EVERY RESULTNot an eligibility
decision.
Every label ends in “review required”, even when every criterion is decided. The score is a share of criteria, not a probability. A clinician checks each one against the current ClinicalTrials.gov record.
Criteria met — review requiredPotential match — review requiredLikely excluded — review required
Patient matchrules · Jev · Groq
INCLUSIONAge 18 to 80 yearsMetage: 61
INCLUSIONStage II (high-risk) or IIIUnknownno fact settles this
EXCLUSIONPrior anti-HER2 therapyMet · blockingmedications: Trastuzumab
EXCLUSIONPregnant or breastfeedingN/Asex: MALE
Judged on your facts alone.
A model’s answer counts only when it quotes one of your facts word for word. PubMed, codes and drug names never touch a match.
Published researchPubMed
P1PMID 39282896Trastuzumab deruxtecan after endocrine therapy in metastatic breast cancer
Kept apart from trial facts.
Abstracts from NCBI PubMed, only when you switch it on. Cited as P1, P2…, every quote checked against the abstract.
Diagnosis codesICD-10-CM · NLM
C56.1 right ovary
C56.2 left ovary
C56.9 unspecified ovary
Listed, never picked for you.
Looked up only when you ask for a code. The right one depends on your record, so Orvant shows the options.
Drug namesRxNorm · NLM
Brand names understood.
Ask about trials “after Tagrisso” and the search also looks for osimertinib, then tells you it did.
What an outside search receivesdirect to NCBI and NLM
I’m a 46 year old woman in Austin with HER2-low breast cancer — any papers on trastuzumab deruxtecan?
Only topic words leave.
Ages, places, pronouns, emails and long numbers are removed first. Your profile is never sent, and no third-party hosted service sits in between.
