1275. Will an App-optimized HIV Self-testing Strategy Work for South Africans? Results from a Large Cohort Study.
Session: Poster Abstract Session: HIV: Diagnosis and Screening
Friday, October 5, 2018
Room: S Poster Hall

HIV self-testing (HIVST) offers a potential for expanded test access; challenges remain in operationalizing rapid personalized linkages and referrals to care. We investigated if an app-optimized personalized HIVST strategy improved referrals, detected new infections and expedited linkages to care and treatment.


In an ongoing cohort study (n=2000) based in South Africa, from Nov 2016-Jan 2018, to participants presenting to self-test at community township based clinics, we offered a choice of the following strategies: a) unsupervised HIVST; b) supervised HIVST. We also observed participants opting for conventional HIV testing (ConvHT) in geographically separated clinics. We observed outcomes (i.e., linkage initiation, referrals, disease detection) and compared it between the two (HIVST vs. ConvHT) for the same duration.


Of 2000 participants, 1000 participants were on HIVST, 599 (59.9%) chose unsupervised HIVST, 401 (40.1%) on supervised HIVST; compared to 1000 participants on ConvHT.

Participants in HIVST vs. ConvHT were comparable- young (mean age 27.7 (SD=9.0) vs. 29.5 (SD=8.4)); female (64.0% vs. 74.7%); poor monthly income<3,000 RAND ($253 USD) (79.9% vs. 76.4 %).

With HIV ST (vs. ConvHT), many more referrals (17.4% (15.1-19.9) vs. 2.6% (1.7-3.8); RR 6.69 [95% CI: 4.47-10.01]), and many new infections (86 (8.6% (6.9-10.5)) vs. 57 (5.7 % (4.3-7.3)); Odds Ratio 1.55 [95% CI 1.1-2.2]) were noted. Break up: 45 infections in supervised HIVST 45 (52.3%); 41 infections in unsupervised HIVST (47.6 %)]. Preference for HIVST was at 91.6%.

With an app-optimized HIVST strategy, linkages to care were operationalized within a day in all participants (99.7% (HIVST) vs. 99.2% (ConvHT); RR 1.005 [95% CI: 0.99-1.01]); 99.8% supervised HIVST, 99.7% unsupervised HIVST.


Our app-optimized HIVST strategy successfully increased test referrals, detected new infections, and operationalized linkages within a day. This innovative, patient preferred strategy holds promise for a global scale up in digitally literate populations worldwide.

Nitika Pant Pai, MD, MPH, PhD1, Ali Esmail, MD2, Gayatri Marathe, MSc1, Suzette Oelofse, MD2, Marietjie Pretorius, MSc2, Megan Smallwood, MSc1, Jana Daher, MSc (PH)1, Ricky Janssen, MSc3, Paramita Saha-Chaudhuri, PhD4, Nora Engel, PhD3 and Keertan Dheda, MBBCh, PhD2, (1)Division of Clinical Epidemiology, McGill University & McGill University Health Centre, Montreal, QC, Canada, (2)Division of Pulmonology, Department of Medicine, UCT Lung Institute, Lung Infection and Immunity Unit, Cape Town, South Africa, (3)Department of Health, Ethics & Society, Maastricht University, Maastricht, Netherlands, (4)Department of Epidemiology, Biostatistics & Occupational Health, McGill University, Montreal, QC, Canada


N. P. Pai, None

A. Esmail, None

G. Marathe, None

S. Oelofse, None

M. Pretorius, None

M. Smallwood, None

J. Daher, None

R. Janssen, None

P. Saha-Chaudhuri, None

N. Engel, None

K. Dheda, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.