REFERENCE / ANNOTATED EHR EXAMPLE
Annotated oncology EHR example
One synthetic lung-cancer patient, told every way it travels — chart screen, clinic notes, codes, claims, FHIR, and JSON — with every caveat annotated.
Chart view
Unstructured notes
FHIR view
Open FHIR bundle
Open JSON transport
Treatment failure concept
ECOG score concept
EHR provenance concept
Patient token
SYN-ONC-0007
Cancer
Malignant neoplasm of upper lobe, right bronchus or lung
Stage / biomarker
Stage IVB; EGFR exon 19 deletion
Linkage confidence
0.94
How to read this page
One synthetic Stage IVB lung-cancer patient travels the full length of the real-world-data pipeline on this page. The record appears five ways: the chart screen a clinician works in, the unstructured notes behind it, the human-readable source records, the FHIR bundle a health system would exchange, and the flat JSON transport a research dataset is built from. Every view is annotated with what each element does — and does not — tell a researcher.
The gap between what the clinician saw and what your dataset contains is where misclassification, missingness, and index-date error come from. The point of this page is to make that gap visible, element by element: the stage that lives only in a note, the ECOG score most patients never get, the stop date that lags the last dose, the ED visit that vanishes into an inpatient claim.
Suggested path: read the chart mock first, then find the same facts in the FHIR view, then use the mapping table to connect any chart element to its FHIR path and its research use.
Educational example only. Every patient, provider, date, and identifier on this page is synthetic. The record is modeled on realistic EHR artifacts for methods education. It is not medical advice and not patient data.
Patient timeline
DiagnosisStage IVB
Outpatientnew lung mass workup
Imaging summarybaseline metastatic disease
Ancillarybaseline staging CT chest/abdomen/pelvis
Ancillarytumor NGS panel resulted
Creatinine0.9 mg/dL
Start LOT 1osimertinib
ECOG Performance Status1
Outpatientsystemic therapy start
CEA21.4 ng/mL
Ancillarypre-visit labs and tumor marker
Oxygen saturation89 %
Emergencyshortness of breath and right-sided pleuritic chest pain
Ancillaryultrasound-guided thoracentesis
Imaging summaryprogressive disease
Ancillaryrestaging CT chest/abdomen/pelvis
Stop therapyradiographic progression / non-response
ECOG Performance Status2
Outpatientprogression visit and treatment change
Candidate failure dateprobable treatment failure / non-response
Start LOT 2carboplatin + pemetrexed + pembrolizumab
Outpatientcycle 1 day 1 chemoimmunotherapy
The chart as the clinician sees it
This is the same record presented the way an EMR (electronic medical record — the software clinicians chart in) shows it on screen. The numbered markers point at chart elements; the notes below each explain what that element does and does not tell a researcher.
Problem list
| Problem | Code | Status | Onset |
|---|---|---|---|
| Malignant neoplasm of upper lobe, right bronchus or lung 2 | C34.11 | active 3 | 2024-01-18 |
| Malignant pleural effusion | J91.0 | active | 2024-07-21 |
| Shortness of breath | R06.02 | resolved | 2024-07-21 |
Medications
| Drug | Dose | Dates | Status |
|---|---|---|---|
| osimertinib | 80 mg oral daily | 2024-02-09 → 2024-08-05 | stopped 4 |
| carboplatin + pemetrexed + pembrolizumab | 2024-08-12 → active | active |
Results flowsheet
| Date | Result | Value | Ref range |
|---|---|---|---|
| 2024-02-05 | Creatinine | 0.9 mg/dL | 0.6-1.2 |
| 2024-07-15 | CEA | 21.4 ng/mL 5 | 0.0-3.0 |
| 2024-07-21 | Oxygen saturation | 89 % | |
| 2024-02-09 | ECOG Performance Status | 1 6 | |
| 2024-08-05 | ECOG Performance Status | 2 | |
| 2024-01-24 | Imaging summary | baseline metastatic disease | |
| 2024-08-01 | Imaging summary | progressive disease |
Recent encounters
| Date | Setting | Reason |
|---|---|---|
| 2024-07-21 | emergency 8 | shortness of breath and right-sided pleuritic chest pain |
| 2024-07-21 | inpatient | malignant pleural effusion |
| 2024-07-22 | ancillary | ultrasound-guided thoracentesis |
| 2024-08-01 | ancillary | restaging CT chest/abdomen/pelvis |
| 2024-08-05 | outpatient | progression visit and treatment change |
| 2024-08-12 | outpatient | cycle 1 day 1 chemoimmunotherapy |
oncology consultation note · 2024-01-18
Thoracic oncology consult for persistent cough and CT-detected right upper-lobe mass. Biopsy from outside hospital reviewed as lung adenocarcinoma. Staging workup ordered; baseline ECOG not yet documented. 7
- 1Patient banner. The banner shows a synthetic token, age, and sex — no name, no birth date. A real chart shows full identity; a research extract shows this.RWEDe-identification happens before data reach the researcher; identity survives only as linkage tokens.
- 2Problem list entry. The problem list pairs a narrative label with an ICD-10-CM code. The problem list is curated by clinicians and is not the same as billing diagnoses.RWEProblem-list codes and claims diagnoses disagree often; case definitions should say which source they use.
- 3Problem status. Each problem carries a status (active / resolved) that clinicians update — or forget to update.RWEStale statuses are a known data-quality trap; do not treat 'active' as verified current disease.
- 4Medication status. The chart shows osimertinib as stopped with a documented reason. The stop date is what the prescriber recorded, not necessarily the last day taken.RWEDiscontinuation endpoints inherit this documentation lag; sensitivity analyses should widen the stop window.
- 5Lab result with flag. CEA 21.4 ng/mL displays with the lab's reference range. The H flag comes from the lab feed, not the EMR.RWEReference ranges vary by lab and over time; harmonize units and ranges before pooling.
- 6ECOG in the flowsheet. ECOG performance status appears because the oncologist scored it at a visit. Many charts never have one.RWEPerformance status is a strong oncology confounder with heavy, non-random missingness.
- 7Clinical note. The note narrative holds details that never reach structured fields: outside pathology review, staging language, clinician reasoning.RWEAbstraction and NLP work on this text; structured-only studies never see it.
- 8ED encounter row. The encounter list shows the care setting for each contact. The ED visit on 2024-07-21 precedes the inpatient stay by hours.RWEUtilization endpoints are built from these setting labels and timestamps; same-day transitions must be merged into episodes.
Unstructured notes
Most of an EHR is not fields — it is documents. These four notes carry the same synthetic patient in the form NLP pipelines and chart abstractors actually receive: abbreviations, copy-forward, unexpanded template macros, a misspelling, and results that arrive after the document is signed. The numbered markers flag the artifacts; the notes below each document explain what they do to research. In the FHIR view, documents like these travel as the DocumentReference resource.
Oncology new-patient consult2024-01-18 · Thoracic medical oncology attending (synthetic)
THORACIC MEDICAL ONCOLOGY — NEW PATIENT CONSULT Date of Service: 01/18/2024 14:32 Dept: Thoracic Onc Clinic Patient: SYNTEST, RWE A MRN: ehrtok_9b7a2d DOB: **/**/**** (66 F) CC: persistent cough x3 mo, RUL mass on OSH CT HPI: 66F, never-smoker, p/w 3 months progressive dry cough and mild DOE. OSH CT chest 12/2023: 3.8 cm RUL mass, mediastinal LAD. OSH CT-guided bx 01/09/2024, slides reviewed at our institution: adenocarcinoma, TTF-1+, c/w lung primary. Referred for onc eval. Denies hemoptysis. Wt loss ~4 lb. No bone pain, no HA, no focal deficits. PMH: HTN, OA. PSH: cholecystectomy. FamHx: mother breast ca, dx 70s. SocHx: never smoker, retired teacher, lives w/ spouse. EtOH occasional. MEDS (reconciled): amlodipine 5 mg PO qd; APAP prn. NKDA. ROS: 10-pt ROS negative except as per HPI. EXAM: T 97.9 HR 82 BP 142/84 RR 16 SpO2 96% RA. NAD, A&Ox3. Lungs: faint end-insp crackles R upper field. CV: RRR, no m/r/g. No supraclavicular LAD. Ext: no edema, no clubbing. DATA: OSH CT report reviewed (images pending upload to PACS). Path per above. Staging labs pending. A/P: 66F w/ newly dx lung adenocarcinoma, RUL primary, radiographic mediastinal LAD — clinically at least IIIB, suspect IV pending staging 1. 1. Staging CT C/A/P w/ contrast + MRI brain 2. Tumor NGS panel + PD-L1 (tissue from OSH block) 3. RTC 2 wks to review; d/w thoracic tumor board 4. Performance status: not formally assessed this visit — appears fit, est 0-1 2 .onclastvisit 3 Electronically signed by ATTENDING, SYNTHETIC MD 01/18/2024 17:05
- 1The structured field stage_at_diagnosis says Stage IVB and carries this note’s date, but the clinician wrote "suspect IV pending staging" and the confirmatory CT was 01/24. The record holds no stage-entry timestamp, so whether the value was back-filled cannot be proven from the data — only suspected.RWEDate the stage assignment, not the diagnosis. Without a provenance timestamp you cannot tell a contemporaneous value from one back-filled with hindsight — so ask the data partner for one.
