Devices · Diagnostics · Post-market evidence
Every catalogue code, in every EHR, traced through real outcomes.
The device and diagnostic world speaks in SKUs, UDIs and lot numbers. Hospital data speaks in procedure codes, free-text and local charge masters. Unison maps the two — and turns every implant, consumable and assay into a queryable real-world evidence stream.
CATALOGUE MAPPING
UDI / SKU → EHR procedure reality
POST-MARKET SURVEILLANCE
PMCF & PSUR evidence from live data
DIAGNOSTIC UTILITY
Assay → downstream clinical impact
REG. SUBMISSIONS
FDA & MDR/IVDR-ready evidence packs
01 · The device-data problem
Devices don't live in OMOP. Until Unison maps them in.
WHY IT'S HARD
- Your catalogue has 10,000+ SKUs. The hospital has CPT / ICD / OPCS codes and free-text.
- UDIs are captured inconsistently. Lot numbers live in supply-chain systems, not the EHR.
- The outcome you care about — revision, explant, assay follow-up — sits one or two hops downstream.
- Each hospital network requires a new SOW and a new harmonisation project.
WHAT UNISON DOES
- Maps your catalogue against the destination EHR's procedure codes, device tables and free-text.
- Resolves UDI fragments and lot-level traces where the data supports it.
- Links device exposure to downstream clinical events — revision, infection, re-intervention, biomarker follow-up.
- Same protocol re-runs across every connected network. New hospital ≠ new project.
02 · The mapping, made legible
From your SKU to the patient's outcome.
MANUFACTURER
Catalogue
SKU · 47201-A
UDI · (01)0085410...
Family · drug-eluting stent
→
HOSPITAL SOURCE
EHR reality
CPT 92928
device.code = "DES"
implant log · local vocab
→
OMOP VIRTUAL
Concept
Device 4034893
Procedure 4182962
harmonised · attributed
→
OUTCOME
Downstream event
Target-lesion revasc.
Infection · 90d
Mortality · 1y
One mapping. Every connected hospital. Every queryable outcome.
03 · What device & diagnostic teams run
Five workflows, one mapping layer.
POST-MARKET SURVEILLANCE
Continuous safety & performance signal
Track revision, explant, infection and re-intervention rates by device family, lot and site. Detect drift before regulators do.
OUTPUT
PMCF-ready signal dashboard · lot-level tracesREAL-WORLD PERFORMANCE
Head-to-head vs comparator in the wild
Pre-specified, matched comparisons of your device vs the comparator actually used by the hospital. Outcomes the surgeon cares about, not just those in the trial.
OUTPUT
PS-matched comparison · KM curves · subgroup tablesDIAGNOSTIC UTILITY
From assay result to clinical decision
Follow the patient from the diagnostic event through to the clinical action it triggered — treatment initiated, escalation avoided, outcome achieved.
OUTPUT
Decision-impact pathway · utility metricsHEALTH-ECONOMIC IMPACT
Cost, LOS, readmission — by device
Same device, different hospitals. Quantify total episode cost, length of stay, readmission and re-intervention across real care pathways.
OUTPUT
Episode cost table · LOS distribution · readmissionREGULATORY SUBMISSIONS
MDR / IVDR / FDA evidence packs
PMCF, PSUR and post-approval study evidence assembled from a living query — not a one-off retrospective. Same methods, every filing.
OUTPUT
Submission-ready tables · replayable UQL artefacts04 · From catalogue to evidence
Map the catalogue once. Every hospital becomes a query target.
01
Ingest catalogue
Your SKU hierarchy, UDIs, device families and known comparators.
02
Map to EHR
Resolve every code to the destination hospital's procedure codes, device tables and free-text notes.
03
Link outcomes
Connect device exposure to downstream events — revision, infection, assay follow-up, readmission.
04
Federate
Every connected hospital becomes a query target. Aggregate-only, custodian-controlled.
05
Submit
Assemble PMCF, PSUR, real-world performance and utility packs from the same living query.
A DAY IN POST-MARKET
"How is the v4 implant performing vs v3 — across the three networks that use both?"
A clinical evidence lead pulls 24-month revision rates, matched on patient and procedural characteristics, across three federated hospital networks. Same query. Every network. Aggregate-only.
Catalogue-aware
v3 and v4 SKUs mapped to EHR procedures
Matched
age, sex, comorbidity, procedural complexity
Aggregate-only
no patient-level movement
Replayable
UQL artefact per network · per cohort
# unison · post-market workspace
> "24m revision · v4 vs v3 · PS-matched · 3 networks."
→ Catalogue: device-family-A mapped · 94% coverage
→ Federated across 3 hospital networks · aggregate-only
Network A · n(v4)=1,214 n(v3)=1,214 matched
24m revision v4 3.1% v3 4.8% HR 0.64 (0.48–0.86)
Network B · n(v4)= 906 n(v3)= 906 matched
24m revision v4 2.8% v3 4.1% HR 0.68 (0.49–0.95)
Network C · n(v4)= 518 n(v3)= 518 matched
24m revision v4 3.4% v3 4.2% HR 0.81 (0.53–1.23)
# uql://query/pmcf-v4-a71b · replayable
05 · Built for device regulators
Aligned to how device and IVD regulators expect evidence to arrive.
Catalogue-aware from day one
Your SKU hierarchy is a first-class concept, not an afterthought. Mappings are versioned and attributable.
Continuous, not retrospective
PMCF and PSUR become a living query — refreshed as the data moves, not scoped as a one-off every cycle.
Defensible in submission
Every number backed by a replayable UQL artefact. Notified body or regulator re-runs, not just reads.
Standards & fit:· OMOP CDM-native· Cyber Essentials Plus· CFR 21 Part 11-ready· EU MDR / IVDR-aligned· FDA PMA/510(k) RWE-ready· UDI-aware· EHDS-aligned
06 · What changes for the function
From "RWE is a retrospective project" to "RWE is a standing capability."
NEW HOSPITAL ONBOARDING
new SOW → new connector
The catalogue mapping is done once. A new network is a new query target, not a new project.
SURVEILLANCE CADENCE
annual cycle → living query
PMCF and PSUR stop being deadlines. They become views on a continuously refreshed dataset.
SUBMISSION DEFENSIBILITY
narrative → replayable artefact
Every number a notified body or FDA reviewer sees is backed by a UQL artefact they can re-run.
Map your catalogue in 6 weeks