AmericasNLI: Machine translation and natural language inference systems for Indigenous languages of the Americas.

Little attention has been paid to the development of human language technology for truly low-resource languages-i.e., languages with limited amounts of digitally available text data, such as Indigenous languages. However, it has been shown that pretrained multilingual models are able to perform crosslingual transfer in a zero-shot setting even for low-resource languages which are unseen during pretraining. Yet, prior work evaluating performance on unseen languages has largely been limited to shallow token-level tasks. It remains unclear if zero-shot learning of deeper semantic tasks is possible for unseen languages. To explore this question, we present AmericasNLI, a natural language inference dataset covering 10 Indigenous languages of the Americas. We conduct experiments with pretrained models, exploring zero-shot learning in combination with model adaptation. Furthermore, as AmericasNLI is a multiway parallel dataset, we use it to benchmark the performance of different machine translation models for those languages. Finally, using a standard transformer model, we explore translation-based approaches for natural language inference. We find that the zero-shot performance of pretrained models without adaptation is poor for all languages in AmericasNLI, but model adaptation via continued pretraining results in improvements. All machine translation models are rather weak, but, surprisingly, translation-based approaches to natural language inference outperform all other models on that task.

Lower Optimal Treatment Adherence Among Youth Living With HIV in a Universal Health Care Setting Where ART Is Available at No Cost.

PURPOSE: We assessed differences in optimal adherence between youth (aged 15-29 years) and adults (aged ≥30 years) enrolled in the British Columbia Centre for Excellence in HIV/AIDS Drug Treatment Program from 2010 to 2016. METHODS: Population-level clinical data were used to compare optimal antiretroviral therapy adherence (≥95%), based on pharmacy refill data, among youth and adults. Unadjusted and adjusted generalized estimating estimates were performed to examine the independent relationship between time-dependent age categories and optimal adherence, adjusting for confounders. Factors associated with optimal adherence among youth were examined. RESULTS: Data for 7,485 individuals living with HIV were included. Median follow-up was 7 years (Q1-Q3: 4-7). Over the study period, the number of individuals categorized as "youth" ranged from 820 in 2010 to 291 in 2016. Multivariable models found youth living with HIV were significantly less likely to be optimally adherent than adults (adjusted odds ratio [aOR] = .55; 95% confidence interval [CI]: .49-.62), after controlling for potential confounders, although youth adherence improved significantly during the study period. Among youth, increasing time-dependent age (aOR = 1.18/year older; 95%CI: 1.11-1.25) and number of years on antiretroviral therapy (aOR = 1.15, 95%CI: 1.10-1.19) were independently associated with optimal adherence, while Hepatitis C-positive serostatus (aOR = .55; 95%CI: .33-.92) and multiple treatment regimen change (aOR = .89/regimen change; 95%CI: .81-.97) were negatively associated with optimal adherence. CONCLUSIONS: Youth were less likely to be optimally adherent throughout the study period. Findings suggest implications for increased youth-centered adherence support, particularly for youth living with HIV concurrently living with Hepatitis C, newly initiating treatment, and going through medication change.

Viral suppression and viral rebound among young adults living with HIV in Canada.

Describe the prevalence and covariates of viral suppression and subsequent rebound among younger (≤29 years old) compared with older adults.A retrospective clinical cohort study; eligibility criteria: documented HIV infection; resident of Canada; 18 years and over; first antiretroviral regimen comprised of at least 3 individual agents on or after January 1, 2000.Viral suppression and rebound were defined by at least 2 consecutive viral load measurements <50 or >50 HIV-1 RNA copies/mL, respectively, at least 30 days apart, in a 1-year period. Time to suppression and rebound were measured using the Kaplan-Meier method and Life Table estimates. Accelerated failure time models were used to determine factors independently associated with suppression and rebound.Younger adults experienced lower prevalence of viral suppression and shorter time to viral rebound compared with older adults. For younger adults, viral suppression was associated with being male and later era of combination antiretroviral initiation (cART) initiation. Viral rebound was associated with a history of injection drug use, Indigenous ancestry, baseline CD4 cell count >200, and initiating cART with a protease inhibitor (PI) containing regimen.The influence of age on viral suppression and rebound was modest for this cohort. Our analysis revealed that key covariates of viral suppression and rebound for young adults in Canada are similar to those of known importance to older adults. Women, people who use injection drugs, and people with Indigenous ancestry could be targeted by future health interventions.

