CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

900 COMPARING METHODS FOR ESTIMATING SEXUAL TRANSMISSION RISK AMONG US ADULTS WITH HIV Yunfeng Tie 1 , Sharoda Dasgupta 1 , Linda Beer 1 , Jennifer Fagan 1 , Qian An 1 , R. L. Shouse 1 1 CDC, Atlanta, GA, USA Background: An accurate assessment of sexual risk behaviors associated with HIV transmission is important for informing the Ending the HIV Epidemic (EHE) initiative. The Medical Monitoring Project (MMP) produces nationally representative estimates for high-risk sex using information from all past year partnerships. To potentially reduce data collection burden, we explored whether most recent sexual partner data was enough to accurately assess high- risk sex among adults with diagnosed HIV. Methods: MMP staff interviewed adults with diagnosed HIV to collect information on demographic characteristics and sexual behaviors with the last 5 partnerships during the past 12 months (P12M); for those with >5 partners, aggregated information on sexual behaviors with additional partners was also collected. Viral load results were abstracted frommedical records. Using weighted data collected 6/2015−5/2018 (n=11,914), we estimated the prevalence of high-risk sex, defined as 1) having ≥1 detectable viral load (≥200 copies/mL) over P12M and 2) having condomless anal or vaginal sex with an HIV-negative or HIV-unknown partner not reported to be using pre-exposure prophylaxis (PrEP). We reported the incremental contributions of each sexual partner to the measure, and compared prevalence of high-risk sex overall and by age, race/ethnicity, and sexual behaviors using data from the most recent partner compared with all partners. Results: Of adults with diagnosed HIV, 58% had anal or vaginal sex in P12M, of whom 44% reported >1 partner and 12% reported >5 partners. A higher percentage of men who had sex with men (MSM), whites, and people aged 18-29 reported having multiple partners. The prevalence of high-risk sex was 6% overall, 11% among MSM, 13% among women who had sex with men, and 15% among persons aged 18-29. Estimates of high-risk sex were lower when information of the last partner only vs. all partners was assessed (4% using last partner vs. 6% using all partners), particularly for MSM (6% vs. 11%), persons aged 18-29 (9% vs. 15%), and Hispanics/Latinos (3% vs. 6%) (Figure). Conclusion: Estimates of high-risk sex using last partnership were not adequate to accurately describe the prevalence of HIV transmission risk— particularly for groups highly affected by HIV, such as MSM and young adults. Using information on all sexual partners may be helpful to identify key populations in need of additional support for HIV prevention and can help inform EHE initiative activities.

899 MORTALITY IN PEOPLE LIVING WITH HIV AND MENTAL HEALTH DISORDERS IN SOUTH AFRICA

Andreas D. Haas 1 , Yann Ruffieux 1 , Ernest Mokotoane 2 , Johannes P. Mouton 2 , Mpho Tlali 2 , Mary-Ann Davies 2 , Matthias Egger 1 , Gary Maartens 2 , for the IeDEA Southern Africa Collaboration 1 University of Bern, Bern, Switzerland, 2 University of Cape Town, Cape Town, South Africa Background: People with mental and substance use disorders (MSD) often die prematurely from suicide, accidents or chronic comorbidities. We quantified mortality from natural and external causes in people living with HIV and MSD enrolled in the Aid for AIDS (AfA) program in South Africa. Methods: AfA is a large South African private sector HIV management program. AfA collects demographic, clinical and laboratory data. We linked the AfA data with mortality and cause of death information (natural vs. external cause) from the South African national population registry and with ICD-10 diagnoses from hospitalization records covering the period 2011-2018. We left-truncated ART records in 2011. HIV+ children and adults who initiated cART from 2001-2018 were followed for up to 15 years on ART. We estimated cumulative mortality using the Kaplan-Meier method. We calculated adjusted hazard ratios (aHR) for associations between MSD and mortality using Cox regression. HR were adjusted for age, gender, CD4 count at ART initiation and year of ART initiation. Results: Out of 219,686 individuals who initiated ART, 9,527 (4.3%) were admitted for an MSD for a median duration of 7 days (IQR 4-14). The cumulative mortality from natural and external causes 15 year after ART initiation was 15.5% (95%CI 14.9-16.1) and 2.3% (CI 2.1-2.6), respectively. The Figure shows aHRs and 95% CIs comparing mortality in ART patients with and without MSD. AHRs for mortality from natural causes were 3.65 (CI 3.33-4.01) for people with mental disorders and 2.27 (CI 1.61-3.20) for people with substance use disorders. AHRs for mortality from external causes were 2.13 (CI 1.57-2.89) for people with mental disorders and 3.79 (CI 2.18-6.59) for people with substance use disorders. Individuals with mental disorders due to organic causes (e.g. dementia) had the largest increase in risk of mortality from natural 13.52 (CI 11.57-15.80) and external 7.04 (CI 3.83-12.94) cause. The risk of mortality from natural causes was about four times higher for people with psychotic, anxiety, other psychiatric disorder, or drug use disorder, and about double for people with mood or alcohol use disorders, compared to people without those disorders. Conclusion: Excess mortality of people with MSD is a major public health concern that warrants action. Differentiated care models that account for the special needs of people living with HIV and MSD might be a promising approach to reduce excess mortality in this vulnerable population.

Poster Abstracts

CROI 2020 337