Integrated medical coding lands in SafetyBase v3

Medical coding used to be a step you did outside SafetyBase. You'd capture an adverse event in the platform, then hand off to a coder who'd look up the right MedDRA term in a separate tool, then paste the code back into the case. Two systems. Two audit trails.

That gap closes with v3.

What's in

SafetyBase v3 ships with four dictionaries integrated directly into the AE capture, drug logging, and medical-history screens:

• MedDRA — adverse events, at PT and LLT levels
• WHO-DD — drug coding, with route and formulation
• ICD-10 — medical history and concomitant conditions
• SNOMED — where the site prefers it (kept optional per project)

The coder still owns the decision. What changes is that they do it in the same case they're already looking at, with search-as-you-type autocomplete, term-hierarchy hints, and a coding-history view that shows every term change with the reason and timestamp.

Why now

Two reasons.

Sponsors ask for it in every RFP. Every trial we've bid on in the past six months required integrated coding. Building it once and shipping it as core beats hand-writing an integration each time.

It's the last piece of the "one platform" story. Kronus for management, Nukleus for capture, SafetyBase for safety — but coding was the seam. Now the seam's gone.

What's next

Auto-coding suggestions using a fine-tuned model trained on our own historical case data. Not automatic — the coder still confirms. But 90% of coding decisions are the same three terms; if we can rank those first, we can shave an hour a day off a busy safety team.

More on that when we're ready to show it working.