Sexual Self-Efficacy and Gender: A Review of Condom Use and Sexual Negotiation Among Young Men and Women in Sub-Saharan Africa.

Sexual self-efficacy (SSE), one's perceived control of or confidence in the ability to perform a given sexual outcome, predicts sexual behavior; however, important questions remain regarding whether gender modifies observed associations. In a comprehensive review of peer-reviewed HIV-prevention literature focusing on youth (ages 10 to 25) in sub-Saharan Africa, we measured and assessed the influence of SSE on condom use and sexual refusal, overall and by gender. Our results, after reviewing 63 publications, show that SSE is inconsistently measured. Most studies measured condom use self-efficacy (CUSE) (96.8%) and/or sexual refusal self-efficacy (SRSE) (63.5%). On average, young men had higher CUSE than young women, while young women had higher SRSE than young men. While cross-sectional studies reported an association between high SSE and sexual behaviors, this association was not observed in interventions, particularly among young women who face a disproportionate risk of HIV acquisition. In all, 25% of intervention studies demonstrated that fostering CUSE increased condom use among young men only, and one of two studies demonstrated that higher SRSE led to reduced frequency of sexual activity for both men and women. Future research and HIV-prevention interventions must be gender targeted, consider improving CUSE for young men, and move beyond limited individual-level sexual behavior change frameworks.

Factors associated with low neighborhood cohesion among women living with HIV impacted by social-structural inequities in British Columbia.

Built and social environments, including one's perception of their environment, are important determinants of health. The intersection of gender and HIV status may complicate the role of neighborhood cohesion in safety, personal well-being, and health outcomes for populations impacted by social and structural inequities. Among women in particular, social cohesion within the neighborhood they reside in may have a greater influence on health outcomes compared to their male counterparts. We sought to examine perception of neighborhood cohesion (validated scale with a range 0-100, with higher scores indicating higher perceived neighborhood cohesion) among women living with HIV, impacted by social-structural inequities, receiving combination antiretroviral therapy, and enrolled in the Longitudinal Investigations into Supportive Ancillary health services (LISA) study in British Columbia, Canada. Cross-sectional data on neighborhood cohesion and socio-demographic data were collected in an interviewer-administered survey. Of the 1,000 LISA participants interviewed, 908 (including 249 women and 659 men) had complete data for the variables of interest. At the bivariate level, women had worse perceived neighborhood cohesion scores compared to men (median: 56 [95% CI: 44-66] vs. 60 [95% CI: 47-71]). Multivariable model results indicated that for women living with HIV in our sample, greater neighborhood cohesion scores were positively associated with stable housing (ß coefficient = 7.85; 95% CI: 3.61, 12.10, p < 0.001), and negatively associated with greater perceived HIV stigma (ß coefficient = -1.19; 95% CI: -2.24 to-0.15; p = 0.025). The results illustrate the gendered nature of experiencing built and social environments, and highlight the need for women-centred interventions to address the social determinants of HIV burden associated with negative perceptions of neighborhood cohesion.

Gender, Sexual Self-Efficacy and Consistent Condom Use Among Adolescents Living in the HIV Hyper-Endemic Setting of Soweto, South Africa.

Within HIV-endemic settings, few studies have examined gendered associations between sexual self-efficacy (SSE), one's confidence or perceived control over sexual behavior, and uptake of HIV prevention behaviors. Using cross-sectional survey data from 417 sexually-experienced adolescents (aged 14-19, median age = 18, 60% female) in Soweto, South Africa, we measured SSE using a 6-item scale (range:0-6) with 'high-SSE' = score > 3 (study alpha = 0.75). Gender-stratified logistic regression models assessed associations between high-SSE and lifetime consistent condom use. A higher proportion of women reported high-SSE (68.7%) than men (49.5%, p < 0.001). We observed no difference in reported consistent condom use by gender (45.5% among women, 45.8% among men; p = 0.943). In confounder models, high-SSE was associated with consistent condom use among men (aOR = 3.51, 95%CI = 1.86-6.64), but not women (aOR = 1.43, 95%CI = 0.74-2.77). Findings highlight that individual-level psychosocial factors are insufficient for understanding condom use and must be considered alongside the relational, social, and structural environments within which young women navigate their sexual lives.

Conspiracy beliefs and knowledge about HIV origins among adolescents in Soweto, South Africa.

