2024 Discussion Series on AI and Global Health Convening #2
At the Bay Area Global Health Alliance’s second convening of the 2024 Discussion Series on AI and Global Health, thought leaders from academia, philanthropy, and healthcare explored the transformative potential of AI in enhancing healthcare delivery by improving data accuracy and addressing biases, amidst ongoing challenges with imperfect data. Their insights shed light on key challenges and unanticipated learnings of working with AI-powered solutions. The meeting highlighted important considerations at this critical juncture of integrating AI in global health.
“I fundamentally believe that AI has the potential to rapidly transform healthcare, but we need to be mindful of how we implement it so that it leads to meaningful change in health outcomes,” said Rebecca Distler, strategist for AI, data, and digital health at the Patrick J. McGovern Foundation.
In this pivotal moment for healthcare innovation, stakeholders are urged to adopt a balanced approach to AI integration and address both high-risk applications and foundational data challenges. As the global health sector embraces AI’s transformative potential, critical questions arise about its necessity and its impact on health outcomes, prompting a strategic and thoughtful approach to technology adoption. What follows are some of the key takeaways.
________________________________________________________________________________________________________________
Key takeaways:
1. The Patrick J. McGovern Foundation: Grounded in Needs, Driven by Implementation: Navigating AI’s Impact on Global Health
- The Patrick J McGovern Foundation advocated for a strategic and thoughtful approach to leveraging AI and data science for public good and social impact in the global health landscape.
- Encouraged asking critical questions about whether AI is truly necessary for a given challenge and addressing fundamental issues around data, digital systems, policy, and governance to enable the transformation of healthcare through AI.
- Recommended a holistic approach to integrating AI, considering the entire ecosystem, from data collection to implementation, and prioritizing human-centered design.
2. inSupply Health: Automate Before Innovate: Building Trust Through Streamlined Processes Paves the Way for Successful AI integration
- inSupply Health advocated for using AI to enhance forecasting for supply chains in public health.
- Advised those new to AI to first start by automating data extraction and cleaning processes to make data more accessible and manageable, building trust and excitement among stakeholders, and then implementing AI machine learning algorithms.
- Recommended implementing user-friendly visuals, and addressing data security concerns to ensure the successful adoption, sustainability of technology solutions, and continuous stakeholder engagement.
3. Palindrome Data: Data-centric Approach and Challenges in Global Health Datasets
- Palindrome Data underscored the multifaceted nature of data challenges in public health and the importance of holistic approaches that integrate technology, policy, and human factors to drive meaningful impact.
- Identified what drives data challenges, including delays in data digitization, data privacy concerns, regulatory complexities, and fragmented data systems.
- Stressed the importance of addressing issues such as interruption in treatment and the need for timely interventions, especially for at-risk patients.
- Addressed the necessity of building trust among end-users through consistent and reliable data and underscored the crucial role of people in effectively utilizing data.
4. Jacaranda Health: From Data Infancy to AI Maturity: Evolving Solutions for Complex Problems
- Jacaranda Health addressed how the implementation of digital health tools has allowed them to engage with expecting mothers, provide information, and triage questions using AI.
- Emphasized the importance of problem-solving even with limited data, illustrating how they simplified feedback collection by asking a single question and leveraging AI to analyze responses.
- Demonstrated how collected data can drive actionable insights, leading to improvements in healthcare facilities, such as reducing wait times based on patient feedback.
- Underscored the importance of adapting data visualization methods to suit the needs of different stakeholders and emphasized the ongoing refinement of AI tools to handle complex information effectively.
- Suggested that AI is not always necessary at the outset but can become invaluable as datasets mature.
________________________________________________________________________________________________________________
In the opening discussion, moderator Krista Donaldson, director of innovation to impact at Stanford Byers Center for Biodesign, and Distler explored AI’s potential to bring about significant positive changes in healthcare delivery, patient care, and health outcomes. Distler suggested that AI holds promise for revolutionizing healthcare delivery by improving decision-making, efficiency, personalization, accessibility, and innovation in the healthcare industry. However, she also underscored the importance of careful consideration of regulatory, governance, and ethical frameworks to ensure that AI innovations lead to meaningful improvements in healthcare. “We can be really smart about the data, but we need to be really smart about the implementation as well if we’re going to see any impact from these tools,” said Distler.
