Ashish Manne, Associate Professor at Ohio State University Comprehensive Cancer Center-James Cancer Hospital and Solove Research Institute, shared on LinkedIn:
“Two Years Building AI for Pancreatic Cancer Taught Me Who Not to Listen To
Many of the loudest voices in healthcare AI have never built anything. They lecture and podcast because they’re key opinion leaders (KOLs). I built an AI tool with Ohio State University Comprehensive Cancer Center-James Cancer Hospital and Solove Research Institute and Microsoft. It took two years to reach a proof-of-concept pilot in pancreatic cancer. Then I talked to 30 healthcare professionals, from oncologists to CTOs, and skipped the AI gurus.
What physicians treating patients every day told me:
- They don’t want AI at the point of care unless they must.
- They accept that evidence gaps exist and aren’t looking to AI to close them.
- They recognize tools built to generate revenue, like finding trial patients so institutions get paid. Don’t insult them by pitching it as help.
- They do want real-world evidence and local treatment patterns, if it meets scientific rigor.
If your strategy is hiring KOLs, you will fail. Over 80% of patients are treated where those KOLs don’t practice.
My worry: in 10-20 years, tools may aim to replace physicians without their consent, and no KOL will come to the rescue. I go deeper on this in my Substack on AI in medicine, where I separate hype from real clinical value.”
Learn more here.
Other articles about Pancreatic Cancer on OncoDaily.