Skip to main content

Screen patients for clinical trial eligibility

Screen your patient population against a trial's inclusion and exclusion criteria using facts extracted from clinical notes, not diagnosis codes alone, then save the matching patients as a named cohort the study team can work from.

Clinical Trial Matching and Eligibility Screening

Who this is for​

Anyone who has to turn eligibility criteria into a real list of patients: clinical research coordinators and study teams running pre-screening, trial feasibility analysts answering site capacity questions, clinical operations and site selection teams at sponsors and CROs, principal investigators assessing whether a protocol is enrollable locally, and the research IT and clinical data teams supporting the study portfolio.

Why it matters​

A sponsor asks whether your site can enroll 40 patients with HER2-low metastatic breast cancer, prior CDK4/6 inhibitor exposure, and ECOG performance status 0 to 1, within six months. Your structured data can answer "metastatic breast cancer." It cannot answer HER2-low, because that reading sits in a pathology narrative, and it cannot answer ECOG, because that sits in an oncology progress note.

Sites that answer feasibility from diagnosis codes alone commit to enrollment targets built on a population that is larger on paper than in the clinic. The correction arrives months later as a missed enrollment milestone.

The same gap appears outside oncology. Cardiology criteria turn on ejection fraction and NYHA class, neurology on symptom onset and prior lines of therapy, rare disease on findings documented in a single specialist letter.

What you gain​

Screening becomes repeatable. When the sponsor amends the protocol, you re-run against the revised criteria rather than restarting chart review. The matched population is saved as a named cohort, so the same definition feeds the screening log, the dataset you hand the study team, and the feasibility answer you send the sponsor. Nothing is re-derived by hand at each step.

Before you start​

  • Patient documents and structured data ingested, including the note types that carry eligibility detail: pathology, oncology and specialty progress notes, imaging reports
  • The trial's inclusion and exclusion criteria in hand
  • Sharing permissions set for the study team members who will receive the cohort or dataset

Step 1: Match patients against a trial​

  1. In the left navigation, go to Agents and Tools > Clinical Trial Matcher.
  2. Start a new trial matching workflow.
  3. Search for a clinical trial by trial identifier, title, condition, intervention, sponsor, or keyword.
  4. If a custom study is being evaluated, paste the trial eligibility criteria into the matcher.
  5. Review the parsed inclusion and exclusion criteria.
  6. Confirm which criteria can be evaluated against the available patient data.
  7. Adjust or remove criteria that are not relevant to the local screening workflow, where criteria editing is available.
  8. Select the target patient population or source scope for matching.
  9. Start the trial matching process.
  10. Wait until the trial job status is Completed.
  11. Review the resulting patient list.

Step 2: Review match results and save matching patients as a cohort​

  1. Open the completed trial matching results.
  2. Review the included, excluded, and potentially eligible patient groups.
  3. Inspect the criteria-level explanations for a sample of matched and excluded patients.
  4. Confirm that the match logic aligns with the intended feasibility or recruitment use case.
  5. Select the patients that should be saved for follow-up review.
  6. Click Create Cohort to begin cohort creation.
  7. Enter the cohort name, description, category, and any linked cohort information.
  8. Use a clear name that references the trial or study objective.
  9. Review the cohort configuration and patient count.
  10. Click Create Cohort to save the matched patient cohort.

Step 3: Build a dataset from the trial cohort​

  1. In the left navigation, go to Data Curation Studio > Dataset Builder.
  2. Select Cohort & Clinical Filter as the dataset mode.
  3. Select the data type required for study review, such as Unstructured for clinical notes and reports.
  4. Click Next to continue to Cohort Selection.
  5. Select the trial cohort created from the matching results.
  6. Click Next to continue to the criteria step for the selected data type.
  7. Apply document filters to include the evidence needed for screening, such as pathology reports, oncology notes, radiology reports, medication records, procedure notes, or lab-related documents.
  8. Click Next to continue to Dataset Details.
  9. Enter the dataset name and description.
  10. Use a name that clearly links the dataset to the trial or feasibility review.
  11. Click Next to continue to Review Dataset.
  12. Review the selected cohort, document filters, data types, and expected dataset contents.
  13. Click Create Dataset to create the dataset.

Step 4: Review documents or run extraction to support screening​

  1. Open the created dataset from Dataset Explorer.
  2. Review the included patients and supporting documents.
  3. Confirm that the dataset contains the evidence needed to validate key eligibility criteria.
  4. Open patient documents where manual review is required.
  5. If targeted evidence needs to be extracted, run Information Extraction on the dataset for trial-relevant facts such as biomarkers, stage, prior therapies, disease progression, lab values, or performance status.
  6. Review extraction results before using them for screening decisions.
  7. Use extracted facts as reviewer context or to refine the dataset, depending on the supported workflow.
  8. Document any criteria that cannot be evaluated from the available data.

Step 5: Share the cohort or dataset with the study team​

  1. Review the saved cohort and dataset before sharing.
  2. Confirm that the patient list, evidence scope, and document filters match the study-team request.
  3. Share the cohort with the study team.
  4. Share the dataset with the study team when deeper document review is needed.
  5. Provide the study team with the cohort, dataset, extracted evidence, or review notes according to the approved team workflow.

Recipe reference​

Each stage of this scenario is also a reusable building block.

These steps have their own pages:

Match patients to a clinical trial and reuse the cohort​

When to use on its own: For trial feasibility, recruitment planning, and pre-screening across therapeutic areas.
Features involved: Clinical Trial Matcher, Cohort Builder, patient list review, saved cohorts.
Edge cases / limitations: Some trial criteria are too vague or unavailable in the source data. Results may need clinical validation.
Value: Moves trial feasibility from manual chart review toward a repeatable screening workflow.

Run Information Extraction on a dataset​

When to use on its own: When you need targeted extraction such as biomarkers, pathology details, staging terms, adverse events, or custom clinical facts.
Features involved: Information Extraction, custom extraction profiles, extraction run history, result review and export.
Edge cases / limitations: Extraction quality depends on source document quality and field definitions. Results should be reviewed before customer-facing or clinical use.
Value: Lets you quickly discover useful facts before committing to a larger curation workflow.