- 2Performance status appears only as an eyeball estimate in prose. The first scored ECOG enters the flowsheet at treatment start, 02/09.RWEBaseline ECOG missingness starts exactly here — 'appears fit' is not a scored covariate.
- 3An unexpanded dot-phrase — a template macro that never fired. Real notes are full of this debris.RWENLP pipelines must survive template artifacts; their frequency is itself a documentation-quality signal.
ED provider note2024-07-21 · Emergency medicine attending (synthetic)
EMERGENCY DEPARTMENT PROVIDER NOTE DOS: 07/21/2024 Arrival 18:42 Triage acuity: ESI 2 Patient: SYNTEST, RWE A (66 F) MRN: ehrtok_9b7a2d Triage vitals 18:48: T 98.4 HR 118 BP 134/78 RR 26 SpO2 89% RA -> 94% on 2L NC 1 CC: "can't catch my breath" HPI: 66F w/ stage IV lung adenoca (EGFR ex19del) on osimertinib since 02/2024, p/w 2 days worsening SOB and R-sided pleuritic CP. No fever/chills. No hemoptysis. Worse lying flat. Denies unilateral leg swelling. Labs drawn 07/15 ahead of onc f/u. EXAM: uncomfortable, speaking in short sentences. Decreased BS R base w/ dullness to percussion. Tachycardic, regular. No JVD. Ext no edema. ED COURSE / MDM: CXR: large R pleural effusuion 2. CTA chest: no PE; large R effusion w/ compressive atelectasis; interval increase in bilateral pulm nodules c/w progression of metastatic dz. Morphine 2 mg IV 19:10 w/ relief. DDx: malignant effusion vs parapneumonic — afebrile, no consolidation, malignant favored 3. D/w hospitalist — admit to medicine, onc notified. IP/thoracics to eval re: drainage in AM. DISPO: admit to medicine, floor bed 23:58. Admit dx: malignant pleural effusion (J91.0). Electronically signed by EDATTENDING, SYNTHETIC MD 07/21/2024 23:41
- 1The structured vitals table captures the single value 89%. The note holds the trajectory and the oxygen response.RWESeverity adjustment from structured vitals alone loses the clinical arc the note preserves.
- 2A misspelling, left in on purpose — real notes contain them.RWEExact-string and dictionary NLP miss misspelled mentions; this is one reason text-mined phenotypes undercount.
- 3The diagnosis at disposition is a weighted hypothesis, not a certainty. The admit code J91.0 crystallizes it anyway.RWEDiagnosis codes inherit whatever hypothesis was leading when the code was assigned; cytology confirmation comes days later.
Discharge summary2024-07-24 · Hospital medicine attending (synthetic)
DISCHARGE SUMMARY Admit: 07/21/2024 Discharge: 07/24/2024 Service: Hospital Medicine Consultants: Medical Oncology, Interventional Radiology Admit dx: malignant pleural effusion (J91.0) Discharge dx: malignant pleural effusion; lung adenocarcinoma stage IV, progression suspected HOSPITAL COURSE: 66F w/ stage IV EGFR+ lung adenoca on osimertinib p/w "2 days worsening SOB and R-sided pleuritic CP" (per ED note) 1. Found to have large R malignant-appearing effusion; CTA negative for PE. 07/22 IR performed US-guided thoracentesis, 1.2 L serosanguinous fluid removed, cytology sent — pending at DC 2. Symptoms improved; SpO2 94% RA at rest by 07/23. Onc consulted — concern for progression on osimertinib given interval nodule growth; outpatient restaging CT arranged, onc f/u 08/05. DISCHARGE MEDS: continue osimertinib 80 mg daily pending 3 onc f/u; amlodipine 5 mg qd; oxycodone 5 mg q6h prn pain, #12 dispensed. DISPO: home w/ home health nursing. F/u: oncology 08/05; restaging CT prior to visit. Patient educated on return precautions, verbalized understanding. [x] Electronically signed by HOSPITALIST, SYNTHETIC MD 07/24/2024 11:22
- 1Copy-forward: the HPI phrase is pasted verbatim from the ED note three days earlier, quotation marks and all.RWECopy-forward inflates term frequency across documents; NLP that counts mentions overweights duplicated text.
- 2Cytology resulted after discharge. The discharge summary is frozen before the facts finish arriving.RWEResults that post-date a document are invisible to it — link labs by result date, never by the note that ordered them.
- 3The discharge med list says continue. The structured record stops the drug 12 days later.RWEMedication-list snapshots are not exposure. Reconcile lists against orders and dispensings before building exposure windows.
Oncology progress note (progression visit)2024-08-05 · Thoracic medical oncology attending (synthetic)
MEDICAL ONCOLOGY PROGRESS NOTE — f/u, progression eval DOS: 08/05/2024 15:40 Dept: Thoracic Onc Clinic Patient: SYNTEST, RWE A (66 F) MRN: ehrtok_9b7a2d INTERVAL HISTORY: s/p admission 07/21-07/24 for malignant R effusion, s/p thoracentesis 1.2 L (cytology: adenocarcinoma, TTF-1+, c/w known lung primary). Breathing improved but fatigue worse — now needs rest after one flight of stairs. DATA REVIEWED: CEA 21.4 (07/15), prior 8.2 — rising. Restaging CT C/A/P 08/01: increased size and number of bilateral pulm nodules, new hepatic lesion, recurrent moderate R effusion. IMPRESSION: progressive metastatic disease compared with 01/24/2024 baseline. EXAM: ECOG 2 today (was 1 1 at treatment start). Decreased BS R base. Otherwise unremarkable. A/P: 66F stage IVB lung adenocarcinoma, EGFR exon 19 del, PROGRESSION on 1L 2 osimertinib (started 02/09/2024) — radiographic + clinical + tumor marker. 1. STOP osimertinib today. 2. Discussed 2L options; plan carboplatin AUC 5 + pemetrexed + pembrolizumab 3 q3wk, C1D1 08/12. Benefits/risks reviewed; consent signed. 3. Labs incl CMP prior to C1D1; calculate Cockcroft-Gault CrCl — give pemetrexed only if CrCl >= 45 mL/min. Carboplatin dosed to AUC by renal function. 4. Consider re-biopsy for resistance profiling (T790M/C797S 4) — pt deferring, revisit after cycle 2. 5. Palliative care referral placed. Electronically signed by ATTENDING, SYNTHETIC MD 08/05/2024 17:58
- 1The functional decline supporting the failure classification lives here and in the flowsheet on the same date.RWEThis note is the source-trace anchor for the ecog-2 observation — provenance you can point an auditor at.
- 2The explicit clinician assessment of treatment failure — the sentence abstraction teams hunt for.RWEThe constructed treatment-failure endpoint (candidate date 08/05) is anchored to exactly this documentation; without the note, the endpoint rests on inference.
- 3This regimen is off-guideline. ASCO recommends against adding anti-PD-(L)1 therapy after progression on an EGFR TKI, and KEYNOTE-789 found no significant PFS or OS benefit. It is here on purpose: real-world data is full of off-guideline care, and this note documents the reasoning that produced it.RWENever assume the observed regimen is the recommended one. Comparative-effectiveness cohorts built from real-world treatment patterns must handle guideline-discordant arms explicitly, not silently drop them.
- 4Resistance testing was considered and deferred — a planned-but-not-done event that leaves no structured trace.RWEAbsence of a resistance result is not absence of resistance. Deferrals are a censoring mechanism only the note records.
Human-readable source records
Each row is synthetic but modeled on realistic EHR artifacts: clinic notes, ED triage/provider documentation, inpatient discharge summaries, radiology reports, laboratory interfaces, procedure notes, pathology addenda, and infusion administration records.