PURPOSE: We examined adolescents' knowledge regarding the origin of HIV/AIDS and correlates of beliefs surrounding conspiracy theories in Soweto, South Africa. Now, a decade post-AIDS denialism, South Africa has the largest antiretroviral therapy roll-out worldwide. However, conspiracy theories stemming from past AIDS denialism may impact HIV prevention and treatment efforts. METHODS: Study participants were recruited through the Kganya Motsha Adolescent Health Centre and the Perinatal HIV Research Unit's Botsha Bophelo Adolescent Health Study (BBAHS). Adolescents were eligible to participate if aged 14-19 years and living in Soweto. We calculated the proportion of adolescents who correctly believed that HIV originated from non-human primates, and used contingency table analysis and logistic regression modeling to describe correlates associated with accurate knowledge and beliefs in conspiracy theories. RESULTS: Of 830 adolescents, 168 (20.2%) participants correctly identified HIV as originating from chimpanzees and one third (n = 71, 8.6%) believed in a conspiracy theory about the origins of HIV, including that it originated from the US government (2.3%), the pharmaceutical industry (2.2%), a vaccine (2.1%), space (1.5%), and a scientist (0.6%). Participants who were more likely to correctly identify the origin of HIV were older, men, and unemployed. Participants who were men, unemployed or students, and who had a parent or close relative who had died of HIV, were more likely to believe in a conspiracy theory regarding the origins of HIV. CONCLUSIONS: Adolescents living in Soweto did not have high levels of accurate knowledge regarding the origins of HIV/AIDS and conspiracy beliefs were present among a small minority of participants. Accurate knowledge of the origins of HIV and debunking myths are important for improving uptake of HIV prevention tools in this population.

The Botsha Bophelo Adolescent Health Study: A profile of adolescents in Soweto, South Africa.

BACKGROUND: Youth between the ages of 15 years to 24 years account for almost half of new HIV infections in South Africa. OBJECTIVES: To describe the study details of the Botsha Bophelo Adolescent Health Study (BBAHS) which was an investigation of HIV risk among adolescents living in Soweto, South Africa. METHODS: Eligibility criteria for the BBAHS included being 14 years - 19 years old and living in one of the 41 identified formal and informal areas in the township of Soweto. A cross-sectional survey was developed between investigators and an adolescent community advisory board consisting of previously validated scales and original questions including demographics, sexual and reproductive health, health service utilisation and psychosocial behaviours. RESULTS: Between 2010 and 2012, interviewers administered surveys among 830 adolescents (57% females), whose median age was 17 years (Q1, Q3: 16, 18), and found that 43% of participants identified their ethnicity as Zulu, 52% reported high food insecurity, 37% reported at least one parent had died, 15% reported living in a shack and 83% identified as heterosexual. Over half of the participants (55%) reported ever having sex (49% of females and 64% of males), 11% of whom initiated sex at < 15 years of age (3% females and 21% males). Almost half (47%) reported ever testing for HIV, 3% (n = 12) of whom self-reported being HIV-positive and 33% (n = 4) were on antiretroviral therapy. CONCLUSION: Our study highlights important individual, relational and structural level determinants of HIV risk for adolescent men and women growing up within HIV hyperendemic settings.

Late initiation of combination antiretroviral therapy in Canada: a call for a national public health strategy to improve engagement in HIV care.

INTRODUCTION: Combination antiretroviral therapy (ART) significantly decreases morbidity, mortality and HIV transmission. We aimed to characterize the timing of ART initiation based on CD4 cell count from 2000 to 2012 and identify factors associated with late initiation of treatment. METHODS: Participants from the Canadian Observational Cohort (CANOC), a multi-site cohort of HIV-positive adults initiating ART naively after 1 January 2000, in three Canadian provinces (British Columbia, Ontario and Québec) were included. Late initiation was defined as a CD4 count <200 cells/mm(3) or an AIDS-defining illness before ART initiation (baseline). Temporal trends were assessed using the Cochran-Armitage test, and independent correlates of late initiation were identified using logistic regression. RESULTS: In total, 8942 participants (18% female) of median age 40 years (Q1-Q3 33-47) were included. The median baseline CD4 count increased from 190 cells/mm(3) (Q1-Q3 80-320) in 2000 to 360 cells/mm(3) (Q1-Q3 220-490) in 2012 (p<0.001). Overall, 4274 participants (48%) initiated ART with a CD4 count <200 cells/mm(3) or AIDS-defining illness. Late initiation was more common among women, non-MSM, older individuals, participants from Ontario and BC (vs. Québec), persons with injection drug use (IDU) history and individuals starting ART in earlier calendar years. In sub-analysis exploring recent (2008 to 2012) predictors using an updated CD4 criterion (<350 cells/mm(3)), IDU and residence in BC (vs. Québec) were no longer significant correlates of late initiation. CONCLUSIONS: This analysis documents increasing baseline CD4 counts over time among Canadians initiating ART. However, CD4 counts at ART initiation remain below contemporary treatment guidelines, highlighting the need for strategies to improve earlier engagement in HIV care.

