Graded, attributed, and signed before it leaves your site.

Burna reads the note, grades every event against the CTCAE version your protocol names, scores attribution for each drug in the regimen, and writes the IRB, sponsor and FDA reports. Your investigator reviews and signs every one.

criterion cited at grade time

CTCAE v6.0 · GI-DIAR-3

Diarrhea, Grade 3 criterion

app.burna.ai

Amara Osei

ENC-2214

Diarrhea

G3

Increase of ≥7 stools per day over baseline; hospitalization indicated...

CTCAE v6.0 · GI-DIAR-3

“Patient reports 8 to 9 loose stools daily since C2D4, admitted for IV fluids.”

audit · graded by Dr. N. Adeyemi · countersigned

G3

Grade 3 · Severe

number and word, color blind safe

Three buyers ask us three different questions.

Cancer centres

Can we take on more trials without adding people?

Your sub-investigator reviews a toxicity in two minutes, and both CTCAE versions are handled for you. The technology fee sits in the study contract, so the trial that needs it also funds it.

For cancer centres

Pharma and biotech

Will the label be right the first time?

Attribution is scored at the encounter, at every site, while the label is still being written. And your protocol stays inside your own cloud.

For pharma

CROs

What does this do to our cost to deliver?

It lowers the line a sponsor benchmarks most easily. Clean data the first time is what keeps you on the panel.

For CROs

Everything your clinician wrote, read as evidence.

The category is called EMR to EDC. Those systems move structured fields: a neutrophil count crosses a threshold, so a row appears. That’s real work, and Burna reads the rest. Most of an adverse event lives in prose. It’s a sentence your clinician typed at the end of a long clinic, and that’s where Burna starts.

What a field-mapping system reads

Day 14 of cycle 2. Patient reports fatigue and fever overnight. Labs drawn this morning: ANC 800 cells/mm³, platelets stable. She's eating less and stopped her afternoon walks.

one value, because a field exists for it

What Burna reads

Day 14 of cycle 2. Patient reports fatigue and fever overnight. Labs drawn this morning: ANC 800 cells/mm³, platelets stable. She's eating less and stopped her afternoon walks.

four findings, three of them only in prose

What people are saying.

“What you guys are doing is a fantastic idea. This is the way that all AE grading should be done in the future.”

Scott Paulson, MDGI medical oncologist, clinical research lead, Texas Oncology

“I've seen a couple of other things like this that are more slapdash. This really looks like you put a lot of care and diligence into it.”

Ken Kobayashi, MDsenior oncology drug safety and regulatory reviewer

“Oncology is inherently complicated and inherently dense. As an oncologist, this does not seem overwhelming to me. I'm not scared of the information.”

Manuj Agarwal, MDradiation oncologist

“I see the most potential on the biotech and pharma side, simply by making AE grading more objective. And especially the attribution part.”

Oliver Overheu, MDmedical oncologist

“I'm responsible for delivering things, and you've allowed me to do it cheaper. That's the only reason I would want to purchase this system.”

John Hayslip, MDCMO, Verastem Oncology

“From somebody who's lived at the health system layer and the technology layer: clearly cancer centers can use this.”

Joel Saltz, MD PhDbiomedical informatics

“I see this allowing us to open up this many more trials, or have this many more patients on study.”

Ashley Waldrepclinical research coordinator, Nebraska Hematology Oncology

“I see it less as trimming the staff and more as being able to conduct more trials with the same size team.”

Jonathan Gerber, MDChief Medical Officer, Burna AI

“It's a question that has been there forever since we started clinical trials. And it's always very polemic.”

Michel Azoulay, MD30 years in drug development

“This question I've asked like a thousand times. Tingling. Is it better than it was before? Does it come and go?”

Usman Shah, MDGI oncologist and phase 1 trialist, Atlantic Health

“I think it's going to be a big thing.”

Islam Sadek, MDmedical oncologist

“A patient who has metastatic disease may not have a quarter year of time to wait for the drug.”

Stefan Glück, MD30 years in drug development

Grading is the first half. The reports are the other half.

A graded event still needs attribution, a narrative and a submission. Burna scores attribution per suspect drug, then writes the IRB report, the sponsor report and the FDA submission from the signed record. Your investigator reviews and signs, and the packet arrives assembled.

The signed record

Diarrhoea · Grade 3 · CTCAE v6.0

Probable, pembrolizumab. Three cited spans.

Signed by your investigator, 21 CFR Part 11.

IRB report

Site-facing safety report

Event, onset, grade and criterion.

The attribution and what it rests on.

Built from the signed record.

Sponsor report

Cross-site graded events

Every event with its evidence span.

The scoring behind each attribution.

The query, answered before it is asked.

FDA submission

MedWatch-ready narrative

Narrative written from the record.

3500A fields already populated.

Your investigator reviews and sends.

What the signed record produces, assembled and ready to send.

Which drug caused it is the harder question.

Burna scores every suspect drug with WHO-UMC and Kramer, carries comorbidities and concomitant medications into the reasoning, and shows your investigator the working before anything is signed. In a combination regimen you see which agent the evidence points at.

“If you're giving a new drug to a thousand people, you're going to have side effects that are unrelated to your drug. That's why we have these labels that are so big… They went out to a Mexican restaurant the night before.”

Charles Balch, MD · former EVP and CEO, ASCO

The field has been measuring this for over a decade.

6 to 7

times in ten that two qualified reviewers agree on the same toxicity

31 to 36%

of attributions change between investigator and central review

215

days, the median window in which sites run mixed protocol versions

That is the published baseline for doing this by hand, measured across the whole field over more than a decade. Burna applies the same criteria the same way every time, and shows you which criterion it applied.

Sources: Hillman 2010; Le-Rademacher 2017; Hong 2020; Tufts CSDD; FDA/NCI attribution workshop 2019.

The event happens on day 9. The visit is on day 21.

A note written at the next appointment is a recollection. Healthnotes captures what happened in your patient’s own words, on the phone they already own, and it arrives as a graded, coded event with attribution attached.

twelve days of recollection

DAY 9

The fever starts

On a Saturday. She waits, because it might pass.

DAY 21

The visit

She is asked to remember a fortnight, and does her best.

Twelve days a site currently has no record of. Illustrative rendering.

AI suggests, clinicians decide. Every grade is reviewed, editable, and signed under 21 CFR Part 11. The clinical decision stays with your clinician.

Read three pages, then take 15 minutes with the founder.