Free · No signup · Source-linked
Find evidence and compose a traceable CER workpaper.
Search PubMed and ClinicalTrials.gov, preserve a reproducible Evidence Pack, and compose source-linked screening, study, outcome, appraisal, and claim-evidence tables.
PubMed literatureClinicalTrials.gov device studiesR2-backed reproducible runsStudy and claim-evidence tables
Define the evidence context
Start with the device identity. Indication, population, and known synonyms make the search more specific without hiding the actual query.
Try an example
About the data
A data product and evidence workpaper before document generation
Each result remains traceable to its official source. The pack preserves the exact search strategy and identifiers. Workpaper composition runs in durable batches and stores every intermediate stage, ready artifact, quality report, and commit manifest in R2 rather than hiding transformation state in memory.
- Does this generate a finished CER?
- No. It first creates a reproducible candidate evidence set, then can compose a source-linked workpaper covering search, screening, studies, outcomes, appraisal, claims, and gaps. The output remains a bounded draft input to CER authoring.
- Where do the records and stored artifacts come from?
- Article metadata comes from the official NCBI PubMed APIs and device-study records from the maintained ClinicalTrials.gov dataset. The submitted context, source responses, ready pack, quality report, and manifest are stored separately in R2.
- How are articles connected to trials?
- Only explicit ClinicalTrials.gov accession identifiers published in the PubMed record are used. The system does not infer a link from similar wording.