Results From a Pilot Study of an Automated Directly Observed Therapy Intervention Using Artificial Intelligence With Conditional Economic Incentives Among Young Adults With HIV

Abstract

Background: Despite improvements in antiretroviral therapy (ART) availability, suboptimal adherence is common among youth with HIV (YWH) and can increase drug resistance and poor clinical outcomes. Our study examined an innovative mobile app-based intervention that used automated directly observed therapy (aDOT) using artificial intelligence, along with conditional economic incentives (CEIs) to improve ART adherence and enhance viral suppression among YWH. Setting: We conducted a pilot study of the aDOT-CEI intervention, informed by the operant framework of Key Principles in Contingency Management Implementation, to improve ART adherence among YWH (18–29) in California and Florida who had an unsuppressed HIV viral load. Methods: We recruited 28 virally unsuppressed YWH from AIDS Healthcare Foundation clinics, who used the aDOT platform for 3 months. Study outcomes included feasibility and acceptability, self-reported ART adherence, and HIV viral load. Results: Participants reported high satisfaction with the app (91%), and 82% said that it helped them take their medication. Comfort with the security and privacy of the app was moderate (55%), and 59% indicated the incentives helped improve daily adherence. Conclusions: Acceptability and feasibility of the aDOT-CEI intervention were high with potential to improve viral suppression, although some a priori metrics were not met. Pilot results suggest refinements which may improve intervention outcomes, including increased incentive amounts, provision of additional information, and reassurance about app privacy and security. Additional research is recommended to test the efficacy of the aDOT-CEI intervention to improve viral suppression in a larger sample.