Data model

Tables and fields behind each analysis

Ask iZO
DB
Synthetic, de-identified, aggregated data. Illustrative demo.
Data model

What data do these analyses need?

Analysis to data map

Select an analysis to highlight the tables it needs

AnalysisBusiness questionRequired tablesKey metrics
CohortWho are our new RA patients?
patientdiagnosislab_resultprescriptioncohort_membership
Cohort size · Median age · % with comorbidity · % biologic-naive
Therapy FlowWhat happens after first-line therapy?
therapy_lineprescriptionproductregion
Switch rate by line · Brand X switch win share · Competitor win share
Drop-offWhere are we losing patients?
prescriptionbenefit_verificationprior_authpayer_planregion
Abandonment rate · PA rejection rate · Top rejection reason · Regional gap
ActWho needs to act on this?
practicepayer_planprior_authprescription
Practices flagged · Plans to engage · PA support volume
ImproveDid our access push work?
prescriptionprior_auththerapy_linedata_refresh
Quarterly abandonment · Switch share trend · Regional gap trend

Tables

13 tables in 5 groups

Patient
Clinical
Rx and access
Reference
Derived

prescription

Every advanced-therapy prescription and how far it got. Drives the drop-off funnel.

Source
Pharmacy claims and specialty hub
Grain
One row per new prescription
Used by
Cohort, Therapy Flow, Drop-off, Act, Improve
FieldTypeDescription
rx_idstringPrescription identifier
patient_tokenstringLink to patient
product_idstringBrand X, competitors, conventional
prescriber_practice_idstringWriting practice
plan_idstringPayer plan billed
written_monthmonthMonth written
statusenumWritten, benefits verified, PA submitted, approved, filled, abandoned, reversed
first_fill_monthmonthNull when abandoned
is_new_to_brandboolNo prior fill of the same product

Primary key Link to another table

Privacy rules for real data

Required before any real data is loaded

  • Patients are identified only by a hashed token from a certified de-identification vendor.
  • Dates are kept at month level and ages in 5-year bands.
  • Geography is rolled up to region; no zip codes or addresses.
  • Results under the minimum cell size are suppressed before display.
  • Field insight is aggregated to practice level, never patient level.
  • Each refresh is logged for the audit-ready report.
Learn and improve.Every data refresh sharpens the next question and the next action.