Castor

The RWE Crossroad: the physician, the patient, or the machine?

Real-world evidence has more ways than ever to capture a patient’s story: directly from patients, from medical records and claims, and increasingly from AI reading the whole chart. So where does the physician still belong, and does the answer even matter once patients authorize their own data?

On-demand

This webinar is on-demand. Get access to the webinar here.

Duration: 40 minutes

Watch Now Watch Now

Records tell you a patient switched from treatment A to treatment B. They don’t tell you why.

That gap, the clinical context behind real-world data, is where one of the sharpest debates in RWE is happening. And it is no longer just physician vs. patient data. It is physician vs. patient vs. AI reading the whole chart.

Matthew Gordon of Aptitude Health and Derk Arts of Castor each take a side they do not fully believe, then drop the act to work through what actually holds up:

When a physician is worth the cost, and when direct-to-patient is enough
Why a GLP-1 study and an oncology study need very different designs
What a physician adds that records and patient-reported outcomes cannot
Where AI fits today and in three years, with human judgment in the loop
Why a patient-authorized data layer turns the whole question into a routing decision

For RWE, HEOR, and medical affairs leaders deciding how to build evidence that holds up. Watch back on-demand now.

Frequently Asked Questions

Does real-world evidence still need physicians?
It depends on the question the study has to answer. Some real-world evidence questions can be answered by pulling standard-of-care and patient-reported data directly, with no active physician role. Others depend on clinical context, the why behind a treatment decision, that only the treating physician can supply. Start from the strategic objective and design the study to fit it.
What is direct-to-patient real-world evidence?
It collects data with the patient at the center rather than routing everything through a site. Patients consent electronically, retrieve their own medical records (in the US via TEFCA and FHIR or a HIPAA release), and contribute patient-reported data, with wearable device data connected via API.
How does AI extract data from medical records without losing clinical context?
The safeguard is human review, not full automation. Castor Catalyst is an AI-driven extraction pipeline with auditable steps and human review built in, so a person confirms what the model pulls before it enters the record. Letting an AI infer missing clinical reasoning on its own invites hallucination, so the role of AI is to surface and structure what is in the record, with a human keeping control.
How do you keep AI-assisted real-world data trustworthy?
By keeping provenance clear. The evidence stays trustworthy only when it is obvious what came from the patient, what came from the clinician, and what was inferred by the model. Castor Catalyst attaches a confidence score and source evidence to every extracted value, and human reviewers stay in control, with risk-based review before data enters the record. Human-in-the-loop is built in by design.