Show Notes
AI tools lost half his chemotherapy log. Another found a procedure in his notes that no doctor had mentioned.
In episode 2 of The Agentic Patient, Faces of Digital Health speaks with Russ Read-Barrow, who has lived with bowel cancer since 2021 and is now treated for stage 4 disease. Read-Barrow looks after AI for the business he works for, and has built his own set of AI tools to track chemotherapy symptoms, manage medication and supplements, prepare for oncology appointments and monitor new research. The conversation covers what has worked, what failed, how clinicians have responded, and why he deliberately avoids using AI to ask about his condition.
The Agentic Patient is a Faces of Digital Health series on how patients use AI in their own care, and what that means for clinicians and health systems.
Guest
Russ Read-Barrow — patient, blogger at FcancerwithAI, responsible for AI at his workplace
Host
Tjaša Zajc — founder and host, Faces of Digital Health
What the conversation covers
• Building an AI chemotherapy symptom tracker with Claude
• Why ChatGPT and Manus lost half of three months of symptom data, and how to back up AI health logs
• "One tool, one job": separating symptom tracking from medication and supplement planning
• Using NotebookLM to summarise medical notes and prepare for oncologist appointments
• Paid versus free AI models and the privacy of personal health data
• How oncologists and surgeons respond to patient-generated AI summaries
• NHS records fragmentation and specialists working without the full picture
• Prompting against AI sycophancy and confirmation bias
• A weekly AI research agent that scans new cancer trials against a tumour profile
• AI, anxiety and control: reassurance, worst-case thinking and when to call the care team
• Agentic AI in practice: Claude Cowork and Dispatch from a hospital bed
Chapters
02:00 Introduction: The Agentic Patient series
02:54 Diagnosis at 40: bowel cancer and smouldering myeloma
04:41 Why paper chemotherapy diaries fail, and the move to AI
07:08 Paid vs free AI models: what happens to patient data
09:55 Lost data: AI memory limits and the Google Docs backup
11:11 How oncologists and surgeons respond to AI-generated summaries
15:53 ChatGPT, Gemini, Claude, Manus: choosing tools, one tool per job
18:34 Prompting against confirmation bias, and a weekly research agent
26:36 Reassurance or worst case: AI and patient anxiety
33:03 Control, self-blame and the language of "treatment failure"
39:20 Agentic AI in practice: Claude Cowork and Dispatch
44:15 How NotebookLM surfaced an undisclosed gallbladder removal
45:49 Has AI ever been wrong? Questioning health misinformation
This conversation reflects one patient's personal experience and is not medical advice. Discuss any changes to your care with your clinical team.
Faces of Digital Health
Website: https://www.facesofdigitalhealth.com
The Agentic Patient series: https://www.facesofdigitalhealth.com/agentic-patient-blog
Newsletter: https://fodh.substack.com
Spotify: https://open.spotify.com/show/4cElKJHrauyP6QJQaCkvdY
Apple Podcasts: https://podcasts.apple.com/gb/podcast/faces-of-digital-health/id1194284040
LinkedIn: https://www.linkedin.com/company/faces-of-digital-health
Russ Read-Barrow's blog: fcancerwith.ai
#TheAgenticPatient #PatientAI #BowelCancer #FacesOfDigitalHealth #DigitalHealth
More from Faces of Digital Health
Her relative said it was nothing. Two weeks of symptom history in ChatGPT said get checked. She was diagnosed at stage one.
Three in ten AI pilots at Sheba Medical Center never scale. Eyal Zimlichman on why, and what Europe's regulation holds back.
When prevention works, nothing happens. That's why it gets cut. Joanna Lane on financing prevention as long-term investment.
Nearly €200m, twenty years of EU health data projects. Henrique Martins on what they built and what EHDS needs next.
One ultrasound tech, four scanning rooms. NVIDIA's David Niewolny on what "augmented autonomy" asks of clinicians.