Outpatient records
| Date | Record | Human-readable source text | Coded elements | RWE use and caveat |
|---|---|---|---|---|
| 2024-01-18 | oncology consultation note | Thoracic oncology consult for persistent cough and CT-detected right upper-lobe mass. Biopsy from outside hospital reviewed as lung adenocarcinoma. Staging workup ordered; baseline ECOG not yet documented. | ICD-10-CM C34.11: Malignant neoplasm of upper lobe, right bronchus or lung; CPT 99205: New patient office visit, high complexity | Incident disease confirmation, specialty-care cohort entry, baseline date context.CaveatNarrative stage terms require abstraction; outside-pathology date may differ from local EHR diagnosis date. |
| 2024-02-09 | oncology treatment-start note | Patient started osimertinib 80 mg by mouth daily for EGFR-mutant metastatic NSCLC. ECOG performance status 1. Baseline creatinine 0.9 mg/dL. Consent documented; toxicity counseling completed. | RxNorm 1721581: osimertinib 80 MG Oral Tablet; LOINC 89247-1: ECOG performance status | Index date, first-line exposure start, baseline functional status.CaveatMedication order and specialty-pharmacy dispense agree on start month but not actual ingestion. |
| 2024-08-05 | oncology progression visit | Reviewed restaging CT and recent hospitalization. Documented progressive disease on osimertinib with ECOG 2. Osimertinib discontinued. Plan: carboplatin, pemetrexed, and pembrolizumab after lab clearance. | RxNorm 1721581: osimertinib 80 MG Oral Tablet; LOINC 89247-1: ECOG performance status | Candidate failure date, discontinuation reason, and clinical non-response classification.CaveatIf stop reason is not structured, NLP or chart abstraction is required to distinguish progression from toxicity. |
| 2024-08-12 | infusion administration record | Cycle 1 day 1 administered: pembrolizumab 200 mg IV, pemetrexed 500 mg/m2 IV, carboplatin AUC 5 IV. Premedications given; infusion completed without reaction. | HCPCS J9271: Injection, pembrolizumab, 1 mg; HCPCS J9305: Injection, pemetrexed, 10 mg; HCPCS J9045: Injection, carboplatin, 50 mg | Observed next-line therapy administration; stronger than order-only evidence for infused therapies.CaveatDose units differ across HCPCS billing, MAR/eMAR, and pharmacy compounding systems. |
Emergency department records
| Date | Record | Human-readable source text | Coded elements | RWE use and caveat |
|---|---|---|---|---|
| 2024-07-21 | ED triage, provider note, and facility charge | Arrived to ED at 18:42 with 2 days of worsening dyspnea and right-sided pleuritic pain. Triage: ESI 2, SpO2 89% on room air, HR 118, RR 26. Chest x-ray showed enlarging right pleural effusion. Treated with oxygen and IV morphine; admitted to oncology floor. | ICD-10-CM R06.02: Shortness of breath; UB-04 revenue center 0450: Emergency room; CPT 99285: Emergency department visit, high severity | Acute-care utilization, symptom onset, and progression-related escalation of care.CaveatED chief complaint is symptom-level and should not be treated as a confirmed endpoint without inpatient/radiology follow-up. |
Inpatient records
| Date | Record | Human-readable source text | Coded elements | RWE use and caveat |
|---|---|---|---|---|
| 2024-07-21/2024-07-24 | inpatient admission, daily notes, and discharge summary | Admitted from ED for hypoxemia due to malignant right pleural effusion. Hospital course: oxygen, ultrasound-guided thoracentesis removing 1.2 L serosanguinous fluid, oncology consult concerned for progression on osimertinib. Discharged home with home-health nursing and oncology follow-up. | ICD-10-CM J91.0: Malignant pleural effusion; MS-DRG 181: Respiratory neoplasms with MCC; UB-04 type of bill 0111: Hospital inpatient admit-through-discharge claim | Inpatient outcome, length of stay, severity, procedure utilization, and post-index resource use.CaveatAdmission date follows ED date; same-day ED facility charges may be bundled into the inpatient claim for some payers. |
Ancillary services records
| Date | Record | Human-readable source text | Coded elements | RWE use and caveat |
|---|---|---|---|---|
| 2024-01-24 | radiology report | CT chest/abdomen/pelvis with contrast: 4.2 cm right upper-lobe mass, multiple bilateral pulmonary nodules, small right pleural effusion, and two liver lesions suspicious for metastases. Impression: metastatic lung cancer, baseline disease burden. | LOINC 24627-2: CT Chest Abdomen Pelvis WO and W contrast; CPT 71260: CT thorax with contrast | Baseline tumor burden and pre-treatment imaging anchor.CaveatRECIST target-lesion measurements are not structured; only radiology narrative is available. |
| 2024-02-02 | molecular pathology result | Tumor NGS addendum: EGFR exon 19 deletion detected; ALK and ROS1 rearrangements not detected; PD-L1 TPS 5%. Report electronically signed by molecular pathology. | LOINC 81247-9: Master HL7 genetic variant reporting panel; HGVS-like EGFR exon 19 deletion: Actionable EGFR alteration | Biomarker-defined cohort eligibility and treatment selection rationale.CaveatExact variant nomenclature is simplified for the synthetic example; real pipelines should retain HGVS and assay metadata. |
| 2024-07-15 | outpatient laboratory panel | Pre-visit labs: creatinine 1.1 mg/dL, AST 42 U/L, hemoglobin 10.8 g/dL, CEA 21.4 ng/mL, up from 8.2 ng/mL in May. Results interfaced from hospital laboratory. | LOINC 2160-0: Creatinine [Mass/volume] in Serum or Plasma; LOINC 2039-6: Carcinoembryonic Ag [Mass/volume] in Serum or Plasma | Tumor-marker progression support and safety-monitoring covariates.CaveatReference ranges and assay platform are not harmonized across sites unless mapped separately. |
| 2024-07-22 | interventional radiology procedure note | Ultrasound-guided right thoracentesis performed; 1.2 L serosanguinous pleural fluid removed. Cytology sent. No immediate complication. | CPT 32555: Thoracentesis with imaging guidance; SNOMED CT 39607008: Thoracentesis | Procedure confirmation, complication severity, and evidence that effusion was clinically actionable.CaveatProcedure may appear in inpatient procedure tables, professional claims, and radiology systems with different timestamps. |
| 2024-08-01 | restaging radiology report | CT chest/abdomen/pelvis: increased size and number of bilateral pulmonary nodules, new liver lesion, and recurrent moderate right pleural effusion. Impression: progressive metastatic disease compared with 2024-01-24 baseline. | LOINC 24627-2: CT Chest Abdomen Pelvis WO and W contrast; RECIST-like abstraction PD: Progressive disease | Primary radiographic support for treatment-failure/non-response endpoint.CaveatRECIST-like category is abstracted from narrative; raw lesion measurements are incomplete. |
Treatment failure evidence
- medrx-1MedicationRequest stopped for progression
- medrx-2New line of therapy started within 7 days
- img-1Restaging CT impression says progressive metastatic disease
- ecog-2ECOG worsened from 1 to 2
- enc-6Emergency encounter for dyspnea and pleural effusion
- enc-7Inpatient admission for malignant pleural effusion
- admin-1Observed second-line infusion administration
Patient context
| Age at index | 66 |
|---|---|
| Sex | female |
| Race / ethnicity | Black or African American; Not Hispanic or Latino |
| Region | US-South |
| Payer type | commercial |
| Subscriber / member | sub_44fd19 / mem_772ac1 |
Treatment and observations
Lines of therapy
| LOT | Drug | Start | Stop | Status |
|---|---|---|---|---|
| 1 | osimertinib | 2024-02-09 | 2024-08-05 | radiographic progression / non-response |
| osimertinib | active | |||
| morphine | active | |||
| 2 | carboplatin + pemetrexed + pembrolizumab | 2024-08-12 | active | active |
| 2 | pembrolizumab + pemetrexed + carboplatin | active | completed |
Observations
| Date | Observation | LOINC | Value | RWE use |
|---|---|---|---|---|
| 2024-02-05 | Creatinine | 2160-0 | 0.9 | Baseline renal covariate. |
| 2024-07-15 | CEA | 2039-6 | 21.4 | Tumor-marker progression support. |
| 2024-07-21 | Oxygen saturation | 59408-5 | 89 | ED severity marker and acute-care eligibility covariate. |
| 2024-02-09 | ECOG Performance Status | 89247-1 | 1 | Baseline functional-status covariate. |
| 2024-08-05 | ECOG Performance Status | 89247-1 | 2 | Clinical deterioration and endpoint-support signal. |
| 2024-01-24 | Imaging summary | baseline metastatic disease | Baseline imaging comparator. | |
| 2024-08-01 | Imaging summary | progressive disease | Clinical response component; not RECIST-confirmed unless source report is available. |
Provenance and data quality
Source trace
| Derived field | Trace |
|---|---|
| line of therapy | MedicationRequest sequence, specialty-pharmacy dispense, and infusion MedicationAdministration |
| treatment failure non response | Medication status reason, restaging imaging, ECOG change, ED/inpatient utilization, next-line therapy |
| claims | Payer extract line-level claims linked to encounter IDs |
| ecog | Structured observation using LOINC 89247-1 |
| ed episode | ED triage note, vitals, facility revenue center, and admission handoff |
| inpatient episode | Admission/discharge timestamps, principal diagnosis, DRG-like grouping, and procedure record |
| ancillary services | Radiology, laboratory, pathology, and procedure systems linked back to encounters |
Data quality flags
| Flag | Severity | Affects |
|---|---|---|
| synthetic_record | info | all fields |
| date_shifted | info | calendar-time interpretation |
| imaging_not_recist_confirmed | medium | treatment_failure_non_response |
| claims_ehr_linkage_probabilistic | medium | claims_linkage |
| oral_adherence_unobserved | medium | osimertinib exposure duration |
| same_day_ed_inpatient_bundling_possible | medium | ED and inpatient utilization counts |
| ancillary_timestamp_granularity_mixed | low | procedure and imaging sequence |
Claims linkage
| Date | Setting | Claim | Encounter |
|---|---|---|---|
| 2024-02-09 | outpatient | claim_77aa | enc-4 |
| 2024-07-21 | emergency | claim_91fd | enc-6 |
| 2024-07-21/2024-07-24 | inpatient | claim_ip_332c | enc-7 |
| 2024-08-12 | outpatient infusion | claim_op_81bc | enc-11 |
The same record as FHIR
FHIR (Fast Healthcare Interoperability Resources) defines the standard resources health systems use to exchange records; they travel as JSON, XML, or Turtle. The bundle below carries selected resources for the same synthetic patient as the chart screen above, reusing the record ids from the transport file. Numbered markers point at specific lines; the notes under each resource explain the field and its research caveat. Download the bundle (pure FHIR, annotations stripped).
