Community oncologists don't need more information, they just need the right information
Oncology does not have an information shortage.
Every week brings new trials, FDA approvals, biomarkers, indications, combinations, guidelines and conference data. An oncologist can open NEJM, search UpToDate, read NCCN or ask an AI to summarize all of it in seconds.
The problem is not finding information, it's knowing which information matters when a patient is sitting in front of you.
What do I need to know before I prescribe this tomorrow?
We recently met a board partner in community oncology who sees more than 30 patients a day across lung cancer, GU malignancies, myeloma and other diseases.
When a new drug is approved, he doesn't need a comprehensive review of the pivotal trial. He needs answers to more practical questions:
What do I need to prescribe with it?
Which toxicities do I actually need to watch for?
When should I see the patient again?
Those questions sound simple, but they're not.
A prescribing oncologist may need to know that a particular supportive medication should be started concurrently. That diarrhea or rash matters more in practice than the long list of adverse events in the label. Or that the patient should have a nurse evaluation two weeks after starting therapy.
That's the difference between knowing about a drug, and knowing how to use it in practice.
The community oncology information problem is different
An academic thoracic oncologist may spend most of their clinical career thinking deeply about lung cancer. But a community oncologist might see lung cancer at 8:00, myeloma at 8:30, prostate cancer at 9:00 and lymphoma at 9:30.
Then the FDA approves something new.
Modern oncology increasingly demands subspecialist-level knowledge from physicians who, by necessity, practice as generalists.
And the pace keeps accelerating.
New biomarkers change treatment selection. New indications move therapies into earlier lines. New combinations introduce unfamiliar toxicities. New evidence changes sequencing.
Keeping up with one type of cancer is difficult. Keeping up with all cancers is impossible.
Nobody needs another inbox
The traditional response to this problem has been to give physicians more information:
Another email
Another publication
Another guideline update
Another slide deck
Another portal
Now AI makes it possible to generate yet another summary of all of the above. But information abundance doesn't solve the problem if physicians still have to figure out what truly matters.
A community oncologist seeing 30 patients a day isn't looking for less rigorous medicine. They're looking for the shortest path from evidence to a good clinical decision.
Clinical relevance is a form of expertise
Knowing what to leave out is sometimes as valuable as knowing what to include.
A trial publication may contain dozens of adverse events. An experienced physician knows which three are likely to change management.
A guideline may list several acceptable regimens. An expert can explain why they usually choose one over another.
A label tells you what can happen. Clinical experience tells you what to anticipate on Monday morning.
That knowledge doesn't necessarily come from another database, it comes from physicians who have already treated these patients.
This is what Living Algorithms are meant to capture
Living Algorithms are designed around a simple idea: physicians shouldn't have to reconstruct the practical meaning of the medical literature every time they make a decision.
Instead of asking: "What does the evidence say?"
We can ask: "What does an experienced physician actually need to know to use this evidence?"
That means combining published evidence with the practical knowledge of clinicians who use these treatments every day.
What matters.
What to watch.
What to prescribe alongside treatment.
When to follow up.
Where the guidelines leave room for judgment.
And what experienced physicians actually do in the grey zones.
AI can make this more powerful: it can retrieve evidence, monitor new publications, summarize trials and help keep medical knowledge current.
But the objective isn't to produce more information. It's to use AI and physician expertise together to identify the information that matters.
Because the community oncologist seeing patient number 27 doesn't need another 12-page paper. They need to know what to do next.