“How do you suggest organizations get feedback when they have a vision of how to use AI or digital health?” asked Donaldson. Distler emphasized the importance of seeking diverse perspectives, engaging with stakeholders, and actively soliciting feedback throughout the AI development process to ensure that organizations’ visions align with the needs and expectations of their target audience.
Distler explained that the Patrick J McGovern Foundation balances meeting the needs of frontline organizations while leveraging a specific type of technological approach, such as AI and data science, through a combination of tailored support, collaborative partnerships, and strategic alignment. Through these partnerships, Distler encourages innovative and behind-the-scenes uses of AI, as they have the potential to transform health systems and support meaningful change.
She also encouraged the audience to question whether AI is always a necessary tool. “Is this really an AI problem? There are so many folks excited about these solutions, but the most exciting use cases are [from] some of the behind-the-scenes work. It becomes very clear when you work with partners who live human-centered design, where it’s not just a sticker they slap on their approaches. We need to start with community needs and think about how AI can help us better serve those communities,” said Distler.
“Where do you think healthcare will be in regard to AI in the next 10 years?” Donaldson asked Distler before concluding the discussion. “I anticipate that large language models are going to accelerate clinical decision support and change the nature of relationships between patients and clinicians,” responded Distler.
The remainder of the discussion allowed the three presenting panelists the opportunity to elaborate on navigating challenges with data, leveraging AI for social impact, and the future direction of healthcare innovation.
Use Case Challenge #1: inSupply Health
Yasmin Chandani, inSupply Health CEO, highlighted her organization’s experience using AI to enhance forecasting for supply chains in public health sectors using AI. Chandani shared examples of how AI implementation led to significant improvements at inSupply Health, such as reducing data processing time from days to minutes, increasing the amount and quality of data analyzed, and tailoring forecasting methods to specific products and regions.
Chandani elaborated on how inSupply Health has implemented AI in their family planning program in Kenya and with their national quantification team for essential medicines in Tanzania. “Before AI, data for our family planning commodity was manually collected from the Kenya Health Information System for all 47 counties, using only 12 months of data, visually checking for outliers and estimating needs using average monthly consumption,” said Chandani. In contrast, with AI integration, inSupply Health’s family planning tool now automatically extracts 3.5 years of data from >10,000 facilities standardizes and cleans the data, including outlier detection and missing value completion. “Our AI-integrated family planning tool also streamlines our original process by reducing data processing time from 2.5 days to 15 minutes” explained Chandani. “Before AI, we had one-size-fits-all in terms of methodology to use, and now we can actually adjust the robustness of the forecast,” said Chandani.
Chandani noted that AI can be perceived as scary, and many initially exercise caution with its implementation. For inSupply Health, overcoming this fear and skepticism required transparent communication, education, and demystifying AI concepts. “In our approach, we unpacked the challenges of AI for our stakeholders and trained them on the concepts before we even started the process,” explained Chandani. “Building confidence and trust in AI technology was critical for stakeholders to buy in. We did this by creating user-friendly visualizations at all stages of the process so that changes were easy to interpret, AI predictions were validated, and decisions were informed,” said Chandani. inSupply’s approach showcased how AI can effectively lead to tangible improvements in efficiency, effectiveness, and user perception, and add value to imperfect data.
Use Case Challenge #2: Palindrome Data
To follow, Lucien de Voux, co-founder at Palindrome Data illustrated how each type of data (whether routinely collected surveillance data and data from research studies) has its own challenges, including delays in data digitization, data privacy concerns, regulatory complexities, and fragmented data systems.
“Great data science is not enough. We can build these amazing predictive models, but if we don’t have a mechanism to distribute that information, our partners aren’t able to consume a predictive algorithm,” said de Voux while elaborating on the importance of digitization of data. Likewise, He highlighted the importance of also capturing data in real-time at the point of care to enable immediate decision-making and interventions, particularly in preventing treatment interruptions and improving patient outcomes. “Predictive AI algorithms are a game changer in terms of the ability to differentiate models of care with patients in-facility,” said de Voux. He also pointed out that inconsistent data can erode trust in the system among end-users and that building trust and adoption relies on both good data and competent people who can effectively leverage that data.
To reduce challenges with data, he advocated for being technology agnostic and embedding AI solutions into existing tools used by healthcare workers, rather than adding new tools to reduce the burden on healthcare workers. Lucien also encouraged closing the loop between predictions, interventions, and outcomes to enable continuous learning and improvement in AI-driven interventions.
Lucien’s overview of the complexities of working with data in public health settings underscored the importance of addressing these challenges to leverage AI solutions for better patient care and outcomes.
Use Case Challenge #3: Jacaranda Health
Sathy Rajasekharan, co-executive director at Jacaranda Health discussed how his team has utilized AI to streamline communication and improve the quality of care for mothers and babies. Through a digital health tool called PROMPTS, mothers receive text messages providing information about their pregnancy or newborn, appointment reminders, and coaching. However, the volume of inquiries—up to 6,000 questions per day from over 120,000 active users—requires efficient prioritizing. Here, AI prioritizes urgent queries, ensuring timely responses while maintaining a human touch. “There’s no way we could be answering all those questions within minutes without using a form of AI to triage the questions, identify the most urgent, and then have a human in the loop to be able to answer the questions and adjust the responses,” said Rajasekharan.
Additionally, AI assists in analyzing feedback from clients, allowing for valuable insights to be shared with healthcare facilities for continuous improvement. Beyond basic queries, Jacaranda Health is leveraging AI’s capabilities in natural language processing to delve deeper into complex narratives shared by users, aiming to automate the classification of subjects discussed and summarize conversations for more efficient data processing and analysis. “While this isn’t perfect, we’ve got a really good model that can detect the primary topic of conversation, and when multiple mothers address multiple issues through PROMPTS, we’re working on using things like large language models, blended with other tools, to really get at the statements they’re making,” concluded Rajasekharan. His group’s journey highlights the evolution of how organizations are integrating AI to address critical healthcare challenges and emphasizing the importance of strategic problem-solving and a gradual build-up of datasets to maximize AI’s potential impact.
Special thanks to Pfizer, sponsor of the Alliance’s AI and Global Health Discussion Series. Stay tuned for the third convening in the discussion series, taking place on Tuesday, June 25 from 8:30-10am PT.
________________________________________________________________________________________________________________
About the Alliance’s AI and Global Health Series:
Following the Alliance’s panel on AI during its 2023 Annual Meeting, and the Alliance’s participation in the White House Office of Science and Technology Policy (OSTP) Roundtable on AI and Global Health, members requested further convenings to understand opportunities and implications for their work.
This series hopes to tackle that – showcasing members who have started integrating AI into their organizations and programs and exploring how best to ensure the responsible and ethical implementation of AI in their initiatives. It hopes to provide paths forward for those new to AI and those who are well-versed in the emerging technologies.
Leveraging AI technologies offers innovative solutions that can positively impact healthcare systems worldwide. In the context of global health, shifts in healthcare within developed markets offer the promise of significant improvements for underserved communities in global health. The impact spans from empowering community-health workers to enhance patient care in remote regions to assisting low and middle-income countries (LMICs) in proactively averting deadly disease outbreaks. There is a growing acknowledgment of the vast potential of AI tools to disrupt traditional tradeoffs in health access, quality, and cost. Health systems in LMICs grapple with significant challenges such as severe shortages of personnel, medical equipment, and other resources, demanding strategic and innovative solutions. AI tools present an exciting opportunity not only to optimize existing resources and address workforce shortages but also to significantly enhance healthcare delivery and outcomes in low-resource settings.
The series will feature experts, practitioners, and stakeholders from diverse sectors to discuss applications of AI in areas such as healthcare delivery, disease surveillance, diagnostics, and treatment. With a focus on education, knowledge sharing, and networking, these interactive virtual sessions are designed to:
-
- Increase our community’s knowledge and awareness of how AI is being integrated into global health programming,
- Surface opportunities and challenges with the use of AI in low-resources settings, emphasizing critical success factors, and
- Build a vibrant community for learning, dialogue and networking.