Patient syn-onc-0007
{
"resourceType": "Patient",
"id": "syn-onc-0007",
"extension": [
{
"url": "http://hl7.org/fhir/us/core/StructureDefinition/us-core-race", 1
"extension": [
{
"url": "ombCategory",
"valueCoding": {
"system": "urn:oid:2.16.840.1.113883.6.238",
"code": "2054-5",
"display": "Black or African American"
}
},
{
"url": "text",
"valueString": "Black or African American"
}
]
},
{
"url": "http://hl7.org/fhir/us/core/StructureDefinition/us-core-ethnicity",
"extension": [
{
"url": "ombCategory",
"valueCoding": {
"system": "urn:oid:2.16.840.1.113883.6.238",
"code": "2186-5",
"display": "Not Hispanic or Latino"
}
},
{
"url": "text",
"valueString": "Not Hispanic or Latino"
}
]
}
],
"identifier": [ 2
{
"system": "https://rwednesdays.com/fhir/naming/ehr-patient-token",
"value": "ehrtok_9b7a2d"
},
{
"system": "https://rwednesdays.com/fhir/naming/claims-patient-token",
"value": "clmtok_51fd90"
}
],
"gender": "female",
"address": [
{
"postalCode": "303", 3
"state": "US-South",
"country": "US"
}
],
"communication": [
{
"language": {
"coding": [
{
"system": "urn:ietf:bcp:47",
"code": "en",
"display": "English"
}
]
},
"preferred": true
}
]
}- 1Race and ethnicity are not core Patient fields. US systems carry them in US Core extensions with OMB category codes.RWEEquity descriptors and subgroup analyses depend on these extensions being populated; missingness is common and differential.
- 2Identifiers are pseudonymized tokens, one per source system. The system URI names the namespace that makes the value unique; the issuing organization would go in Identifier.assigner.RWETokenized identifiers are what link EHR and claims records for the same person; linkage error enters here.
- 3Only the 3-digit ZIP prefix is kept and birthDate is absent — two of the HIPAA Safe Harbor steps. This bundle is synthetic, not a Safe Harbor extract: full month and day survive on every date, and Safe Harbor allows a ZIP3 only when its population exceeds 20,000.RWEDe-identification limits geographic and age precision. Read 45 CFR 164.514(b)(2) for the full identifier list rather than assuming ZIP3 plus no birth date is sufficient.
- §There is no name, address line, or birth date anywhere in this resource. A real EHR Patient resource would have all three.RWEResearch extracts drop direct identifiers; the study index date replaces birth date arithmetic done inside the health system.
Condition cond-1
{
"resourceType": "Condition",
"id": "cond-1",
"clinicalStatus": {
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/condition-clinical",
"code": "active"
}
]
},
"verificationStatus": { 1
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/condition-ver-status",
"code": "confirmed"
}
]
},
"category": [
{
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/condition-category",
"code": "problem-list-item",
"display": "Problem List Item"
}
]
}
],
"code": {
"coding": [
{
"system": "http://hl7.org/fhir/sid/icd-10-cm", 2
"code": "C34.11",
"display": "Malignant neoplasm of upper lobe, right bronchus or lung"
}
],
"text": "Lung adenocarcinoma, right upper lobe"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"onsetDateTime": "2024-01-18", 3
"stage": [ 4
{
"summary": {
"text": "Stage IVB"
},
"type": {
"coding": [
{
"system": "http://snomed.info/sct",
"code": "260998006",
"display": "Clinical staging"
}
]
}
}
]
}- 1FHIR separates whether the problem is active (clinicalStatus) from how certain the diagnosis is (verificationStatus).RWEA single diagnosis code on one claim is weak evidence. Case definitions usually require confirmed status, repeat codes, or linked treatment.
- 2The billing vocabulary ICD-10-CM appears under code.coding.system as a canonical URI. The free-text label rides along in text.RWECohort entry usually keys on code.coding.code; the text field is what abstractors and NLP read.
- 3Onset is the clinical onset date, not the date the code was first recorded.RWEIncident-user and new-diagnosis designs need the first-recorded date; onset vs recorded-date confusion shifts index dates.
- 4Stage lives in a dedicated stage element, here only as text. Structured AJCC stage coding is rare outside oncology modules.RWEStage is a critical confounder and is often only in narrative or a tumor registry, not in the coded EHR problem list.
Encounter enc-6
{
"resourceType": "Encounter",
"id": "enc-6",
"status": "finished",
"class": { 1
"system": "http://terminology.hl7.org/CodeSystem/v3-ActCode",
"code": "EMER",
"display": "emergency"
},
"priority": {
"text": "ESI 2"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"period": { 2
"start": "2024-07-21T18:42:00-04:00",
"end": "2024-07-21T23:58:00-04:00"
},
"reasonCode": [ 3
{
"text": "shortness of breath and right-sided pleuritic chest pain"
}
],
"location": [
{
"location": {
"display": "Hospital ED"
}
}
]
}- 1Encounter.class carries the care setting as a v3-ActCode: EMER means emergency. Inpatient is IMP, office visit is AMB.RWEED-visit and hospitalization endpoints are built from class plus period; misclassifying class corrupts utilization outcomes.
- 2Timestamps carry timezone offsets. This ED stay starts 18:42 and ends 23:58 the same day.RWESame-day ED-to-inpatient transitions are why ED visits get bundled into inpatient claims; episode logic must merge them.
- 3The chief complaint is free text here, not a coded reason. That is typical for ED encounters.RWEPresenting-symptom analyses often need NLP; the coded diagnosis for this episode arrives later on the inpatient record.
Encounter enc-7
{
"resourceType": "Encounter",
"id": "enc-7",
"status": "finished",
"class": {
"system": "http://terminology.hl7.org/CodeSystem/v3-ActCode",
"code": "IMP",
"display": "inpatient encounter"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"period": {
"start": "2024-07-21",
"end": "2024-07-24"
},
"reasonCode": [
{
"coding": [
{
"system": "http://hl7.org/fhir/sid/icd-10-cm",
"code": "J91.0",
"display": "Malignant pleural effusion"
}
]
}
],
"hospitalization": { 1
"admitSource": {
"text": "emergency department"
},
"dischargeDisposition": {
"text": "home with home health"
}
},
"location": [
{
"location": {
"display": "7 East oncology floor"
}
}
]
}- 1The admission carries admitSource (came from the ED) and dischargeDisposition (home with home health) — the fields that turn a date range into an episode.RWEReadmission and post-acute-care outcomes are built from disposition; an ED-sourced admission also signals the ED visit should be merged into this episode, not counted separately.
Encounter enc-8
{
"resourceType": "Encounter",
"id": "enc-8",
"status": "finished",
"class": { 1
"system": "http://terminology.hl7.org/CodeSystem/v3-ActCode",
"code": "IMP",
"display": "inpatient encounter"
},
"serviceType": {
"text": "interventional radiology"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"period": {
"start": "2024-07-22",
"end": "2024-07-22"
},
"reasonCode": [
{
"text": "ultrasound-guided thoracentesis"
}
],
"location": [
{
"location": {
"display": "IR procedure suite"
}
}
],
"partOf": {
"reference": "Encounter/enc-7"
}
}- 1The IR procedure visit is a child encounter: class IMP with partOf pointing at the admission enc-7. The source system filed it as a separate "ancillary" visit, which is why the transport file and this bundle describe the same event differently.RWEEncounter granularity varies by source system. Roll child encounters up via partOf before counting visits, or one admission becomes three.
MedicationRequest medrx-1
{
"resourceType": "MedicationRequest",
"id": "medrx-1",
"extension": [ 1
{
"url": "https://rwednesdays.com/fhir/StructureDefinition/line-of-therapy",
"valueInteger": 1
}
],
"status": "stopped", 2
"statusReason": {
"text": "radiographic progression / non-response"
},
"intent": "order",
"medicationCodeableConcept": { 3
"coding": [
{
"system": "http://www.nlm.nih.gov/research/umls/rxnorm",
"code": "1721581",
"display": "osimertinib 80 MG Oral Tablet"
}
]
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"authoredOn": "2024-02-09",
"dosageInstruction": [
{
"text": "80 mg oral daily"
}
]
}- 1Line of therapy is NOT a FHIR element. This is a custom extension, and its url is where a real feed must publish a StructureDefinition defining the field. No definition is published at this illustrative URL.RWELine of therapy is a constructed variable. An undefined extension is an undocumented algorithm — ask for the definition and the code that produced it before using a vendor-supplied value.
- 2status stopped plus statusReason records that the drug was discontinued and why — but R4 MedicationRequest has no field for WHEN. authoredOn is the original prescription date. The 2024-08-05 stop date exists only in the chart and the transport file.RWEA dated discontinuation endpoint cannot come from these fields alone; it needs Provenance, an eventHistory reference, or the source system’s own stop-date column.
- 3The drug is coded in RxNorm, the US standard drug vocabulary; the code identifies drug, strength, and form.RWEExposure definitions map product-level codes up to ingredient level; a code list built at the wrong level miscounts exposure.
- §This is an order. It proves intent to treat, not that the drug was picked up or taken.RWEOrder-only exposure misclassifies never-starters as exposed; pair orders with dispensing records.
