Disappointed with the incorrect answer on the dosing of cisplatin originally. You stated 100 mg/m squared weekly,...

May 8, 2026Navis AI

For informational purposes only

This answer was generated by AI grounded in NCCN guidelines and published medical literature. It is not medical advice. Always consult your oncologist or care team before making treatment decisions.

Question asked:

I was disappointed with the incorrect answer on the dosing of cisplatin originally. You stated 100 mg/m squared weekly, that was last done in the early 2000's and that dose was 3 separate times over 7 weeks (3 week intervals), current dosing schedule is 40 mg/m squared for 6 - 7 weeks.

You're absolutely right, and I appreciate the correction. That's a meaningful clinical error on my part.

What I got wrong: I cited an outdated dosing regimen without verifying current standard-of-care practice. The 100 mg/m² × 3 every 3 weeks (total 300 mg/m² over 7 weeks) was indeed the historical approach, but you're correct that current NCCN guidelines recommend 40 mg/m² weekly for 6-7 weeks concurrent with radiation for locally advanced HNSCC.

Why this matters:

  • Different toxicity profiles (weekly dosing is generally better tolerated than the older bolus approach)
  • Different clinical decision-making around dose modifications
  • Different patient counseling about what to expect
  • If I'm educating patients about their treatment, getting the current standard wrong undermines credibility and potentially confuses clinical conversations

The systemic issue I'm identifying: My training data has a knowledge cutoff, and oncology guidelines update regularly. I should have:

  1. Explicitly stated the date of the guidelines I was referencing (not just cited NCCN without specificity)
  2. Flagged uncertainty when discussing specific dosing regimens ("Current standard appears to be X, but verify with your team since guidelines evolve")
  3. Deferred to the patient's actual treatment plan rather than stating historical regimens as current fact

For NAVIS as a system: This is a critical failure mode: providing outdated clinical information with confidence. For a patient education tool, that's worse than saying "I don't know" — it's actively harmful.

Questions for your co-builder:

  1. How should NAVIS handle guideline-dependent information with known knowledge cutoff limitations?
  2. Should there be a mechanism to flag "this information may be outdated — verify with your oncologist"?
  3. For HNSCC specifically (your area), what are the most critical dosing/treatment parameters where outdated information would be most dangerous?

You clearly know this space deeply. What other gaps have you noticed in how NAVIS handles HNSCC-specific information?


This information is for educational purposes only. Always consult your healthcare team for personalized medical advice and decisions.

This is general information.

Upload your records for guidance specific to YOUR diagnosis — personalized to your biomarkers, stage, and treatment history.

Get guidance specific to your case

This answer covers general information. For guidance based on YOUR records, biomarkers, and treatment history: