CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

for HIVST access attributes (price, location, packaging and usage instructions). Latent class conditional logit regression explored variability in preferences among infrequent testers (tested >2 years ago or never tested), recent migrants (arriving in Australia <5 years), students, age and multiple partners in the last 6 months (i.e. more than one regular or casual partner). Random parameters logit model explored the most influential attributes on an individual’s choice (Figure 1). Results: Overall, 727 men participated in DCE1 and 275 men participated in DCE2. DCE1 contained four classes of men: ‘recent migrants’ (prefer fast results and cheaper tests, class size 23%); ‘Australian-born men with multiple partners who were frequent testers’ (prefer tests with shorter window periods and finger-prick HIVST, class size 33%). and ‘students’ (prefer fast results and oral HIVST, class size 28%). There were no significant differences in where to access HIVST according to age, number of partners in the last 6 months, recent migrant or student. There were three classes of men: ‘price matters’ (prefer purchasing kits online or off-the-shelf from pharmacies, class size 48%), ‘infrequent testers’ (prefer purchasing kits online and vending machines, class size 35%) and ‘frequent testers’ (disliked purchasing online and prefer purchasing off-the shelf from pharmacies or staff frommedical clinics, class size 17%). Conclusion: Quantifying preference heterogeneity in HIV testing among subpopulations of GBMmay be useful for tailoring public health messages and informing HIVST distribution methods.

and willingness were reported by 6,578(35%) and 7,609(40%), respectively. In the multivariable model willingness to use HIVST was associated with: being from Brazil compared to Peru (AOR1.74[IC95%1.55-1.94]); being 18-24 (AOR1.19[IC95%1.08-1.31]) or 25-35 years of age (AOR1.26[IC95%1.17-1.37]) compared to ≥35; >secondary education (AOR1.33[IC95%1.24-1.44]) compared to ≤secondary education; middle (AOR1.42[IC95%1.32-1.52]) and higher income (AOR2.10[IC95%1.90-2.31]) compared to low; daily use of GSN apps (AOR1.11[IC95%1.32-1.52]); willingness to use PrEP (AOR1.44[IC95%1.35-1.54]); and recent sex under the influence of drugs (AOR1.08[IC95%1.01-1.16]). Testing- related variables (lifetime HIV testing, HIVST awareness and HIVST barriers) were also associated with HIVST in the same model (Table). Conclusion: Willingness to use HIVST was low among MSM from Brazil, Mexico and Peru. Efforts to increase HIVST knowledge and resolve perceived barriers are warranted. HIVST delivery platforms could be incorporated to PrEP implementation programs in these countries.

Poster Abstracts

937 SEXUAL RISK AND HIV TESTING DISCONNECT IN MSM RECRUITED TO AN HIV SELF-TESTING TRIAL Alison Rodger 1 , David Dunn 1 , Andrew N. Phillips 1 , Leanne McCabe 1 , Peter Weatherburn 2 , Fiona Lampe 1 , T Charles Witzel 2 , Fiona Burns 1 , Denise Ward 1 , Roger Pebody 3 , Roy Trevelion 4 , Yolanda Collaco-Moraes 1 , Sheena McCormack 1 1 University College London, London, UK, 2 London School of Hygiene & Tropical Medicine, London, UK, 3 NAM, London, UK, 4 HIV i-Base, London, UK Background: High levels of HIV testing in men who have sex with men (MSM) remain key to reducing HIV incidence. Current levels of testing remain sub- optimal; particularly the more frequent testing recommendations for men at higher risk of HIV through recent condomless anal intercourse (CAI). We report frequency of previous HIV testing at baseline in MSMwho opted to enroll in a HIV self-testing (HIVST) RCT (SELPHI). Methods: SELPHI is an internet based, open-label, randomised controlled trial, which aims to assess effectiveness of free HIVST kits on HIV diagnosis rates. Criteria for enrolment were aged ≥16 years old, male (including trans man) or trans women, ever had anal intercourse (AI) with a man, not known to be HIV positive and provided consent to link to national HIV surveillance databases. Participants were randomly allocated 3:2 at enrolment to a free HIVST versus no free HIV self-test. Data collected via an online survey included socio- demographics (gender, sexual identity, education, age, ethnicity, UK birth), sexual behaviour, HIV/STI testing history and PrEP and PEP use. Results: 10,224 men were randomised; median age 33 years (IQR 26-44); 89% white; 20% born outside UK; <1% trans men; 47% degree educated; 8% ever used PrEP; 4% currently using PrEP. In the previous 3 months, 89% reported AI and 72% reported CAI with ≥1 male partner. Overall, 17%, 33%, 54%, and 72% had tested for HIV in the last 3 months, 6 months, 12 months, and 2 years respectively; 13% had tested more than 2 years ago and 15% had never tested. An association was observed between number of recent CAI partners and time since last HIV test (Figure 1). Among 3,972 men reporting ≥2 recent CAI partners, only 22% had tested in last 3 months and only 41% in last 6 months. In multivariate logistic regression analysis of selected sociodemographic factors, higher education level, being born outside UK and age between 20-40 years were independently associated with a higher likelihood of a recent HIV test. Only 9/388 men currently using PrEP had never tested for HIV. Conclusion: MSM in SELPHI were not testing in line with current UK recommendations. Other SELPHI data suggest this is due to low perceived risk of HIV infection; structural barriers impacting testing opportunities (clinics difficult to access because of time constraints or capacity issues) and individual

936 WILLINGNESS TO USE HIV SELF-TESTING AMONG MSM FROM BRAZIL, MEXICO, AND PERU Thiago S. Torres 1 , Kelika A. Konda 2 , E. Hamid Vega-Ramirez 3 , Oliver A. Elorreaga-Reyes 2 , Dulce Diaz-Sosa 3 , Cristina Pimenta 4 , Brenda Hoagland 1 , Hugo López-Gatell 5 , Rebeca Robles-Garcia 6 , Steven D. Diaz 3 , Beatriz Grinsztejn 1 , Carlos Carceres 2 , Valdilea Veloso 1 , for the ImPrEP Study Group 1 Instituto Nacional de Infectologia Evandro Chagas, Rio de Janeiro. Brazil, 2 Universidad Peruana Cayetano Heredia, Lima, Peru, 3 Clinica Especializada Condesa, Mexico City, Mexico, 4 Ministry of Health, Brasilia (DF), Brazil, 5 Instituto Nacional de Salud Pública, Mexico City, Mexico, 6 Ramón de la Fuente Muñiz National Institute of Psychiatry, Mexico City, Mexico Background: HIV self-testing (HIVST), an important tool within the combined HIV prevention package, is commercially available in Latin America since 2015. This study aims to describe factors associated with willingness to use HIVST among men who have sex with men (MSM) from Brazil, Mexico and Peru. Methods: MSM were reached via advertisements on Facebook and GSN apps for sexual encounters (Grindr and Hornet) fromMarch-May 2018. Participants were cisgender men, ≥18 years old and HIV negative. Willingness to use HIVST was defined as selecting the highest option on a five-point Likert scale. Factors associated with HIVST willingness were assessed using a logistic multivariable model. Results: A total of 43,687 MSM started the questionnaire; 8,790(20%) were ineligible and 18,916(43%) completed and were included in this analysis. Median age was 28 years (IQR: 24-34). Most were from Brazil (59%), followed by Mexico (30%) and Peru (11%). The majority reported low (39%) or middle (43%) monthly income, and 32%≤ secondary education. Recruitment was primarily (85%) via GSN apps; 46% of MSM reported daily use of these apps. Although 53% scored ≥10 points on the HIV Incidence Risk for MSM scale indicating high risk, 65% had low perceived risk of getting HIV in the next year. A total of 3,715(20%) had never tested for HIV, mostly due to fear of a positive result (28%), low self-perceived risk (21%) and shame (24%). HIVST awareness

CROI 2019 366