Vets, denialists and rememberers: social typologies of patient adherence and non-adherence to HAART from the perspective of HIV care providers.

For many people living with HIV/AIDS taking highly active antiretroviral therapy (HAART) is difficult due to various individual and social factors, including the side effects of these medications, HIV/AIDS stigma and poor patient-provider relationships. Most studies that examine barriers to and facilitators of adherence to HAART have been conducted with people on these medications, which is critical to improving adherence among various HIV-affected groups. Less attention has been paid to the experiences of HIV care providers, which is an important gap in the literature considering the key role they play in the delivery of HAART and the management of patient treatment plans. This paper presents findings from a qualitative pilot study that explored how HIV care providers assess adherence and non-adherence to HAART among their HIV-positive patients in Vancouver, British Columbia. Drawing upon individual interviews conducted with HIV physicians (n = 3), social service providers (n = 3) and pharmacists (n = 2), this discussion focuses on the social typologies our participants use to assess patient success and failure related to adherence. Eleven unique categories are featured and the diversity within and across these categories illustrate a broad spectrum of adherence-related behaviours among patients and the social meanings providers attribute to these behaviours. As one of the first explorations of the social typologies used by HIV care providers to assess patient performance on HAART, these data contribute valuable insights into the experiences of providers within the context of adherence-related care delivery.

The impact of unstable housing on emergency department use in a cohort of HIV-positive people in a Canadian setting.

The social-structural challenges experienced by people living with HIV (PHA) have been shown to contribute to increased use of the emergency department (ED). This study identified factors associated with frequent and nonurgent ED use within a cohort of people accessing antiretroviral therapy (ART) in a Canadian setting. Interviewer-administered surveys collected socio-demographic information; clinical variables were obtained through linkages with the provincial drug treatment registry; and ED admission data were abstracted from the Department of Emergency Medicine database. Multivariate logistic regression was used to compute odds of frequent and nonurgent ED use. Unstable housing was independently associated with ED use (adjusted odds ratio [AOR] =1.94, 95% confidence interval [CI] 1.24-3.04]), having three or more ED visits within 6 months of the interview date [AOR: 2.03 (95% CI: 1.07-3.83)] and being triaged as nonurgent (AOR = 2.71, 95% CI: 1.19-6.17). Frequent and nonurgent use of the ED in this setting is associated with conditions requiring interventions at the social-structural level. Supportive housing may contribute to decreased health-care costs and improved health outcomes amongst marginalized PHA.

High rates of lifetime and recent violence observed among harder-to-reach women living with HIV.

This study examined the prevalence and correlates associated with lifetime and recent violence among a sample of harder-to-reach human immunodeficiency virus (HIV)-positive women living in British Columbia (BC), Canada. Interviewer-led surveys with 1000 participants gathered quantitative data on social, economic, and structural characteristics such as housing status, relationship status, family structure, history of drug use, and sexual behavior. Logistic regression analysis modeled associations between independent variables and experiences of violence at baseline. Cox regression analyses with time-dependent covariates determined correlates of lifetime and recent violence among HIV-positive women. Of the 249 women in the study, an overwhelming proportion of women (81%) reported experiences of violence in their lifetime. Among those, 22% reported recent experiences of violence, and 56% of the women reported more than five violent episodes in their lifetime. Lifetime violence was independently associated with HIV-related stigma (adjusted odds ratio [AOR] = 2.18, 95% confidence interval [CI] = 1.14-5.70), previous tobacco use (AOR = 2.79, 95% CI = 1.10-7.07), ever having a drinking problem (AOR = 2.82, 95% CI = 1.28-6.23), and ever having received care for a mental health condition (AOR = 2.42, 95% CI = 1.06-5.52). Recent violence was associated with the current illicit drug use (AOR = 2.60, 95% CI = 1.14-5.90), and currently residing in unstable housing (AOR = 2.75, 95% CI = 1.31-5.78). This study underscores the need to consider potential experiences of historical and current violence as part of comprehensive care for women living with HIV.

Factors associated with late initiation of highly active antiretroviral therapy among young HIV-positive men and women aged 18 to 29 years in Canada.

Initiating highly active antiretroviral therapy (HAART) with low CD4 counts or AIDS-defining illnesses (ADIs) increases risk of treatment failure and death. We examined factors associated with late initiation among 18- to 29-year-olds within the Canadian Observational Cohort (CANOC) collaboration, a multi-site study of HIV-positive persons who initiated HAART after 2000. Late initiation was defined as beginning HAART with a CD4 count <200 cells/mm(3) and/or having a baseline ADI. Multivariable logistic regression was used to identify independent correlates of late initiation. In total, 1026 individuals (422 from British Columbia, 400 from Ontario, and 204 from Quebec) met our age criteria. At HAART initiation, median age was 27 years (interquartile range, 24, 28 years). A total of 412 individuals (40%) identified as late initiators. Late initiation was associated with female gender, age >25 years at initiation, initiating treatment in earlier years, and having higher baseline viral load. The high number of young adults in our cohort starting HAART late indicates important target populations for specialized services, increased testing, and linkages to care.

Implications for HIV prevention: lesbian, gay and bisexual adolescents in urban South Africa are at increased risk of living with HIV.

Among adolescents (14-19 years) in Soweto, South Africa, 14% (4/29) who identified as lesbian, gay or bisexual (3 female and 1 male) reported living with HIV compared with only 2.3% (8/350) who identified as heterosexual (P = 0.002). These data indicate an urgent need to utilize evidence-based strategies to support access to HIV prevention education and services for lesbian, gay and bisexual youth.

Women-specific HIV/AIDS services: identifying and defining the components of holistic service delivery for women living with HIV/AIDS.

INTRODUCTION: The increasing proportion of women living with HIV has evoked calls for tailored services that respond to women's specific needs. The objective of this investigation was to explore the concept of women-specific HIV/AIDS services to identify and define what key elements underlie this approach to care. METHODS: A comprehensive review was conducted using online databases (CSA Social Service Abstracts, OvidSP, Proquest, Psycinfo, PubMed, CINAHL), augmented with a search for grey literature. In total, 84 articles were retrieved and 30 were included for a full review. Of these 30, 15 were specific to HIV/AIDS, 11 for mental health and addictions and four stemmed from other disciplines. RESULTS AND DISCUSSION: The review demonstrated the absence of a consensual definition of women-specific HIV/AIDS services in the literature. We distilled this concept into its defining features and 12 additional dimensions (1) creating an atmosphere of safety, respect and acceptance; (2) facilitating communication and interaction among peers; (3) involving women in the planning, delivery and evaluation of services; (4) providing self-determination opportunities; (5) providing tailored programming for women; (6) facilitating meaningful access to care through the provision of social and supportive services; (7) facilitating access to women-specific and culturally sensitive information; (8) considering family as the unit of intervention; (9) providing multidisciplinary integration and coordination of a comprehensive array of services; (10) meeting women "where they are"; (11) providing gender-, culture- and HIV-sensitive training to health and social care providers; and (12) conducting gendered HIV/AIDS research. CONCLUSIONS: This review highlights that the concept of women-specific HIV/AIDS services is a complex and multidimensional one that has been shaped by diverse theoretical perspectives. Further research is needed to better understand this emerging concept and ultimately assess the effectiveness of women-specific services on HIV-positive women's health outcomes.

Social-structural factors associated with supportive service use among a cohort of HIV-positive individuals on antiretroviral therapy.

As mortality rates decrease in British Columbia, Canada, supportive services (e.g. housing, food, counseling, addiction treatment) are increasingly conceptualized as critical components of care for people living with HIV/AIDS. Our study investigates social and clinical correlates of supportive service use across differing levels of engagement. Among 915 participants from the Longitudinal Investigations into Supportive and Ancillary health services (LISA) cohort, 742 (81%) reported using supportive services. Participants were nearly twice as likely to engage daily in supportive services if they self-identified as straight (95% confidence interval [CI], adjusted odds ratio [AOR]: 1.69), had not completed high school (95% CI, AOR: 1.97), had an annual income of < $15,000 (95% CI, AOR: 1.81), were unstably housed (95% CI, AOR: 1.89), were currently using illicit drugs (95% CI, AOR: 1.60), or reported poor social capital in terms of perceived neighborhood problems (95% CI, AOR: 1.15) or standard of living (95% CI, AOR: 1.70). Of interest, after adjusting for sociodemographic and socioeconomic variables, no clinical markers remained an independent predictor of use of supportive services. High service use by those demonstrating social and clinical vulnerabilities reaffirms the need for continued expansion of supportive services to facilitate a more equitable distribution of health among persons living with HIV.