MedicationDispense disp-1
{
"resourceType": "MedicationDispense",
"id": "disp-1",
"status": "completed",
"medicationCodeableConcept": {
"coding": [
{
"system": "http://www.nlm.nih.gov/research/umls/rxnorm",
"code": "1721581",
"display": "osimertinib 80 MG Oral Tablet"
}
]
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"authorizingPrescription": [ 1
{
"reference": "MedicationRequest/medrx-1"
}
],
"quantity": {
"value": 30,
"unit": "tablet",
"system": "http://unitsofmeasure.org",
"code": "{tbl}"
},
"daysSupply": { 2
"value": 30,
"unit": "days",
"system": "http://unitsofmeasure.org",
"code": "d"
},
"whenHandedOver": "2024-02-09"
}- 1The dispense points back to the order it fills. References use ResourceType/id form.RWEThis link is how order-to-fill latency and primary non-adherence are measured.
- 2daysSupply and quantity describe what the pharmacy handed over on the whenHandedOver date. Coded UCUM units (system + code) make the quantity machine-comparable; a bare unit string does not.RWEAdherence measures need more than this one field: fill dates, multiple fills, a defined observation window, and an overlap rule for early refills.
- §Dispensing confirms availability, not ingestion. For oral oncolytics this is the standard exposure proxy and its known limit.RWEState the proxy explicitly in methods: exposure equals dispensed supply, not observed intake.
Observation lab-2
{
"resourceType": "Observation",
"id": "lab-2",
"status": "final",
"category": [
{
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/observation-category",
"code": "laboratory"
}
]
}
],
"code": {
"coding": [
{
"system": "http://loinc.org",
"code": "2039-6", 1
"display": "Carcinoembryonic Ag [Mass/volume] in Serum or Plasma"
}
],
"text": "CEA"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"effectiveDateTime": "2024-07-15",
"valueQuantity": { 2
"value": 21.4,
"unit": "ng/mL",
"system": "http://unitsofmeasure.org",
"code": "ng/mL"
},
"interpretation": [
{
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/v3-ObservationInterpretation",
"code": "H",
"display": "High"
}
]
}
],
"referenceRange": [ 3
{
"low": {
"value": 0.0,
"unit": "ng/mL"
},
"high": {
"value": 3.0,
"unit": "ng/mL"
}
}
]
}- 1Labs are coded in LOINC; 2039-6 is serum CEA. The same test can arrive under different local codes mapped to one LOINC.RWELab-based covariates and endpoints need LOINC code lists plus unit checks; local-code leakage causes silent missingness.
- 2The result is a quantity with an explicit UCUM unit. 21.4 ng/mL sits far above the reference range, so interpretation is H.RWEAlways harmonize units before modeling; a mixed mg/L and ng/mL series looks like biological signal but is a unit artifact.
- 3The lab ships its own reference range with each result; ranges differ across labs and over time.RWEDichotomizing on a single fixed cutoff ignores between-lab range differences; consider lab-specific normalization.
Observation ecog-1
{
"resourceType": "Observation",
"id": "ecog-1",
"status": "final",
"category": [ 1
{
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/observation-category",
"code": "survey"
}
]
}
],
"code": {
"coding": [
{
"system": "http://loinc.org",
"code": "89247-1",
"display": "ECOG Performance Status score"
}
]
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"effectiveDateTime": "2024-02-09",
"valueInteger": 1, 2
"interpretation": [
{
"coding": [
{
"system": "http://loinc.org",
"code": "LA9623-5",
"display": "Restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature"
}
]
}
]
}- 1Category survey distinguishes clinician-scored instruments from laboratory results.RWEExtraction pipelines that only pull category laboratory will silently drop ECOG, weight, and vitals.
- 2LOINC 89247-1 is a 0-5 score, so the value is an integer. The matching answer-list code (LA9623-5 = ECOG 1) belongs in interpretation, which is how the mCODE oncology profile models it.RWEPerformance status is a strong confounder in oncology and is missing for most patients in routine data; here it exists because the oncologist documented it at treatment start.
Observation egfr-1
{
"resourceType": "Observation",
"id": "egfr-1",
"status": "final",
"category": [
{
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/observation-category",
"code": "laboratory"
}
]
}
],
"code": {
"coding": [
{
"system": "http://loinc.org",
"code": "69548-6", 1
"display": "Genetic variant assessment"
}
],
"text": "EGFR exon 19 deletion"
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"effectiveDateTime": "2024-02-02",
"valueCodeableConcept": {
"coding": [
{
"system": "http://loinc.org",
"code": "LA9633-4",
"display": "Present"
}
]
},
"note": [ 2
{
"text": "Tumor NGS panel: EGFR exon 19 deletion detected. Source record anc-path-1."
}
]
}- 1Genomic results often arrive as a generic variant-assessment observation (LOINC 69548-6) with the actual variant in free text or notes.RWEBiomarker cohorts built only from structured fields undercount; the variant name frequently lives in the pathology narrative.
- 2The note preserves the source narrative and points at the originating pathology record (anc-path-1).RWEProvenance back to the source report is what lets an abstractor verify the biomarker call during validation.
Procedure proc-1
{
"resourceType": "Procedure",
"id": "proc-1",
"status": "completed",
"code": {
"coding": [
{
"system": "http://www.ama-assn.org/go/cpt",
"code": "32555", 1
"display": "Thoracentesis with imaging guidance"
}
]
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"performedDateTime": "2024-07-22",
"encounter": { 2
"reference": "Encounter/enc-8"
}
}- 1Procedures bill in CPT; 32555 is image-guided thoracentesis. Hospitals may also code the same event in ICD-10-PCS.RWEProcedure-based endpoints need code lists across CPT, HCPCS, and ICD-10-PCS; one vocabulary alone undercounts.
- 2The reference points at Encounter/enc-8, included here so the downloadable bundle resolves. Real extracts are rarely this tidy: references to resources outside the delivered set are common.RWECount unresolved references before trusting episode logic; referential integrity is a data-quality check, not a given.
Coverage cov-1
{
"resourceType": "Coverage",
"id": "cov-1",
"status": "active",
"type": {
"text": "commercial"
},
"subscriberId": "sub_44fd19", 1
"beneficiary": {
"reference": "Patient/syn-onc-0007"
},
"period": { 2
"start": "2023-12-01",
"end": "2024-12-31"
},
"payor": [
{
"display": "Synthetic Commercial Health Plan"
}
],
"relationship": {
"coding": [
{
"system": "http://terminology.hl7.org/CodeSystem/subscriber-relationship",
"code": "self"
}
]
}
}- 1The subscriber identifier is hashed and matches the transport file. relationship self says the patient is the subscriber; a dependent would differ.RWEFamily-level linkage and payer switches are tracked through these identifiers; a payer switch looks like disenrollment.
- 2The coverage period is the enrollment window: 2023-12-01 through 2024-12-31.RWEPerson-time denominators, baseline lookback, and censoring all come from enrollment periods, not visit dates. Medical-versus-pharmacy benefit detail is not in this resource — check the payer eligibility file.
- §Coverage is the FHIR view of eligibility data. In practice enrollment truth usually comes from the claims warehouse, not the EHR.RWEWhen EHR and claims enrollment disagree, prefer the payer eligibility file; document the precedence rule.
DocumentReference docref-1
{
"resourceType": "DocumentReference",
"id": "docref-1",
"status": "current",
"type": { 1
"coding": [
{
"system": "http://loinc.org",
"code": "11506-3",
"display": "Progress note"
}
]
},
"subject": {
"reference": "Patient/syn-onc-0007"
},
"date": "2024-08-05T17:58:00-04:00",
"author": [
{
"display": "Thoracic medical oncology attending (synthetic)"
}
],
"description": "Oncology progress note, progression visit (full text)",
"content": [
{
"attachment": {
"contentType": "text/plain",
"language": "en",
"data": "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", 2
"title": "Oncology progress note 2024-08-05"
}
}
]
}- 1The note type is LOINC-coded (11506-3, progress note). The code says what kind of document exists — nothing about what it contains.RWEDocument-type codes let you count documentation events and select note corpora; content requires the attachment.
- 2The note text travels base64-encoded inside content.attachment.data. Decode this string and you get the complete progression note shown above, byte for byte.RWEMany research extracts drop attachments at ingestion. When that happens, the stage narrative, the response assessment, and the clinician's reasoning vanish without any missing-data flag.
- §This is how unstructured notes travel in FHIR: one DocumentReference per document, with the text inline or behind a URL. Narrative can also arrive via Composition, DiagnosticReport, or a resource’s own text element.RWEBefore concluding a feed has no notes, check every document, attachment, and narrative interface — not just DocumentReference.
