Even great oncologists forget things
Oncology has become extraordinarily complicated. New drugs, new combinations, new biomarkers and new indications arrive almost every week. Even physicians who spend their careers treating cancer struggle to keep up.
But there's another challenge we don't talk about enough: sometimes the problem isn't knowing what to do. It's remembering everything that needs to be done.
Knowing the right treatment regimen is only half the job
Consider a patient starting a new cancer treatment. The oncologist knows which regimen to prescribe and understands the evidence supporting it. But what about everything surrounding that decision?
Does the patient need prophylactic antibiotics? Anticoagulation? Additional molecular testing? A dermatology or ophthalmology consultation? How frequently should laboratory tests be performed? Which adverse effects warrant immediate intervention?
These aren't necessarily difficult clinical questions. They're details that an experienced oncologist already knows but can easily overlook during a busy clinic.
As a senior community oncologist recently told us: "I know what drug to give, but that's the skeleton. I lack everything around it to make sure the patient's care is done right."
That's an important distinction. Choosing the right treatment is essential, but delivering that treatment safely and effectively requires dozens of additional decisions.
The human cost of a busy clinic
Community oncologists routinely move between patients with entirely different cancers, treatment histories and personal circumstances. A physician might spend an hour discussing end-of-life care with one patient, then immediately see another patient who needs a complex treatment decision.
There's rarely time to reset.
The oncologist we spoke with described this experience in candid terms: "My dream is to have a continuous measure of what I'm doing right and what I'm doing wrong, because I can't always practice at my best."
He explained that after an emotionally difficult consultation, his mind can remain focused on the previous patient even as he's trying to make decisions for the next one. This isn't a lack of expertise or commitment. It's the reality of practicing medicine under enormous cognitive and emotional pressure.
And it raises an important question: Why do we expect physicians to remember every detail of increasingly complex cancer care, every time, and under every circumstance?
Clinical decision support should do more than recommend treatments
Most clinical decision-support tools focus on helping physicians choose the appropriate therapy. That's valuable, but it addresses only part of the problem.
Imagine a system that doesn't simply recommend a treatment regimen but also reminds the physician about the supportive care, monitoring and follow-up that accompany the regimen.
A patient is starting a particular therapy. The system surfaces relevant prophylaxis considerations. A molecular test hasn't been ordered. The system flags it. A treatment requires additional monitoring. The physician receives a timely reminder.
These aren't attempts to replace clinical judgment, they're ways to help physicians apply their knowledge more consistently. The goal isn't to tell an experienced oncologist something they don't know. It's to make sure important details don't get lost in the demands of a busy day.
What if every oncologist could practice at their best?
We're building Living Algorithms that capture how experienced physicians approach complex clinical decisions. But these decisions involve more than selecting a drug.
They also include the practical details that determine whether a patient receives comprehensive, high-quality care.
By organizing this expertise into structured clinical pathways, we can help physicians see not only which treatment options are available, but also the considerations that accompany each choice.
And as AI becomes increasingly integrated into clinical workflows, that knowledge could eventually be surfaced automatically, when and where it's needed.
The opportunity isn't simply to make oncologists more knowledgeable. It's to help them consistently practice at the level of knowledge they already possess.
The best physicians still need help
There's a tendency to think of clinical decision support as something primarily intended for less experienced physicians. But the more experienced an oncologist becomes, the more aware they may be of everything that can go wrong.
The physician we spoke with put it simply: "For me, it's about sleeping better at night."
He wasn't asking for an AI system to make his decisions. He wanted tools that would help him reduce avoidable mistakes and feel more confident that he had done everything possible for his patients. That's a different way to think about medical AI.
The question isn't whether AI can outperform an oncologist on a clinical benchmark. It's whether AI, combined with expert clinical knowledge, can help an oncologist deliver better care on a difficult Tuesday afternoon.
The best oncologists don't need technology to replace their judgment. They need technology that helps them remember everything their judgment depends on.