Leveraging Microsoft’s Azure OpenAI, Ontada streamlines oncology data analysis, reducing processing times by 85% to support clinical decisions and improve patient care for over 3000 oncologists across 30 states, according to Wanmei Ou, PhD, vice president of product, data analytics, and artificial intelligence (AI) at Ontada.
Ontada, a McKesson business unit, has revolutionized oncology data management with its flagship electronic health record (EHR) system, iKnowMed, used by more than 3000 community-based oncologists and supporting millions of patients. Rigorous model development and ongoing monitoring ensure data quality, while future innovations focus on expanding real-world data applications in oncology research and addressing health equity in rural communities.
Wanmei Ou, PhD, vice president of product, data analytics, and artificial intelligence (AI) at Ontada, discusses how by partnering with Microsoft, Ontada has reduced oncology data processing from 6 months to just weeks, alleviating documentation burdens and enhancing clinical decision support.
Ou explained the importance of using AI and natural language processing (NLP) to unlock unstructured data in oncology, as 70% to 80% of critical patient information is embedded in nonstandardized documents due to the interdisciplinary nature of oncology care. These data span across diverse systems and involve various specialists, including oncologists, surgeons, and primary care providers.
AI applications in this field serve 3 primary use cases, according to Ou. The first use is extracting real-world evidence from unstructured data for organized access. Second, AI applications reduce documentation burdens, allowing physicians to focus more on patient care than data entry. Third, these applications provide clinical decision support by integrating patient context, treatment guidelines, and insurance policies to recommend optimal next steps.
Ontada processes a massive amount of unstructured data—around 150 million documents from 2 to 3 million patients’ longitudinal records, said Ou. Due to the sheer scale, even Microsoft hadn’t previously encountered such a high-volume processing need within a short time frame. Through close collaboration and multiple on-site meetings with Microsoft’s technical specialists, Ontada leveraged the new Azure OpenAI, reducing processing time from an expected 6 months to just 2 to 3 weeks. This partnership has enabled Ontada to handle vast data more efficiently, and the solution may benefit other organizations facing similar high-volume data challenges, Ou noted.
AI in Health Care: Balancing Governance, Innovation, and Trust
September 2nd 2025In this conversation with Reuben Daniel, associate vice president of artificial intelligence at UPMC Health Plan, we dive into how UPMC Health Plan builds trust with providers and members, discuss challenges of scaling AI effectively, and hear about concrete examples of AI's positive impact.
Listen
Stagnation in Reimbursement Keeps Biomedical Innovation From Reaching All Patients, COA Panel Says
September 9th 2025Panelists at the Community Oncology Alliance Payer Exchange Summit discuss the urgent need for innovative reimbursement models in cancer care to match advancements in biomedical technology and drug discovery.
Read More
Current and Emerging Options for Uncommon EGFR- and Exon 20 Insertion–Mutated NSCLC
September 8th 2025Uncommon EGFR mutations in non–small cell lung cancer (NSCLC) remain challenging to treat, but new tyrosine kinase inhibitors, bispecific antibodies, and a proposed “PACCage insert” framework provide opportunities to advance precision therapy.
Read More