In the box
Oncology data from the EHR, with biomarkers and staging as named columns.
Clinical Extract reads biomarker and staging Observations by their LOINC codes, and genomic results by the gene studied. Each result lands in a named column: ER, PR, HER2, PD-L1, EGFR, KRAS, BRAF, ALK, MSI and TNM, ready to analyze.
1 Biomarkers
Each biomarker result becomes a named column.
Oncology data extraction from an EHR is mostly Observations: a receptor status here, a mutation there, each a coded resource with its own LOINC code. The oncology lens reads them by code and by gene and writes each result to a named column with the code system in its description. A study's biomarker data extraction ends as a set of columns, one per marker.
2 Staging
TNM staging data, clinical and pathologic.
- Clinical stage
- cT, cN, cM and the clinical stage group (LOINC 21908-9), from the staging Observations at diagnosis.
- Pathologic stage
- pT, pN, pM and the pathologic stage group, from the post-surgical staging Observations.
- Edition and date
- The AJCC edition where the record carries it, and the date of each staging Observation, as their own columns.
- Histology and grade
- The morphology code from Condition and the grade Observation, beside the stage.
3 Treatment
The first course, coded.
Systemic therapy comes from MedicationRequest by RxNorm: the first chemotherapy agent ordered and its date, the regimen's agents as their own columns. Surgery comes from Procedure by SNOMED CT or CPT; radiation from the procedure and its dates. A registry maps them to its first-course items through the crosswalk on the registry page. The same columns pre-fill a registry abstract, as in abstracting a breast case from coded fields.
4 mCODE
Reads mCODE FHIR profiles and plain LOINC-coded Observations.
mCODE, HL7's minimal Common Oncology Data Elements, profiles the cancer diagnosis, staging, biomarkers, treatment and outcomes on FHIR resources. An EHR that writes mCODE-profiled Observations is read by profile; one that writes plain LOINC-coded Observations is read by code; the columns fill either way. For questions about FHIR versions themselves, our comparison of the FHIR releases is the reference.
5 Questions
Questions about oncology data
What is biomarker data extraction?
Pulling the biomarker results a study or a registry needs out of the EHR as structured values: the receptor statuses, the molecular markers, the gene results. In FHIR they are Observations; the oncology lens reads them by LOINC code or by the gene studied and writes each one to a named column.
Which biomarkers come out as columns?
ER, PR, HER2 and PD-L1 from protein results; EGFR, KRAS, BRAF, ALK and MSI from genomic results; plus the tumor markers the protocol names. Each column’s LOINC code and source are in the data dictionary.
What TNM staging data is exported?
Clinical and pathologic T, N and M with the stage group, from the staging Observations (LOINC 21908-9 for the clinical stage group among them), with the AJCC edition where the record carries it, each as its own column.
What is mCODE?
HL7’s minimal Common Oncology Data Elements, a FHIR implementation guide that profiles the cancer diagnosis, staging, biomarkers, treatment and outcomes. The oncology lens reads mCODE-profiled Observations and plain LOINC-coded ones alike, so a site’s columns fill whether or not its EHR writes mCODE.
Does it cover genomic results?
Genomic Observations are read by the gene studied (EGFR, KRAS, BRAF, ALK, MSI) and written to their own columns, with the source and an example in the dictionary.
Next
See Clinical Extract run on one of your studies.
Tell us which EHR you run and what the study or registry needs. We reply within one business day to set a meeting time.