Estimating the Cost-Effectiveness of HIV Self-Testing in the United States Using Net Benefit Regression

Cost-effectiveness analysis of HIV self-testing using patient-level data from a randomized clinical trial can inform HIV prevention funding decisions. Cost-effectiveness analysis using net-benefit regression addresses the sampling uncertainty in the trial data and the variability of policymakers' willingness to pay (WTP).

We used published data from a 12-month longitudinal randomized clinical trial that enrolled 2665 men who had sex with men randomly assigned to the self-testing arm (participants receiving self-test kits) and control arm (participants receiving standard-of-care), and the self-testing arm identified 48 additional new HIV cases. We used net-benefit regression to investigate the cost-effectiveness of an HIV self-testing intervention, which compared the incremental cost per new HIV diagnosis with policymakers' WTP thresholds. We addressed the uncertainties in estimating the incremental cost and the policymakers' WTP per new diagnosis through the incremental net-benefit (INB) regression and cost-effectiveness acceptability curve (CEAC) analyses.

From the health care provider's perspective, the INB analysis showed a positive net benefit of HIV self-testing compared with standard-of-care when policymakers' WTP per new HIV diagnosis was $9365 (95% confidence interval: $5700 to $25,500) or higher. The CEAC showed that the probability of HIV self-testing being cost-effective compared with standard-of-care was 58% and >99% at a WTP of $10 000 and $50 000 per new HIV diagnosis, respectively.

The INB and CEAC analyses suggest that HIV self-testing has the potential to be cost-effective for relatively low values of policymakers' WTP.