Chart, FHIR, and RWE mapping
| Chart element | FHIR element | Code system | RWE use and caveat |
|---|---|---|---|
| Patient banner token | Patient.identifier | site token systems | Person linkage across EHR and claims; linkage error enters here. |
| Banner race / ethnicity | Patient.extension (US Core) | OMB categories | Equity descriptors; missingness is common and differential. |
| Problem list entry | Condition.code.coding | ICD-10-CM | Cohort entry and case definitions. |
| Problem status | Condition.clinicalStatus + verificationStatus | FHIR terminology | Certainty tiers in case definitions; statuses go stale. |
| Stage (often in the note) | Condition.stage.summary | text / AJCC | Key confounder; frequently narrative-only. |
| Medication list entry | MedicationRequest.medicationCodeableConcept | RxNorm | Exposure definition at drug or ingredient level. |
| Med stopped + reason | MedicationRequest.status + statusReason | FHIR terminology | Discontinuation endpoints — but R4 carries no stop date; get it from Provenance or the source column. |
| Pharmacy fill | MedicationDispense.daysSupply + whenHandedOver | UCUM | Exposure windows and gaps; PDC/MPR also need multiple fills and a defined window. |
| Flowsheet lab value | Observation.valueQuantity + referenceRange | LOINC + UCUM | Covariates and lab-based endpoints; harmonize units. |
| Flowsheet ECOG | Observation.valueInteger (+ interpretation) | LOINC 89247-1 + answer list | Oncology confounder with heavy missingness. |
| Clinical note | DocumentReference.content.attachment | LOINC document type | NLP and abstraction; check every narrative interface before declaring notes absent. |
| ED visit header | Encounter.class + period | v3-ActCode | Utilization endpoints and episode construction. |
| Procedure entry | Procedure.code | CPT | Procedure endpoints; needs multi-vocabulary code lists. |
| Insurance chip | Coverage.period | payer eligibility | Person-time denominator, lookback, censoring. |
JSON transport
Raw JSON
{
"transport_meta": {
"schema": "rwednesdays.synthetic_oncology_ehr.v0.2",
"style": "OpenEMR/FHIR-R4-like longitudinal patient bundle with human-readable service records",
"synthetic": true,
"date_shifted": true,
"contains_phi": false,
"generated_at": "2026-07-06T12:00:00-04:00",
"license_note": "Synthetic example for methods education only; not medical advice and not patient data."
},
"patient": {
"synthetic_patient_id": "SYN-ONC-0007",
"pseudonymized_ids": {
"ehr_patient_token": "ehrtok_9b7a2d",
"claims_patient_token": "clmtok_51fd90",
"linkage_confidence": 0.94
},
"demographics": {
"age_at_index": 66,
"sex": "female",
"race": "Black or African American",
"ethnicity": "Not Hispanic or Latino",
"language": "English",
"region": "US-South",
"synthetic_zip3": "303"
},
"_rwe": "Cohort eligibility, baseline covariates, subgroup checks, equity descriptors, and linkage denominator."
},
"diagnoses": [
{
"id": "cond-1",
"resource_type": "Condition",
"code": {
"system": "ICD-10-CM",
"code": "C34.11",
"display": "Malignant neoplasm of upper lobe, right bronchus or lung"
},
"clinical_status": "active",
"verification_status": "confirmed",
"onset_date": "2024-01-18",
"stage_at_diagnosis": "Stage IVB",
"biomarkers": [
{
"name": "EGFR exon 19 deletion",
"result": "positive",
"date": "2024-02-02",
"source_record": "anc-path-1"
}
],
"_rwe": "Incident cancer diagnosis, disease site, stage, and molecular eligibility subgroup."
},
{
"id": "cond-2",
"resource_type": "Condition",
"code": {
"system": "ICD-10-CM",
"code": "J91.0",
"display": "Malignant pleural effusion"
},
"clinical_status": "active",
"verification_status": "confirmed",
"onset_date": "2024-07-21",
"_rwe": "Acute progression-related complication supporting treatment-failure classification."
},
{
"id": "cond-3",
"resource_type": "Condition",
"code": {
"system": "ICD-10-CM",
"code": "R06.02",
"display": "Shortness of breath"
},
"clinical_status": "resolved",
"verification_status": "confirmed",
"onset_date": "2024-07-21",
"_rwe": "ED presenting symptom; useful for symptom-to-diagnosis pathway analyses but not a confirmed disease endpoint."
}
],
"encounters": [
{
"id": "enc-1",
"resource_type": "Encounter",
"class": "outpatient",
"date": "2024-01-18",
"provider_type": "thoracic oncology",
"location": "Cancer Center North Clinic",
"reason": "new lung mass workup",
"source_record": "out-onc-1",
"_rwe": "Specialty outpatient workup and first confirmed diagnosis context."
},
{
"id": "enc-2",
"resource_type": "Encounter",
"class": "ancillary",
"date": "2024-01-24",
"provider_type": "radiology",
"location": "Hospital imaging suite",
"reason": "baseline staging CT chest/abdomen/pelvis",
"source_record": "anc-rad-1",
"_rwe": "Baseline disease-burden measurement."
},
{
"id": "enc-3",
"resource_type": "Encounter",
"class": "ancillary",
"date": "2024-02-02",
"provider_type": "molecular pathology",
"location": "Reference laboratory interface",
"reason": "tumor NGS panel resulted",
"source_record": "anc-path-1",
"_rwe": "Biomarker eligibility and subgroup definition."
},
{
"id": "enc-4",
"resource_type": "Encounter",
"class": "outpatient",
"date": "2024-02-09",
"provider_type": "medical oncology",
"location": "Cancer Center North Clinic",
"reason": "systemic therapy start",
"source_record": "out-onc-2",
"_rwe": "Index date anchor for first-line systemic therapy."
},
{
"id": "enc-5",
"resource_type": "Encounter",
"class": "ancillary",
"date": "2024-07-15",
"provider_type": "laboratory",
"location": "Hospital outpatient lab",
"reason": "pre-visit labs and tumor marker",
"source_record": "anc-lab-2",
"_rwe": "Tumor-marker trend and treatment-monitoring evidence."
},
{
"id": "enc-6",
"resource_type": "Encounter",
"class": "emergency",
"date": "2024-07-21",
"admit_time": "18:42",
"discharge_time": "23:58",
"provider_type": "emergency medicine",
"location": "Hospital ED",
"triage_acuity": "ESI 2",
"chief_complaint": "shortness of breath and right-sided pleuritic chest pain",
"source_record": "ed-visit-1",
"_rwe": "Acute utilization and possible progression signal; ED symptom record precedes inpatient principal diagnosis."
},
{
"id": "enc-7",
"resource_type": "Encounter",
"class": "inpatient",
"admit_date": "2024-07-21",
"discharge_date": "2024-07-24",
"provider_type": "hospital medicine with oncology consult",
"location": "7 East oncology floor",
"admit_source": "emergency department",
"discharge_disposition": "home with home health",
"principal_diagnosis": "malignant pleural effusion",
"source_record": "ip-stay-1",
"_rwe": "Hospitalization episode, length of stay, severity, and downstream outcome/resource-use signal."
},
{
"id": "enc-8",
"resource_type": "Encounter",
"class": "ancillary",
"date": "2024-07-22",
"provider_type": "interventional radiology",
"location": "IR procedure suite",
"reason": "ultrasound-guided thoracentesis",
"source_record": "anc-proc-1",
"_rwe": "Procedure confirmation and evidence that the effusion was clinically significant."
},
{
"id": "enc-9",
"resource_type": "Encounter",
"class": "ancillary",
"date": "2024-08-01",
"provider_type": "radiology",
"location": "Hospital imaging suite",
"reason": "restaging CT chest/abdomen/pelvis",
"source_record": "anc-rad-2",
"_rwe": "Radiographic progression signal."
},
{
"id": "enc-10",
"resource_type": "Encounter",
"class": "outpatient",
"date": "2024-08-05",
"provider_type": "medical oncology",
"location": "Cancer Center North Clinic",
"reason": "progression visit and treatment change",
"source_record": "out-onc-3",
"_rwe": "Candidate treatment-failure date and clinical decision to stop first-line therapy."
},
{
"id": "enc-11",
"resource_type": "Encounter",
"class": "outpatient",
"date": "2024-08-12",
"provider_type": "infusion oncology",
"location": "Cancer Center infusion suite",
"reason": "cycle 1 day 1 chemoimmunotherapy",
"source_record": "out-inf-1",
"_rwe": "Next-line therapy start confirms line switch."
}
],
"service_records": [
{
"id": "out-onc-1",
"setting": "outpatient",
"date": "2024-01-18",
"record_type": "oncology consultation note",
"human_readable": "Thoracic oncology consult for persistent cough and CT-detected right upper-lobe mass. Biopsy from outside hospital reviewed as lung adenocarcinoma. Staging workup ordered; baseline ECOG not yet documented.",
"coded_elements": [
{
"system": "ICD-10-CM",
"code": "C34.11",
"display": "Malignant neoplasm of upper lobe, right bronchus or lung"
},
{
"system": "CPT",
"code": "99205",
"display": "New patient office visit, high complexity"
}
],
"linked_resources": [
"enc-1",
"cond-1"
],
"rwe_use": "Incident disease confirmation, specialty-care cohort entry, baseline date context.",
"data_quality_notes": "Narrative stage terms require abstraction; outside-pathology date may differ from local EHR diagnosis date."
},
{
"id": "anc-rad-1",
"setting": "ancillary",
"date": "2024-01-24",
"record_type": "radiology report",
"human_readable": "CT chest/abdomen/pelvis with contrast: 4.2 cm right upper-lobe mass, multiple bilateral pulmonary nodules, small right pleural effusion, and two liver lesions suspicious for metastases. Impression: metastatic lung cancer, baseline disease burden.",
"coded_elements": [
{
"system": "LOINC",
"code": "24627-2",
"display": "CT Chest Abdomen Pelvis WO and W contrast"
},
{
"system": "CPT",
"code": "71260",
"display": "CT thorax with contrast"
}
],
"linked_resources": [
"enc-2",
"img-0"
],
"rwe_use": "Baseline tumor burden and pre-treatment imaging anchor.",
"data_quality_notes": "RECIST target-lesion measurements are not structured; only radiology narrative is available."
