Read clinical and claims paperwork, and flag what looks misread.
Claim forms, lab reports and prescriptions, with code, provider-ID and reference-range validation that surfaces misread values.
Built for: Payers, third-party administrators, clinics and health-tech teams handling faxes, scans and forms.
See it on a sample
Run the checks yourself
No sign-up. Read a sample case, see what gets flagged, then correct a value the way a reviewer would.
Lab report - fictional patient
One report, four results, each printed with its reference range and an H / L / N flag.
Lab reports
Chemistry panel
- Potassium (mmol/L) - flag NReference range 3.5 to 5.1
- 41
- Glucose, fasting (mg/dL) - flag HReference range 70 to 99
- 180
- Hemoglobin (g/dL) - flag NReference range 12.0 to 15.5
- 13.8
- Creatinine (mg/dL) - flag NReference range 0.6 to 1.2
- 0.9
Automatic checks
Run the pack to read the report and check it.
Fictional sample data, checked in your browser. The product extracts and validates what is printed; it does not interpret results or advise on care.
Try it on your own filesIn this pack
Every document is verified on its own
Claim forms
Patient, policy, provider, diagnoses and procedure charges.
- Provider NPI passes its Luhn checksum
- Diagnosis codes are well-formed ICD-10
- Procedure charges add up to the total charge
- Date of birth is before the service date
Lab reports
Patient, specimen and every analyte with unit, range and flag.
- Each H/L/N flag agrees with the value against its reference range - a misread decimal shows up here
- Specimen and report dates are in order
Prescriptions
Patient, prescriber and each prescribed item with strength and directions.
- Patient, prescriber and issue date are present
- At least one prescribed item was read
How it works
A batch of claim packets
- 1Upload claim forms with their lab reports and prescriptions.
- 2Codes, provider IDs and totals are validated on each document.
- 3Lab flags are checked against their reference ranges, which exposes misread decimals.
- 4Flagged items go to a reviewer; the rest are ready to export.
Checks flag what does not add up; they do not guarantee every value is correct. Flagged items are queued for a person with the source page highlighted, and the human review tool can be switched on or off for your organization.
Try the Healthcare & claims pack on your own files.
Create a case, drop in a real folder of documents and see what gets caught. Part of our Healthcare and insurance solutions.
FAQ