In the heart of India, a quiet revolution is taking place, driven by the combination of artificial intelligence and the simple mobile phone. Dr. Aparna Hegde, a dedicated advocate for maternal health and the Founder, Chairperson, and Managing Trustee of ARMMAN, has partnered with Dr. Aparna Taneja, Research Lead at Google DeepMind India, to address one of India's most urgent issues: minimizing preventable deaths during pregnancy and childbirth.
Their tool of choice? An innovative AI-powered system designed to predict which expectant mothers are at the greatest risk of disengaging from vital healthcare information programs. This allows ARMMAN, Dr. Hegde's non-profit organization, to deliver timely and targeted support.
We discovered that nearly 30%-40% of women were not consistently engaging with the calls; they either weren't answering or weren't fully listening. With limited staff, it was challenging to reach all these women. We needed a way to identify those at high risk of dropping out and prioritize our outreach efforts.
I recognized that AI could guide us in the right direction. With pro bono support from Google’s research team, we developed a model that helped us retain about 30% of new and expectant mothers who were at the highest risk of disengagement from the program.
We've observed that when women listen to the information, the health outcomes are remarkable. We can reach out to more women each week, re-engage them, and ultimately save lives thanks to AI.
By analyzing millions of anonymized call records, the program enables ARMMAN to identify participants who are most at risk of disengagement, allowing us to prioritize them for additional personalized outreach by our call center staff and community partners, including in-person support and resources for specific concerns.
The model was developed in accordance with Responsible AI principles and underwent thorough review by multiple ethics boards.
At ARMMAN, we have been providing our free mobile voice call service for over a decade, delivering timely and targeted preventive health information to expectant and new mothers in underserved communities, in languages and at times that suit them. We found that regular use of this service is linked to improved behavioral and health outcomes for women and their babies, with 17% more infants achieving healthy birth weights in their first year, 25% more pregnant women taking vital iron and folic acid (IFA) tablets, and a nearly 48% increase in women’s awareness of family planning methods.
By integrating the AI model developed with pro bono support from Google’s research team, we managed to retain approximately 30% of the new and expectant mothers at the highest risk of dropping out from the program. This allowed more women to receive and act on the healthcare information provided through our program, leading to better health outcomes for them and their babies.
Initially piloted with 175,000 women in 2019, the AI powering this program has since been scaled to reach approximately 350,000 women who have used the service, said Aparna Hegde.