Estimating the Cost of Rapid Home HIV testing within a Randomized Controlled Trial in the United States

Background: Approximately 1.1 million people in the U.S. are living with HIV, and 15% are unware of their infection. HIV self-testing programs could increase HIV testing and new diagnoses among people at high-risk of HIV infection who may not access conventional testing services; however, data on the cost of such programs are limited. Using data from a randomized controlled trial (RCT) of rapid home HIV self-testing (eSTAMP: Evaluation of Rapid HIV Self-Testing among MSM Project), we conducted a cost analysis of the trial.

Methods: The study enrolled Internet-recruited men who have sex with men (MSM), who reported unknown or negative HIV status (NCT02067039). All participants were provided information on local HIV testing services in the community. Each participant assigned to the intervention arm initially received 4 rapid self-tests: 2 oral fluid and 2 finger-stick tests. Participants had the option to use either type of test, distribute tests to others, and request replacement tests every 3 months. Participants reporting positive tests results were offered confirmatory testing. Follow-up surveys were conducted quarterly for 12 months to collect testing outcomes. We conducted the cost analysis based on trial expenditure data reported to CDC from March 2015 to December 2016. We estimated total trial expenditures, including trial development and implementation expenditures. The trial development expenditures primarily included design and production of home test kit packaging, Internet site design, Internet system certification and monitoring, and online advertising and recruitment. The trial implementation expenditures included trial administration, HIV test kits purchase and delivery, facility overhead and contracting fees, time for staff to provide linkage to care, and incentives for participants.

Results: A total of 2,665 men were randomly assigned either to the intervention arm (n=1,325) or the control arm (n=1,340). HIV testing during the study period was reported by 969 (73%) intervention arm participants and 358 (27%) control arm participants. Among those who reported testing, intervention arm participants took a total of 5,136 tests (5.3 tests/person), and 97% of the participants used a study HIV test. Control arm participants took 537 tests (1.5 tests/person). Testing during the RCT identified 22 new HIV diagnoses in the intervention arm and 11 new diagnoses in the control arm (IAS, 2017). The total trial expenditure was $2.64 million, of which $0.97 million (37%) was spent on trial development. Of $1.67 million spent on the trial implementation, $0.95 million (57%) was for the intervention arm. The average implementation expenditure per intervention arm participant completing at least one study test was $1,015, per study test conducted was $226, and per new HIV diagnosis reported was $43,198.

Conclusion: This trial showed high acceptance of self-testing, greater willingness to test multiple times, and a doubling of the number of new HIV diagnoses, compared with the control arm. The cost per new diagnosis falls within what has been considered cost-effective in other publications.