How USAID is tapping into the power of AI to advance HIV solutions

In our modern world where data flows abundantly, artificial intelligence (AI) encompasses a broad range of technologies to process data and demonstrates significant potential to transform multiple sectors, including health care. 

USAID, through PEPFAR, helps partner countries to gain new and deeper insights into their HIV epidemics, including identifying and predicting transmission patterns, treatment outcomes, and program effectiveness. This new information will help countries make informed decisions, allocate resources efficiently, and implement targeted interventions to end HIV as a public health threat by 2030.

Here’s how:

Transforming patient care

At the country level, HIV programs strive to retain all individuals diagnosed with HIV on daily antiretroviral treatment, which is critical to achieving viral load suppression. Through the use of AI, the world is entering a new era of personalized treatment where tailored care plans—guided by data insights, predictive algorithms, and patient privacy protections—maximize treatment outcomes and mitigate stigma and discrimination. 

For example, in Mozambique and Nigeria, USAID supports partner governments to use predictive analytics and machine learning to improve the identification of people on HIV treatment who are at greatest risk of treatment interruption. This includes examining patterns of appointment visits, sociodemographic factors, and parts of the clinical history that can interfere with consistent treatment engagement and adherence at PEPFAR-funded, USAID-supported facilities. Early identification of potential interruption in treatment enables health providers to tailor support interventions to reduce barriers to care and prevent people from falling behind or experiencing an interruption in treatment.

Building more resilient health systems

In support of our country partners, USAID is spearheading the integration of AI and data science to build more resilient and patient-centered health systems. Using these technologies, coupled with privacy protections, for disease surveillance and early detection allows for swift responses to local outbreaks and potential pandemics. 

USAID also applied AI techniques to estimate the size of the population of adolescent girls and young women at risk of acquiring HIV in Uganda. These data are helping Determined, Resilient, Empowered, AIDS-free, Mentored, and Safe (DREAMS) programs understand how to better reach young women for program enrollment, use resources efficiently to attain maximum impact, and monitor its reach.

Optimizing program management 

AI-powered data analytics provide invaluable insights into where diseases are most prevalent and where health care is needed most. Predictive models can be used to forecast health care demand to optimize program management. Notably, USAID and partners use analytics modeling to enhance these efforts and track progress toward HIV epidemic control.

For example, USAID supported the development of a tool that uses machine learning to   automate the routine monitoring of data reported by thousands of PEPFAR-supported health facilities. With a high level of accuracy, the tool identifies health facilities that are reporting atypical data, enabling targeted interventions to improve data and program quality. 

USAID also supported the South African National Department of Health to develop and deploy a machine learning-based model to deduplicate more than 20 million patient records. A full index of every record in the database would have considered 200 trillion record pairs, and any deduplication algorithm would have taken years to run. But, the machine learning model reduced the number of record pairs to millions, and the algorithm ran in a matter of days. The deduplicated database of clinic records provides a true accounting of the statistics of the epidemic in a locale and enhances understanding of what services should be provided to meet the needs of patients. 

The future of AI and the HIV response

AI, when used responsibly and with robust rules for privacy and security, presents unprecedented opportunities to improve patient care, bolsters health outcomes, and builds more resilient health systems globally. 

Through collaborative efforts and responsible use of these technologies, USAID is finding ways to support partner countries' use of AI to efficiently reach our collective goal of ending HIV as a public health threat. 

--

ABOUT THE AUTHOR: Melaku Dessie is the Acting Chief for the Strategic Information, Evaluation, and Informatics Division of the USAID Office of HIV/AIDS. Cassie Vasiloff is a Communications Advisor for the USAID Office of HIV/AIDS