity of the brain.

Hypoxia, a noxious and hyperventilatory stimulus and a modifier of neuronal metabolism, could influence cortical function. In this study we attempted to assess any such influence, its determinants, and particularly the role in it of the accompanying hypoxic emotional distress. We addressed the issue by examining the associations among EEG, ventilation, and anxiety during progressive poikilocapnic hypoxia (end-point SaO(2) approximately 75%) in 12 awake healthy volunteers (mean age 27.5 +/-0.7 yr). All subjects hyperventilated in response to hypoxia and 3 of them had a high level of anticipatory anxiety that forced one person to discontinue the test. We failed to show any major effect of hypoxia on the EEG pattern analyzed by visual inspection or wavelet power spectra. Therefore, no relationship between the ventilatory and cortical activity responses to hypoxia could be established. Cortical activity changes appeared, however, in the subjects who experienced emotional distress during the test. These changes were apparent on an expanded analysis of the EEG signal by the use of the Lempel-Ziv complexity that takes into account the ordering of variations in the signal, rather than only the relative frequency of events analyzed by the Shannon entropy. The Lempel-Ziv complexity offers promise as a novel method for unraveling fine and otherwise unexpressed alterations in cortical bioelectrical activity.

Wavelet mapping of sleep spindles in young patients with epilepsy.

Using the wavelet mapping of sleep spindles we investigated influence of focal epilepsy on spindle generation. We found that the maximum of sleep spindle intensity is usually localized away from the epileptic focus. We discuss the possibility of the application of wavelet mapping for localization of epileptic foci prior to epileptic neurosurgery.

Factors associated with interest in initiating treatment for hepatitis C Virus (HCV) infection among young HCV-infected injection drug users.

OBJECTIVE: We sought to identify factors associated with interest in receiving therapy for hepatitis C virus (HCV) infection among HCV-infected injection drug users (IDUs) in 3 United States cities. METHODS: IDUs aged 18-35 years who were HCV-infected and seronegative for human immunodeficiency virus underwent surveys on behaviors, experience, and interest in treatment for HCV infection and readiness to quit drug use. RESULTS: Among treatment-naive IDUs (n=216), 81.5% were interested in treatment for HCV infection, but only 27.3% had seen a health-care provider since receiving a diagnosis of HCV infection. Interest in treatment for HCV infection was greater among IDUs with a high perceived threat of progressive liver disease, those with a usual source of care, those without evidence of alcohol dependence, and those with higher readiness scores for quitting drug use. Interest in treatment for HCV infection was 7-fold higher among IDUs who were told by their health-care provider that they were at risk for cirrhosis or liver cancer. CONCLUSIONS: Improving provider-patient communication and integrating treatments for substance abuse and HCV may increase the proportion of IDUs who initiate treatment for HCV infection.

Acceptance and completion of hepatitis B vaccination among drug users in New York City.

Hepatitis B (HBV) vaccination rates remain low among drug users. We examined correlates of vaccine acceptance and completion in two ongoing prospective studies of young injecting and non-injecting drug users in New York City. Street recruited drug users were enrolled at one of two neighbourhood locations (Harlem and the Bronx) between 2000 and 2004 and completed risk behaviour questionnaires and HBV testing. Free HBV vaccination was offered. Among 1117 participants, 26.1% (275) had a previous HBV infection, 57.9% (610) were susceptible to HBV, and 16.0% (169) had serological evidence of previous vaccination. Of the 610 participants susceptible to HBV, 466 (76.4%) returned for their results and were offered vaccination; 53.9% (251) received at least one dose of the vaccine (acceptors). Correlates of vaccine acceptance included older age, public assistance as main income source, and being recruited in the Bronx. Daily crack users were significantly less likely to initiate the vaccine series. Among 240 vaccine acceptors, 98 (40.8%) completed all three doses. Daily injectors, Hispanics, and those recruited in Harlem were less likely to complete the vaccination series. HBV vaccination acceptance among drug users seems likely in programmes that are convenient and offer remuneration; however, extended efforts are needed to improve series completion.

Use-effectiveness of the female versus male condom in preventing sexually transmitted disease in women.

BACKGROUND: Data are limited on the female condom's effectiveness against STDs. GOAL: The goal was to compare STD rates between women given small-group education on, and free supplies of, either female or male condoms. STUDY DESIGN: Female patients at an STD clinic (n = 1442) were randomly assigned to condom type and followed via medical records for STDs (gonorrhea, chlamydia, early syphilis, or trichomoniasis). RESULTS: In an intention-to-treat analysis, the odds ratio for a comparison of STD occurrence between the female and male condom groups was 0.75 (95% confidence interval [CI], 0.56-1.01), and it did not change with adjustment. In a second analysis among women returning for subsequent screening, incidence rates for the first new postintervention STD per 100 woman-months of observation were 6.8 in the female condom group and 8.5 in the male condom group (rate ratio = 0.79 [CI, 0.59-1.06]). CONCLUSION: Compared with those provided with male condoms alone, women counseled on, and provided with, female condoms fared no worse and experienced a nonsignificant reduction in STDs.

Achieving safer sex with choice: studying a women's sexual risk reduction hierarchy in an STD clinic.

