MOPDB0101
Share
 
Title
Increased levels of current alcohol use are associated with worse HIV care cascade outcomes among HIV-positive adults in rural Kenya and Uganda in the SEARCH trial
Presenter
Sarah Puryear
Authors
S. Puryear1, J. Ayieko2, D. Kwarisiima3, J.A. Hahn1, L. Balzer4, E. Charlebois1, T. Clark1, C.R. Cohen1, E.A. Bukusi2, M. Kamya3,5, M. Petersen6, D. Havlir1, G. Chamie1
Institutions
1University of California - San Francisco, San Francisco, United States, 2Kenya Medical Research Institute (KEMRI), Nairobi, Kenya, 3Infectious Diseases Research Collaboration, Kampala, Uganda, 4University of Massachusetts, Amherst, Biostatistics & Epidemiology, Amherst, United States, 5Makerere University, Kampala, Uganda, 6University of California Berkeley, School of Public Health, Berkeley, United States

Background: Alcohol use is common and associated with poor clinical outcomes among PLHIV. However, there are limited data on the impact of alcohol use across the HIV care cascade in sub-Saharan Africa.
Methods: SEARCH is a cluster-randomized HIV “test-and-treat” trial in 32 rural Kenyan and Ugandan communities. We evaluated baseline (2013-14) HIV care cascade outcomes and alcohol use by AUDIT-C. “Alcohol use” included any current use (AUDIT-C>0) and was stratified by level: never (0), low (1-3 men/1-2 women), medium (4-5 men/3-5 women), high (6-7), very high (8-12). Baseline population-wide HIV testing identified 13,991 HIV-positive adults (≥ 15 years): 11,396 (82%) completed alcohol screening. Logistic regression evaluated associations between alcohol use and cascade metrics (HIV diagnosis, ART uptake, viral suppression), adjusting for sex, age, mobility, marriage, education, occupation, wealth, and community clustering.
Results: Among 11,396 HIV+ adults, 1828 (16%) reported alcohol use: 7% of women (514/7,302); 32% of men (1314/4,094). Levels of drinking were low (30%), medium (15%), high (19%), and very high (36%). Drinkers were significantly less likely to know their HIV-positive status (44% [95%CI: 39-49%]) than non-drinkers (61% [95%CI: 55-66%]) and to be virally suppressed (37% [95%CI: 32-41%]) than non-drinkers (51% [95%CI: 48-54%]), findings noted at every level of drinking (Figure). In multivariate analyses comparing to non-drinkers, alcohol users were significantly less likely to know their HIV status (aOR 0.63, 95%CI: 0.55-0.72). If diagnosed, drinkers were less likely to be on ART (aOR 0.53, 95%CI: 0.44-0.63). Among those on ART, there was no significant association between alcohol and viral suppression. Overall, alcohol use was associated with significantly lower odds of viral suppression (aOR 0.56, 95%CI: 0.47-0.67).
Conclusions: Current alcohol use was associated with lower viral suppression: results suggest this may be due to decreased HIV diagnosis and ART use. Tailored interventions for individuals who use alcohol may be needed to optimize cascade outcomes.


Percent of HIV-positive adults achieving HIV care cascade targets at baseline by alcohol use level
[ Percent of HIV-positive adults achieving HIV care cascade targets at baseline by alcohol use level]