},
{
"id": "anc-path-1",
"setting": "ancillary",
"date": "2024-02-02",
"record_type": "molecular pathology result",
"human_readable": "Tumor NGS addendum: EGFR exon 19 deletion detected; ALK and ROS1 rearrangements not detected; PD-L1 TPS 5%. Report electronically signed by molecular pathology.",
"coded_elements": [
{
"system": "LOINC",
"code": "81247-9",
"display": "Master HL7 genetic variant reporting panel"
},
{
"system": "HGVS-like",
"code": "EGFR exon 19 deletion",
"display": "Actionable EGFR alteration"
}
],
"linked_resources": [
"enc-3",
"cond-1"
],
"rwe_use": "Biomarker-defined cohort eligibility and treatment selection rationale.",
"data_quality_notes": "Exact variant nomenclature is simplified for the synthetic example; real pipelines should retain HGVS and assay metadata."
},
{
"id": "out-onc-2",
"setting": "outpatient",
"date": "2024-02-09",
"record_type": "oncology treatment-start note",
"human_readable": "Patient started osimertinib 80 mg by mouth daily for EGFR-mutant metastatic NSCLC. ECOG performance status 1. Baseline creatinine 0.9 mg/dL. Consent documented; toxicity counseling completed.",
"coded_elements": [
{
"system": "RxNorm",
"code": "1721581",
"display": "osimertinib 80 MG Oral Tablet"
},
{
"system": "LOINC",
"code": "89247-1",
"display": "ECOG performance status"
}
],
"linked_resources": [
"enc-4",
"medrx-1",
"disp-1",
"ecog-1"
],
"rwe_use": "Index date, first-line exposure start, baseline functional status.",
"data_quality_notes": "Medication order and specialty-pharmacy dispense agree on start month but not actual ingestion."
},
{
"id": "anc-lab-2",
"setting": "ancillary",
"date": "2024-07-15",
"record_type": "outpatient laboratory panel",
"human_readable": "Pre-visit labs: creatinine 1.1 mg/dL, AST 42 U/L, hemoglobin 10.8 g/dL, CEA 21.4 ng/mL, up from 8.2 ng/mL in May. Results interfaced from hospital laboratory.",
"coded_elements": [
{
"system": "LOINC",
"code": "2160-0",
"display": "Creatinine [Mass/volume] in Serum or Plasma"
},
{
"system": "LOINC",
"code": "2039-6",
"display": "Carcinoembryonic Ag [Mass/volume] in Serum or Plasma"
}
],
"linked_resources": [
"enc-5",
"lab-2"
],
"rwe_use": "Tumor-marker progression support and safety-monitoring covariates.",
"data_quality_notes": "Reference ranges and assay platform are not harmonized across sites unless mapped separately."
},
{
"id": "ed-visit-1",
"setting": "emergency_department",
"date": "2024-07-21",
"record_type": "ED triage, provider note, and facility charge",
"human_readable": "Arrived to ED at 18:42 with 2 days of worsening dyspnea and right-sided pleuritic pain. Triage: ESI 2, SpO2 89% on room air, HR 118, RR 26. Chest x-ray showed enlarging right pleural effusion. Treated with oxygen and IV morphine; admitted to oncology floor.",
"coded_elements": [
{
"system": "ICD-10-CM",
"code": "R06.02",
"display": "Shortness of breath"
},
{
"system": "UB-04 revenue center",
"code": "0450",
"display": "Emergency room"
},
{
"system": "CPT",
"code": "99285",
"display": "Emergency department visit, high severity"
}
],
"linked_resources": [
"enc-6",
"cond-3",
"claim_91fd"
],
"rwe_use": "Acute-care utilization, symptom onset, and progression-related escalation of care.",
"data_quality_notes": "ED chief complaint is symptom-level and should not be treated as a confirmed endpoint without inpatient/radiology follow-up."
},
{
"id": "ip-stay-1",
"setting": "inpatient",
"date": "2024-07-21/2024-07-24",
"record_type": "inpatient admission, daily notes, and discharge summary",
"human_readable": "Admitted from ED for hypoxemia due to malignant right pleural effusion. Hospital course: oxygen, ultrasound-guided thoracentesis removing 1.2 L serosanguinous fluid, oncology consult concerned for progression on osimertinib. Discharged home with home-health nursing and oncology follow-up.",
"coded_elements": [
{
"system": "ICD-10-CM",
"code": "J91.0",
"display": "Malignant pleural effusion"
},
{
"system": "MS-DRG",
"code": "181",
"display": "Respiratory neoplasms with MCC"
},
{
"system": "UB-04 type of bill",
"code": "0111",
"display": "Hospital inpatient admit-through-discharge claim"
}
],
"linked_resources": [
"enc-7",
"cond-2",
"anc-proc-1",
"claim_ip_332c"
],
"rwe_use": "Inpatient outcome, length of stay, severity, procedure utilization, and post-index resource use.",
"data_quality_notes": "Admission date follows ED date; same-day ED facility charges may be bundled into the inpatient claim for some payers."
},
{
"id": "anc-proc-1",
"setting": "ancillary",
"date": "2024-07-22",
"record_type": "interventional radiology procedure note",
"human_readable": "Ultrasound-guided right thoracentesis performed; 1.2 L serosanguinous pleural fluid removed. Cytology sent. No immediate complication.",
"coded_elements": [
{
"system": "CPT",
"code": "32555",
"display": "Thoracentesis with imaging guidance"
},
{
"system": "SNOMED CT",
"code": "39607008",
"display": "Thoracentesis"
}
],
"linked_resources": [
"enc-8",
"proc-1",
"cond-2"
],
"rwe_use": "Procedure confirmation, complication severity, and evidence that effusion was clinically actionable.",
"data_quality_notes": "Procedure may appear in inpatient procedure tables, professional claims, and radiology systems with different timestamps."
},
{
"id": "anc-rad-2",
"setting": "ancillary",
"date": "2024-08-01",
"record_type": "restaging radiology report",
"human_readable": "CT chest/abdomen/pelvis: increased size and number of bilateral pulmonary nodules, new liver lesion, and recurrent moderate right pleural effusion. Impression: progressive metastatic disease compared with 2024-01-24 baseline.",
"coded_elements": [
{
"system": "LOINC",
"code": "24627-2",
"display": "CT Chest Abdomen Pelvis WO and W contrast"
},
{
"system": "RECIST-like abstraction",
"code": "PD",
"display": "Progressive disease"
}
],
"linked_resources": [
"enc-9",
"img-1"
],
"rwe_use": "Primary radiographic support for treatment-failure/non-response endpoint.",
"data_quality_notes": "RECIST-like category is abstracted from narrative; raw lesion measurements are incomplete."
},
{
"id": "out-onc-3",
"setting": "outpatient",
"date": "2024-08-05",
"record_type": "oncology progression visit",
"human_readable": "Reviewed restaging CT and recent hospitalization. Documented progressive disease on osimertinib with ECOG 2. Osimertinib discontinued. Plan: carboplatin, pemetrexed, and pembrolizumab after lab clearance.",
"coded_elements": [
{
"system": "RxNorm",
"code": "1721581",
"display": "osimertinib 80 MG Oral Tablet"
},
{
"system": "LOINC",
"code": "89247-1",
"display": "ECOG performance status"
}
],
"linked_resources": [
"enc-10",
"medrx-1",
"ecog-2",
"img-1"
],
"rwe_use": "Candidate failure date, discontinuation reason, and clinical non-response classification.",
"data_quality_notes": "If stop reason is not structured, NLP or chart abstraction is required to distinguish progression from toxicity."
},
{
"id": "out-inf-1",
"setting": "outpatient",
"date": "2024-08-12",
"record_type": "infusion administration record",
"human_readable": "Cycle 1 day 1 administered: pembrolizumab 200 mg IV, pemetrexed 500 mg/m2 IV, carboplatin AUC 5 IV. Premedications given; infusion completed without reaction.",
"coded_elements": [
{
"system": "HCPCS",
"code": "J9271",
"display": "Injection, pembrolizumab, 1 mg"
},
{
"system": "HCPCS",
"code": "J9305",
"display": "Injection, pemetrexed, 10 mg"
},
{
"system": "HCPCS",
"code": "J9045",
"display": "Injection, carboplatin, 50 mg"
}
],
"linked_resources": [
"enc-11",
"medrx-2",
"admin-1",
"claim_op_81bc"
],
"rwe_use": "Observed next-line therapy administration; stronger than order-only evidence for infused therapies.",
"data_quality_notes": "Dose units differ across HCPCS billing, MAR/eMAR, and pharmacy compounding systems."
}
],
"medications": [
{
"id": "medrx-1",
"resource_type": "MedicationRequest",
"drug": "osimertinib",
"rxnorm": "1721581",
"intent": "order",
"status": "stopped",
"start_date": "2024-02-09",
"stop_date": "2024-08-05",
"dose": "80 mg oral daily",
"status_reason": "radiographic progression / non-response",
"line_of_therapy": 1,
"_rwe": "Exposure, line-of-therapy start/stop, and treatment discontinuation signal."