A flexible, risk-reduction approach, as compared with a single method approach, may increase sexually transmitted disease (STD)/HIV protection for women attending STD clinics. A brief intervention was tested in an observational study of 292 STD clinic patients in three distinct cohorts. These included subjects counseled on (1) the "women's safer sex hierarchy of prevention methods" (hierarchy cohort, n = 118), including the female condom (FC), male condom (MC), diaphragm, cervical cap, and spermicides, (2) MC only (n = 62), or (3) FC (n = 112) only. We evaluate method use and level of protection achieved at 6-month follow-up among the women in the hierarchy cohort and compare the level of unprotected sex across the three cohorts, using ordinal logistic regression analyses and an imputation procedure to account for attrition. In the hierarchy cohort, the MC, FC, spermicidal film, foam, suppository, and diaphragm were used with main partners by 80%,46%, 37%, 28%, 17%, and 5% of women, respectively. Spermicides were used frequently, mainly in conjunction with condoms. As compared with hierarchy subjects, both MC cohort subjects (OR = 2.3, p = 0.01) and FC cohort subjects (OR = 1.6, p = 0.11) were more likely to report 100% unprotected sex. The tendency for subjects to move toward higher levels of protection was observed most strongly in the hierarchy group. Hierarchical-type counseling, compared with single method counseling, leads to increased protection during sex among women at high risk of STD/HIV infection and should be implemented in STD clinics.

Health care utilization among young adult injection drug users in Harlem, New York.

PURPOSE: This cross-sectional study investigated the predictors for and patterns of health care utilization among young adult injection drug users (IDUs). METHODS: The subjects were 206 IDUs, ages 18-29, who were street-recruited from Harlem, New York. Participants were interviewed about their drug use, health conditions, and use of services such as health care, needle exchange programs (NEPs), and drug treatment in the preceding 6 months. Data were analyzed using logistic regression. RESULTS: Health insurance was associated with use of health care both among NEP users [AOR (adjusted odds ratio) 10.66] and non-NEP users (AOR 2.45). Use of health care was independently associated with drug treatment (AOR 2.58), being gay/bisexual (AOR 3.86), and negatively associated with injecting cocaine (AOR 0.56). Half the participants (49%) had used health care in the previous 6 months; 48% were uninsured. Many participants who did not use health services reported a condition that would have warranted medical care. IMPLICATIONS: Health insurance was strongly associated with use of health care, particularly among those who attend NEPs. Young adult IDUs may benefit from increased efforts to help them arrange and maintain health insurance coverage, potentially at NEPs. NEPs may be connecting young IDUs with health insurance to medical care through referrals.

Prevalence, incidence, and correlates of chlamydia and gonorrhea among young adult injection drug users.

PURPOSE: To measure prevalence, incidence, and correlates of chlamydia and gonorrhea among injection drug users (IDUs). METHODS: Participants (n = 2129; 63% male, 52% white, ages 18-30 years) in five US cities were tested for chlamydia and gonorrhea by urine LCR assay and completed a standardized questionnaire about demographics and recent sexual behavior. Logistic regression identified correlates of prevalent infection; incidence rates were calculated from 6-month follow-up data. RESULTS: Chlamydia prevalence was 5.2% and did not differ by gender. Gonorrhea prevalence was 0.2% among men and 2.0% among women, P < .001. Among men, younger age [OR (95% CI): 0.89 (0.83-0.96)], age at sexual debut [0.91 (0.83-0.99)], and African American race [2.92 (1.53-5.59)] were associated with chlamydia. Among women, age at sexual debut [1.16 (1.02-1.31)] and commercial sex [1.96 (1.03-3.74)] were associated with chlamydia, and with gonorrhea [1.27 (1.04-1.56)] and [5.17 (1.66-16.11)], respectively. At 6 months, the cumulative incidence of chlamydia was 1.7% among men and 4.4% among women, P = .03; no men and 1.3% of women tested positive for gonorrhea, P = .01. IMPLICATIONS: Prevalence and correlates of chlamydia and gonorrhea were similar to other samples, suggesting that screening criteria need not be modified for IDU populations. The number of behavioral correlates identified was limited; perhaps unmeasured sexual-network-level factors play a role in determining sexually transmitted disease (STD) prevalence.

Female-initiated barrier methods for the prevention of STI/HIV: where are we now? Where should we go?

The female condom has been on the US market for over 8 years and was hailed as a method that would allow women greater control in protecting themselves against unwanted pregnancy and sexually transmitted infections. However, since its launch, promotion of the female condom has met with challenges that vary from provider bias against the device, concerns about efficacy, and doubts about whether it will be used. While daunting, many of these challenges are not unique to the female condom. In fact, they parallel those of the tampon when it was first promoted in the US for menstrual hygiene in the early 1930s. Many providers were initially opposed to the tampon; early users found the tampon inadequate. Ten years after the introduction of the tampon and despite mass marketing, acceptability was mixed and use was modest (25% prevalence in a 26-city survey). Similar to female condom use observed in current-day prospective studies, users did not use the tampon exclusively or even predominantly. The story of the tampon demonstrates that a method does not have to be adopted by all users immediately to play an important role, and that even initially controversial methods can become widely accepted as mainstream. The early history of the tampon sheds perspective, and hope, on where we stand with the female condom. While much-needed work is under way to develop microbicides, we should take advantage of the fact that the female condom is already available and promote it to those in need now.