About Our Speakers for Navigating Imperfect Data:
Rebecca Distler, Strategist for AI, Data, and Digital Health, The Patrick J. McGovern Foundation. Rebecca is a Strategist for AI, Data, and Digital Health at the Patrick J. McGovern Foundation, a philanthropy advancing AI and data solutions to create a thriving, equitable, and sustainable future for all. Rebecca has spent the last decade building strategic partnerships and programs with technology companies, foundations, non-profits, and governments to advance innovation in global health. Rebecca previously served as Global Health Programs Lead at ID2020, working on privacy-preserving digital credentials for COVID-19 vaccines, and as Director of Global Health Initiatives at Element, working with partners across Africa and Asia to develop AI digital identity platforms for use in immunization and other health programs. Rebecca holds a BA in Political Science from Yale University and a Masters in Health Policy and Global Health from the Yale School of Public Health. She is a Term Member at the Council on Foreign Relations and a World Economic Forum Global Shaper; she was previously selected as a Forbes Ignite Impact Fellow, AI XPRIZE Semi-Finalist, and Gavi INFUSE Pacesetter.
Yasmin Chandani, CEO, inSupply Health. Yasmin is the CEO of inSupply Health, an East African health advisory firm dedicated to improving people’s access to essential health products and services. Yasmin has spent over 25 years supporting and advising national governments, NGOs and multilateral partners in the strategy, design, implementation and measurement of strong, sustainable supply chains for health. She has worked in 15 countries and served as a director of multi-country, multi-year, complex initiatives, leading teams to develop pioneering supply chain solutions for emergent HIV/AIDS programs and for a range of health programs including family planning, community health, immunization, malaria and essential medicines. Considered a supply chain thought leader, Yasmin is known for her rigorous attention to quality and a successful track record in contextualizing and adapting innovations for building people-centered, responsive supply chains. She is a tireless advocate for supply chain professionalization and the preparation of next generation supply chain professionals, especially women.
Sathy Rajasekharan, Co-Executive Director, Jacaranda Health. Sathy Rajasekharan is the Co-Executive Director at Jacaranda Health, and oversees the organization’s mission of delivering low-cost, sustainable solutions through public hospitals to improve maternal and newborn health outcomes. Prior to joining Jacaranda Health, Sathy worked at the Clinton Health Access Initiative (CHAI), and led CHAI’s work providing technical assistance in Health Financing and Supply Chains to the Eswatini Ministry of Health. He has held previous positions in the innovation space in Montreal, where he helped to develop and commercialize technologies, including digital tools. Sathy holds a Ph.D. in Neuroscience. He was born in a government hospital in Lusaka, Zambia.
Krista Donaldson, Director of Innovation to Impact, Stanford Byers Center for Biodesign, and Bay Area Global Health Alliance board member. As Director of Innovation to Impact at Stanford Byers Center for Biodesign, Krista Donaldson’s work focuses on ensuring that design tools and processes are broadly applicable across global markets. She is also part of the team establishing the East Africa Biodesign Program, which kicked off in early 2023. Stanford’s Biodesign program advances health outcomes and equity through innovation education, translation, and policy. As the former CEO of Equalize Health (formerly D-Rev), Donaldson led the design and scaling of disruptive medical devices to address global health inequities. To date, nearly 1M people – mostly children and young people – have been treated by one of Equalize Health’s products in 70+ countries. Peter Singer of the Effective Altruism movement called Equalize Health “one of the world’s best charities” because of its cost effectiveness and exemplary end-to-end processes. Donaldson has been recognized as a World Economic Forum Technology Pioneer, TED speaker, and a GLG Social Impact Fellow. She was also named one of Fast Company’s “50 Designers Shaping the Future.” Prior to Equalize Health, she was an Economic Officer at the U.S. Department of State where she managed part of Iraq’s reconstruction portfolio. She also worked at KickStart International (Kenya), and the design firm IDEO (USA). Donaldson holds a master’s degree in Product Design and a Ph.D. in Mechanical Engineering from Stanford University.
Lucien de Voux, Co-Founder, Palindrome Data. Lucien aims to improve lives by operationalizing AI in demanding, under-resourced environments. He is a creative market strategist, who blends business and technology competencies to achieve results in the risk and high-tech product development domains. Palindrome Data, a data analytics startup he cofounded in South Africa, now consumes his working hours and has him doing everything from product strategy to field deployments.