},
{
"id": "disp-1",
"resource_type": "MedicationDispense",
"drug": "osimertinib",
"rxnorm": "1721581",
"dispense_date": "2024-02-09",
"days_supply": 30,
"quantity": 30,
"pharmacy_type": "specialty pharmacy",
"_rwe": "Dispensing confirms drug availability but not ingestion for oral oncolytic exposure."
},
{
"id": "admin-ed-1",
"resource_type": "MedicationAdministration",
"drug": "morphine",
"route": "IV",
"date_time": "2024-07-21T19:10:00-04:00",
"dose": "2 mg",
"encounter": "enc-6",
"_rwe": "ED treatment intensity marker; not an anticancer exposure."
},
{
"id": "medrx-2",
"resource_type": "MedicationRequest",
"drug": "carboplatin + pemetrexed + pembrolizumab",
"status": "active",
"start_date": "2024-08-12",
"line_of_therapy": 2,
"_rwe": "Post-failure regimen switch and next-line therapy."
},
{
"id": "admin-1",
"resource_type": "MedicationAdministration",
"drug": "pembrolizumab + pemetrexed + carboplatin",
"status": "completed",
"date_time": "2024-08-12T10:32:00-04:00",
"encounter": "enc-11",
"line_of_therapy": 2,
"_rwe": "Observed infused therapy administration; confirms start of second-line regimen."
}
],
"procedures": [
{
"id": "proc-1",
"resource_type": "Procedure",
"code": {
"system": "CPT",
"code": "32555",
"display": "Thoracentesis with imaging guidance"
},
"date": "2024-07-22",
"encounter": "enc-8",
"status": "completed",
"_rwe": "Ancillary procedure used to characterize severity and progression-related utilization."
}
],
"observations": [
{
"id": "lab-1",
"resource_type": "Observation",
"category": "laboratory",
"loinc": "2160-0",
"name": "Creatinine",
"date": "2024-02-05",
"value": 0.9,
"unit": "mg/dL",
"reference_range": "0.6-1.2",
"_rwe": "Baseline renal covariate."
},
{
"id": "lab-2",
"resource_type": "Observation",
"category": "laboratory",
"loinc": "2039-6",
"name": "CEA",
"date": "2024-07-15",
"value": 21.4,
"unit": "ng/mL",
"reference_range": "0.0-3.0",
"trend": "rising from 8.2",
"_rwe": "Tumor-marker progression support."
},
{
"id": "vital-ed-1",
"resource_type": "Observation",
"category": "vital-signs",
"loinc": "59408-5",
"name": "Oxygen saturation",
"date": "2024-07-21",
"time": "18:48",
"value": 89,
"unit": "%",
"_rwe": "ED severity marker and acute-care eligibility covariate."
},
{
"id": "ecog-1",
"resource_type": "Observation",
"category": "survey",
"loinc": "89247-1",
"name": "ECOG Performance Status",
"date": "2024-02-09",
"valueInteger": 1,
"_rwe": "Baseline functional-status covariate."
},
{
"id": "ecog-2",
"resource_type": "Observation",
"category": "survey",
"loinc": "89247-1",
"name": "ECOG Performance Status",
"date": "2024-08-05",
"valueInteger": 2,
"_rwe": "Clinical deterioration and endpoint-support signal."
},
{
"id": "img-0",
"resource_type": "Observation",
"category": "imaging-summary",
"date": "2024-01-24",
"valueCode": "baseline metastatic disease",
"source": "radiology report abstraction",
"_rwe": "Baseline imaging comparator."
},
{
"id": "img-1",
"resource_type": "Observation",
"category": "imaging-summary",
"date": "2024-08-01",
"valueCode": "progressive disease",
"source": "radiology report abstraction",
"_rwe": "Clinical response component; not RECIST-confirmed unless source report is available."
}
],
"treatment_failure_non_response": {
"constructed_endpoint": true,
"candidate_failure_date": "2024-08-05",
"classification": "probable treatment failure / non-response",
"certainty": "moderate",
"signals": [
{
"source_id": "medrx-1",
"signal": "MedicationRequest stopped for progression"
},
{
"source_id": "medrx-2",
"signal": "New line of therapy started within 7 days"
},
{
"source_id": "img-1",
"signal": "Restaging CT impression says progressive metastatic disease"
},
{
"source_id": "ecog-2",
"signal": "ECOG worsened from 1 to 2"
},
{
"source_id": "enc-6",
"signal": "Emergency encounter for dyspnea and pleural effusion"
},
{
"source_id": "enc-7",
"signal": "Inpatient admission for malignant pleural effusion"
},
{
"source_id": "admin-1",
"signal": "Observed second-line infusion administration"
}
],
"_rwe": "Composite endpoint combining clinical response evidence, treatment-pattern proxy, acute utilization, and intercurrent-event handling."
},
"claims_linkage": {
"payer_type": "commercial",
"subscriber_id_hash": "sub_44fd19",
"member_id_hash": "mem_772ac1",
"enrollment": [
{
"start": "2023-12-01",
"end": "2024-12-31",
"medical": true,
"pharmacy": true
}
],
"claim_refs": [
{
"claim_id_hash": "claim_77aa",
"date": "2024-02-09",
"setting": "outpatient",
"hcpcs": "99215",
"linked_encounter": "enc-4"
},
{
"claim_id_hash": "claim_91fd",
"date": "2024-07-21",
"setting": "emergency",
"revenue_center": "0450",
"hcpcs": "99285",
"linked_encounter": "enc-6"
},
{
"claim_id_hash": "claim_ip_332c",
"date": "2024-07-21/2024-07-24",
"setting": "inpatient",
"type_of_bill": "0111",
"ms_drg": "181",
"linked_encounter": "enc-7"
},
{
"claim_id_hash": "claim_op_81bc",
"date": "2024-08-12",
"setting": "outpatient infusion",
"revenue_center": "0636",
"hcpcs": "J9271",
"linked_encounter": "enc-11"
}
],
"_rwe": "Coverage continuity, utilization capture, EHR-claims concordance, and subscriber/member identifier separation."
},
"provenance": {
"source_systems": [
"synthetic OpenEMR-like EHR",
"synthetic radiology information system",
"synthetic laboratory information system",
"synthetic eMAR/infusion system",
"synthetic adjudicated claims"
],
"mapping_version": "oncology-ehr-rwe-map-0.2",
"source_trace": {
"line_of_therapy": "MedicationRequest sequence, specialty-pharmacy dispense, and infusion MedicationAdministration",
"treatment_failure_non_response": "Medication status reason, restaging imaging, ECOG change, ED/inpatient utilization, next-line therapy",
"claims": "Payer extract line-level claims linked to encounter IDs",
"ecog": "Structured observation using LOINC 89247-1",
"ed_episode": "ED triage note, vitals, facility revenue center, and admission handoff",
"inpatient_episode": "Admission/discharge timestamps, principal diagnosis, DRG-like grouping, and procedure record",
"ancillary_services": "Radiology, laboratory, pathology, and procedure systems linked back to encounters"
}
},
"data_quality_flags": [
{
"flag": "synthetic_record",
"severity": "info",
"affects": "all fields"
},
{
"flag": "date_shifted",
"severity": "info",
"affects": "calendar-time interpretation"
},
{
"flag": "imaging_not_recist_confirmed",
"severity": "medium",
"affects": "treatment_failure_non_response"
},
{
"flag": "claims_ehr_linkage_probabilistic",
"severity": "medium",
"confidence": 0.94,
"affects": "claims_linkage"
},
{
"flag": "oral_adherence_unobserved",
"severity": "medium",
"affects": "osimertinib exposure duration"
},
{
"flag": "same_day_ed_inpatient_bundling_possible",
"severity": "medium",
"affects": "ED and inpatient utilization counts"
},
{
"flag": "ancillary_timestamp_granularity_mixed",
"severity": "low",
"affects": "procedure and imaging sequence"
}
],
"source_references": [
{
"label": "OpenEMR FHIR API documentation",
"url": "https://github.com/openemr/openemr/blob/master/Documentation/api/FHIR_API.md"
},
{
"label": "HL7 FHIR R4 Encounter",
"url": "https://hl7.org/fhir/R4/encounter.html"
},
{
"label": "HL7 FHIR R4 Condition",
"url": "https://hl7.org/fhir/R4/condition.html"
},
{
"label": "HL7 FHIR R4 MedicationRequest",
"url": "https://hl7.org/fhir/R4/medicationrequest.html"
},
{
"label": "HL7 FHIR R4 MedicationAdministration",
"url": "https://hl7.org/fhir/R4/medicationadministration.html"
},
{
"label": "HL7 FHIR R4 Observation",
"url": "https://hl7.org/fhir/R4/observation.html"
},
{
"label": "HL7 FHIR R4 Procedure",
"url": "https://hl7.org/fhir/R4/procedure.html"
},
{
"label": "mCODE ECOG Performance Status",
"url": "https://build.fhir.org/ig/HL7/fhir-mCODE-ig/StructureDefinition-mcode-ecog-performance-status.html"
},
{
"label": "FDA Real-World Data guidance",
"url": "https://www.fda.gov/media/152503/download"
}
]
}