Acceptability of the Reality female condom and a latex prototype.

We report on the comparative acceptability of a prototype latex female condom and the polyurethane Reality female condom. We also identified factors associated with acceptability, measured via a composite index with domains related to ease of insertion, noise, and comfort during insertion and use. There were 135 couples in this randomized crossover trial. The average age was 30 years; more than 60% had education beyond high school; 40% were married; and participants were at low risk for sexually transmitted disease and pregnancy (due to the investigational status of the prototype). Participants were asked to use three of each of the study condoms during a 6-week period. Acceptability ratings on 12 items were summed into a composite index for each participant by condom type. The index midpoint (range) for females was 48 (12-84), and it was 32 (8-56) for males, with lower scores indicating higher acceptability (men completed only a subset of the acceptability questions). Both condoms were equally acceptable: Mean scores were 37 and 40 for the women's ranking of the prototype and Reality, respectively (P =.07) and 29 and 30 for men's rankings, respectively (P =.35). Multiple regression models to predict acceptability scores by gender were somewhat uninformative (most R(2) values were less than 0.10). Nevertheless, minority ethnicity (African American or Hispanic vs. white) was associated with higher acceptability by both genders for both condom types. Among women, for both condom types, less education (less than high school compared with high school or beyond) was associated with higher acceptability. Female condom acceptability may not be equally distributed across demographic groups, which is important for health educators to keep in mind when promoting the female condom.

Collaboration among community members, local health service providers, and researchers in an urban research center in Harlem, New York.

The Urban Research Center at the Center for Urban Epidemiologic Studies brings together community members and researchers working in Harlem, New York. A Community Advisory Board (CAB) composed of community members, service providers, public health professionals, and researchers was formed to assist the Center's research and interventions and to guide community partnerships. Through a collaborative process, the CAB identified three public health problems-substance use, infectious diseases, and asthma-as action priorities. To deal with substance use, the Center created a Web-based resource guide for service providers and a "survival guide" for substance users, designed to improve access to community services. To deal with infectious diseases, the Center is collaborating with local community-based organizations on an intervention that trains injection drug users to serve as peer mentors to motivate behavior change among other injection drug users. To deal with asthma, the Center is collaborating with community child care providers on an educational intervention to increase asthma awareness among day care teaching staff, enhance communication between staff and families, and improve the self-management skills of children with asthma. The Center's experience has demonstrated that active communities and responsive researchers can establish partnerships that improve community health.

Structural integrity of the female condom after a single use, washing, and disinfection.

The Reality(R) female condom is approved for use during a single act of intercourse, but is expensive relative to other barrier methods. Re-use is a potential strategy to reduce its per-use cost. We tested the structural integrity of female condoms (n = 318) after a single act of vaginal intercourse. We also measured the impact of laboratory washing (1, 5, or 10 times) with and without disinfection on the structural integrity of unused condoms. Structural integrity was measured via 5 tests: seam tensile strength, water leakage, air-burst, tear propagation, and device dimensions. No degradation in device structural integrity occurred after a single use when compared to control for seam tensile (16.0 vs.15.7 mPa; p = 0.558); water leakage (1.9% vs. 0.9%; p = 0.618); air burst (3.9 vs. 3.6 kPa; p <0.001); or tear propagation (344.6 vs. 336.8 psi; p = 0.313). Mean length was slightly increased [single use vs. control (177.9 vs. 172.5 mm; p <0.001)]. No consistent pattern of structural degradation emerged across all wash/disinfection groups. Our data suggest the structural integrity of the female condom remains intact after a single use and cleaning.

Male-condom and female-condom use among women after counseling in a risk-reduction hierarchy for STD prevention.

BACKGROUND: A concern with hierarchy messages, which promote male condoms and female-controlled barrier methods along a prevention continuum, is that they may discourage condom use. GOAL: To measure male-condom and female-condom use among women who received hierarchy counseling and compare this with women counseled about condoms only. STUDY DESIGN: Three observational cohorts that correspond to prevention message received were assembled, and consisted of female sexually transmitted disease clinic patients who were counseled about male condoms, female condoms, or a hierarchy message. The hierarchy message promoted male and female condoms, the diaphragm and cervical cap, spermicides, and withdrawal, in descending order of effectiveness against sexually transmitted diseases. After counseling, women were interviewed and returned for follow-up visits at 2 weeks, 4 months, and 6 months. The outcome was the mean proportion of male condom- or female condom-protected coital acts at each follow-up visit in the hierarchy cohort. The outcome was dichotomized as high (> or = 70% of coital acts protected) or low (< 70%), and generalized estimating equations were used to compare observed follow-up condom use with baseline within the hierarchy cohort and observed follow-up condom use between cohorts. It was assumed that condom use in persons not present at 6 months was equal to baseline levels, and condom use estimates were calculated for each full cohort that was initially enrolled.