When cancer patients stop treatment, it is rarely a simple decision. Many face barriers that have little to do with medicine, such as the cost of travel, lost income, food insecurity, or the challenge of balancing care with family responsibilities. During the pandemic, these pressures deepened, and national data showed that thousands of Americans, particularly those with advanced breast and prostate cancers, were not returning to care once it was disrupted.
Pfizer's Oncology CAUSE programme, standing for Collaborative Actions to Understand and Solve for Equity, set out to address this. The company saw an opportunity to use artificial intelligence to bridge one of healthcare's most persistent divides. These social and practical barriers prevent people from getting the treatment they need.
The pilot began in Texas, where the data showed the steepest decline in post-pandemic cancer diagnoses. Pfizer’s oncology team, led by Carmen White, collaborated with more than forty community-based organisations, hospitals, and advocacy groups across Dallas and Houston. Together, they designed a tool to match patients with the support services that would help them return to care, whether that meant transportation, accommodation, or mental health support.
“The community told us clearly that solutions have to come from within,” said White. “Our job was to listen, understand the realities people face every day, and then use technology to remove friction from the system.”
Technology with a Human Touch
The AI-driven matching tool works through community health workers and patient navigators, who are already trusted figures in the healthcare ecosystem. By entering basic patient information, such as diagnosis, age, and ZIP code, they can instantly generate a list of relevant local resources, complete with eligibility details and feedback from other users.
To make the system practical in real-world settings, the team used two forms of AI. Generative AI scans the broader ecosystem for new or updated services, while static AI creates a feedback loop. Community health workers can record whether a resource responded promptly, provided support in the patient's language, and met the patient's needs, ensuring that information is continuously refined.
White described the project’s intent as both technological and emotional. “AI can be cold if you let it,” she said. “We built this tool to feel like an extension of the community health worker’s role, not a replacement for it. It saves time, but more importantly, it restores connection.”
The system operates independently rather than being embedded in a hospital infrastructure, allowing it to be used anywhere with minimal training. The tool is intuitive, accessible, and adaptable, allowing any patient advocate or navigator to use it immediately.
From Proof of Concept to Scalable Impact
Pfizer's collaboration with Leal Health, an AI company founded by a cancer survivor, provided the technical foundation. The platform was built using real patient data from IQVIA to ensure recommendations reflected local realities. The initial focus was on metastatic breast and prostate cancer, where gaps in follow-up care were widest and urgency greatest.
After the proof-of-concept succeeded in Texas, Pfizer issued a competitive Request for Proposal (RFP) to transfer ownership of the tool to a nonprofit organisation, ensuring transparency and long-term sustainability. The Patient Advocate Foundation was chosen to manage the rollout and to extend access to hospitals, advocacy networks, and new geographies.
White emphasised why this handover mattered. “Equity doesn’t happen if innovation stays locked inside a company,” she said. “We wanted this tool to live in the world, to belong to the people and organisations who work closest with patients every day.”
Building for Global Reach
Pfizer is now adapting the model for international use through its Accord programme, starting with pilots in Africa. Countries such as Kenya, Tanzania, and Nigeria have been identified as potential partners, given their rising cancer rates and access-to-care gaps. The mapping plan prioritises regions where both diagnosis and treatment delays are most acute.
The tool’s design allows for this scalability. It requires no integration with national systems, accommodates local resource databases, and can be rapidly localised for language or regulation. Its underlying principle remains the same, connecting patients to care by removing avoidable barriers.
White reflected on what the Texas pilot revealed. “Every data point is a story of someone trying to get help,” she said. “AI can surface those stories faster, but it’s people who turn them into solutions. That’s what this project is really about.”
The Broader Meaning of Equity
Beyond its technical success, Oncology CAUSE shows that meaningful progress in health equity comes from collaboration, not competition. By building with communities instead of for them, Pfizer and its partners have created a model that blends technology, empathy, and accountability.
In White’s words, “You can’t fix inequity from the outside. Real change happens when communities see themselves in the solution. That’s when trust begins, and that’s when people come back to care.”
Find out more about the winners and finalists at the Reuters Events Global Pharma Awards 2025 - visit the website here.
