Drop in the protocol PDF (or a text file). It is parsed on your own server – only protocol text is ever sent to a model, never patient data.
Each criterion keeps a link to the page it came from. A clinician reviews the list – and can edit, merge or add – before anything is built on it.
The Atomizer splits each criterion into single concepts and keeps the logic that joins them – each piece is defined and tested on its own.
Diagnosis any time before time zero
Dialysis procedure or status code in the 90 days before time zero
Transplant procedure or transplant status, any time before
The agent searches each terminology with tools, adds ICD-9 via the CMS crosswalk and links every code to your EHR’s local names.
N18.6End-stage renal diseaseZ99.2Dependence on renal dialysisZ94.0Kidney transplant statusZ49.*Encounter for dialysis care585.6End stage renal diseaseV45.11Renal dialysis statusV42.0Kidney replaced by transplantV56.*Encounter for dialysis98979-8eGFR, CKD-EPI 202162238-1eGFR, CKD-EPI4548-4Hemoglobin A1c17856-6Hemoglobin A1c, HPLC1991302semaglutide475968liraglutide1551291dulaglutide60548exenatideDelimited text, CSV or FHIR: CIPHER profiles the files, maps the columns to a standard schema and builds a read-only copy.
cipher_patientsdemographics, death, record span0cipher_encountersvisits, admissions, settings0cipher_diagnosesICD-10 and ICD-9, primary flag0cipher_labsresults, units, reference ranges0cipher_medicationsorders, ingredients, classes0cipher_notestext, with facts extracted locally0The agent writes a structured definition; the platform compiles read-only SQL, runs it on the EHR copy and reports met, unknown and not met.
{
"domain": "lab",
"lab": "egfr",
"use": "latest",
"value": { "op": "<", "value": 30 },
"window": { "from": { "n": -12, "unit": "months" } }
}Combine definitions with AND / OR / NOT, switch criteria off, move thresholds – counts and demographics update instantly.
Eligible and possibly-eligible patients, screening worksheets with the evidence, refresh alerts and outcome rates for sample-size planning.
Time zero at treatment start, eligibility at time zero, outcomes followed over time, propensity-weighted comparisons and balance.
The Eligibility Extractor copies the criteria verbatim; the Study Design agent finds the arms, time zero, outcomes and follow-up.
Agents pick from a phenotype library and standard labs; the platform compiles the SQL, adds ICD-9 equivalents and converts units.
Synthetic boundary patients, a blind second builder, an independent reviewer – and the judgement calls laid out with their patient counts.
“No record” is not treated as “does not have it” unless the record covers the window. Unknowns become checks, not silent exclusions.
Confirmed patients, possibly eligible patients with the exact checks they need, and what changed since the last data refresh.
Shown to the study team only – patient data never reach a model.
The same eligibility applied three years before the data cut-off – then followed forward. Event rates drive event-driven sample sizes.
| Outcome | Events | Rate / 1,000 py | 1-year risk |
|---|---|---|---|
| Hospitalization for heart failure primary | 412 | 41.3 | 3.8% |
| All-cause death | 298 | 29.6 | 2.7% |
| MI hospitalization | 121 | 11.8 | 1.1% |
New users only (12-month washout), 12 months of baseline records, enrolment 2015–2022, follow-up to the outcome, death, end of records or 3 years.
Kaplan–Meier cumulative incidence by strategy, propensity-score weights, robust hazard ratios, balance diagnostics and an E-value.
Baseline characteristics from ClinicalTrials.gov (only the NCT number is sent), and the demographic footprint of every criterion.
| Characteristic | Trial | EHR | Δ |
|---|---|---|---|
| Age, mean | 63.1 | 65.4 | +2.3 |
| Female | 38.2% | 47.9% | +9.7 |
| Black race | 5.1% | 17.8% | +12.7 |
| HbA1c, mean | 8.1 | 7.9 | −0.2 |
| Heart failure | 10.0% | 14.6% | +4.6 |
| Criterion | Removes | Female | Black | ≥ 65 |
|---|---|---|---|---|
| I3 HbA1c 6.5–11% | 3,410 | 51% vs 46 | 15% vs 18 | 48% vs 44 |
| E3 eGFR < 30 | 412 | 55% vs 47 | 31% vs 17 | 71% vs 45 |
| E5 Pregnancy | 38 | 100% vs 47 | 21% vs 18 | 0% vs 46 |