Identifying self-perceived HIV-related stigma in a population accessing antiretroviral therapy.

This study identifies factors associated with self-perceived HIV-related stigma (stigma) among a cohort of individuals accessing antiretroviral therapy in British Columbia, Canada. Data were drawn from the Longitudinal Investigations into Supportive and Ancillary Health Services study, which collects social, clinical, and quality of life (QoL) information through an interviewer-administered survey. Clinical variables (i.e., CD4 count) were obtained through linkages with the British Columbia HIV/AIDS Drug Treatment Program. Multivariable linear regression was performed to determine the independent predictors of stigma. Our results indicate that among participants with high school education or greater the outcome stigma was associated with a 3.05 stigma unit decrease (95% CI: -5.16, -0.93). Having higher relative standard of living and perceiving greater neighborhood cohesion were also associated with a decrease in stigma (-5.30 95% CI: -8.16, -2.44; -0.80 95% CI: -1.39, -0.21, respectively). Lower levels of stigma were found to be associated with better QoL measures, including perceiving better overall function (-0.90 95% CI: -1.47, -0.34), having fewer health worries (-2.11 95% CI: -2.65, -1.57), having fewer financial worries (-0.67 95% CI: -1.12, -0.23), and having less HIV disclosure concerns (-4.12 95% CI: -4.63, -3.62). The results of this study show that participants with higher education level, better QoL measures, and higher self-reported standards of living are less likely to perceive HIV-related stigma.

Cohort Profile: Longitudinal Investigations into Supportive and Ancillary health services.

The Longitudinal Investigations into Supportive and Ancillary health services (LISA) study is a cohort of people living with HIV/AIDS who have ever accessed anti-retroviral therapy (ART) in British Columbia, Canada. The LISA study was developed to better understand the outcomes of people living with HIV with respect to supportive services use, socio-demographic factors and quality of life. Between July 2007 and January 2010, 1000 participants completed an interviewer-administered questionnaire that included questions concerning medical history, substance use, social and medical support services, food and housing security and other social determinants of health characteristics. Of the 1000 participants, 917 were successfully linked to longitudinal clinical data through the provincial Drug Treatment Program. Within the LISA cohort, 27% of the participants are female, the median age is 39 years and 32% identify as Aboriginal. Knowledge translation activities for LISA include the creation of plain language summaries, internet resources and arts-based engagement activities such as Photovoice.

Gender differences in clinical outcomes among HIV-positive individuals on antiretroviral therapy in Canada: a multisite cohort study.

BACKGROUND: Cohort data examining differences by gender in clinical responses to combination antiretroviral therapy (ART) remain inconsistent and have yet to be explored in a multi-province Canadian setting. This study investigates gender differences by injection drug use (IDU) history in virologic responses to ART and mortality. METHODS: Data from the Canadian Observational Cohort (CANOC) collaboration, a multisite cohort study of HIV-positive individuals initiating ART after January 1, 2000, were included. This analysis was restricted to participants with a follow-up HIV-RNA plasma viral load measure and known IDU history. Weibull hazard regression evaluated time to virologic suppression (2 consecutive measures <50 copies/mL), rebound (>1000 copies/mL after suppression), and all-cause mortality. Sensitivity analyses explored the impact of presumed ART use in pregnancy on virologic outcomes. RESULTS: At baseline, women (1120 of 5442 participants) were younger (median 36 vs. 41 years) and more frequently reported IDU history (43.5% vs. 28.8%) (both p<0.001). Irrespective of IDU history, in adjusted multivariable analyses women were significantly less likely to virologically suppress after ART initiation and were at increased risk of viral load rebound. In adjusted time to death analysis, no differences by gender were noted. After adjusting for presumed ART use in pregnancy, observed gender differences in time to virologic suppression for non-IDU, and time to virologic rebound for IDU, became insignificant. CONCLUSIONS: HIV-positive women in CANOC are at heightened risk for poor clinical outcomes. Further understanding of the intersections between gender and other factors augmenting risk is needed to maximize the benefits of ART.