Extract data from prescription orders
A prescription order is the instruction that tells a pharmacy what to dispense, and a pharmacy intake team reads dozens an hour, often from scans and faxes that have to be transcribed before anything reaches a patient. Each order names the drug, its strength, the directions for use, the quantity, and the refills, plus the prescriber and the patient. A retail or mail-order pharmacy processing a high volume sees orders for a maintenance medication at a specific strength, a controlled substance that triggers extra checks, and a compounded preparation, each written by a different prescriber on a different system, some printed from an EHR and some handwritten on a pad and faxed.
Difficulty concentrates in the sig and the drug identity. Directions, the sig, are written in a shorthand that has to be expanded correctly: "1 tab PO BID" means one tablet by mouth twice daily, and "ii gtt OU QHS" means two drops in both eyes at bedtime. Misreading the sig becomes a dispensing error. Drug names must resolve to a specific product and strength, and the National Drug Code (NDC) when present pins it exactly. Quantity and days supply have to be consistent with the sig. A controlled substance carries a schedule and, on an electronic order, the prescriber's DEA number in its standard two-letter-then-seven-digit format. Refills and the date written govern how long the order is valid.
Talonic reads the prescription order and returns the drug, strength, sig expanded to plain language, quantity, refills, and prescriber detail including the DEA number when present, so a pharmacy intake system receives structured data and flags controlled substances for the checks they require. Written 2026-05-30 and printed from the EHR, an order billed in USD carries the drug, the strength, and the sig before the US pharmacy dispenses and posts the claim to the payer.
What gets extracted from prescription orders
How extraction works for prescription orders
Prescription orders arrive as e-prescribing transmissions printed to PDF, EHR printouts, and handwritten scripts faxed to the pharmacy, so the format ranges from structured to handwritten. Talonic classifies the order and maps it to the prescription schema in the Field Registry, which captures the drug, strength, sig, quantity, refills, prescriber, and patient. Sig shorthand is expanded to plain language: "1 tab PO BID" becomes one tablet by mouth twice daily, with the dose, route, and frequency held as discrete fields. Drug identity resolves to a product and strength, and a National Drug Code is captured when present. Prescriber DEA numbers are validated against the standard format, and a controlled-substance order is tagged with its schedule. Every value returns with a confidence score and pixel-region provenance under DIN SPEC 91491 conformity, so a pharmacist can verify the drug and sig against the source order before dispensing.
Sample extraction
A single-item prescription order printed from an EHR
{
"patient_name": "Helen Park",
"patient_dob": "1968-02-14",
"drug_name": "Lisinopril",
"strength": "10 mg",
"dosage_form": "tablet",
"sig_raw": "1 tab PO QD",
"sig_expanded": "1 tablet by mouth once daily",
"dose": "1 tablet",
"route": "oral",
"frequency": "once daily",
"quantity": 90,
"days_supply": 90,
"refills": 3,
"prescriber_name": "Dr. Alan Reyes",
"prescriber_dea": "BR1234563",
"date_written": "2026-05-30",
"controlled_substance": false
}Frequently asked
Does it expand the sig shorthand?
Sig text is expanded reliably. Directions written as "1 tab PO BID" or "ii gtt OU QHS" become plain language split into discrete dose, route, and frequency fields, because a misread sig is a dispensing risk rather than a formatting nuisance.
How are controlled substances handled?
Controlled-substance orders are tagged with their schedule, and the prescriber DEA number is captured and checked against its standard two-letter-then-seven-digit format, so an intake system can route the order through the additional verification it requires.
Can it read handwritten and faxed prescriptions?
Handwritten and faxed scripts are supported. They run through the same schema with per-field confidence scoring, so a low-confidence drug name or strength is flagged for a pharmacist to verify rather than dispensed on a guess.
What fields are extracted from prescription orders?
Talonic returns prescription orders as schema-validated, typed fields. Common fields include Patient Name, Drug Name, Strength, Sig (Directions), and more, each normalized (dates to ISO 8601, amounts as numbers) and mapped to a stable key so the output shape stays the same across layouts.
How accurate is extraction from prescription orders, and how is confidence reported?
Every extracted cell carries a confidence score from 0.0 to 1.0 and a provenance pointer back to the source page and region, so low-confidence values can be reviewed against the original before the data is trusted downstream. There is no single accuracy number: confidence is reported per field so you can gate on it.
Can I use prescription orders extraction in production?
Yes. The same engine behind this guide is available as a production REST API and Node SDK with sync, async, and streaming modes, schema versioning, signed webhooks, and EU-resident processing. Start free with an API key, then scale on usage-based pricing.
What does it cost to extract data from prescription orders?
There is a free tier for prototyping and agent evaluation with no credit card. Paid usage is credit-based at 1,000 credits per euro: page ingestion is 100 credits per page and registry-resolved queries are free. See talonic.com/pricing for current rates.
Ready to extract from your own prescription orders?
Author note
Reviewed by Talonic engineering, schema review · last reviewed 2026-06-15