Extract data from patient intake forms
A patient intake form is the packet a clinic hands a new patient on a clipboard, and it is the front door to the medical record. Front-desk and medical-records staff retype these by the hundred, because the data on them has to land in the EHR before a visit can be billed: patient demographics, the insurance carrier and policy number, the reason for the visit, the medical and medication history, allergies, and the signed consents. A primary-care practice or an urgent-care chain onboarding patients all day receives intake forms as completed paper scans, PDFs filled in a portal, and faxed packets from referring providers, each laid out differently and many in a patient's own handwriting.
The challenge is that an intake form mixes structured fields, checkbox grids, and free text, often handwritten. Demographics are discrete (name, date of birth, address, phone), but the medical history is a checkbox list where a patient ticks conditions such as hypertension or Type 2 diabetes and adds notes in the margin. Medication and allergy lists are free text, and a missed allergy is a safety issue. The insurance section carries a carrier such as Aetna, a member ID, a group number, and a policy holder relationship that has to map to the right billing fields. Consent and HIPAA acknowledgment sections need a signature and a date to be valid, and an unsigned consent has to be caught at intake rather than discovered at billing.
Talonic reads the intake form and returns demographics, insurance, history, medications, allergies, and consent status as structured fields, with checkbox grids captured as selected and unselected items and handwriting scored for confidence, so front-desk staff confirm rather than transcribe. A patient seen 2026-05-30 with a 1968-02-14 date of birth hands over demographics, an Aetna member ID, and a history the front desk keys into the EHR; the US clinic stores PHI in the EMR, tags the provider NPI, scans the PDF, runs OCR, checks the consent, and bills the CPT codes in USD.
What gets extracted from patient intake forms
How extraction works for patient intake forms
Patient intake forms come from clipboard scans, patient-portal PDFs, and faxed referral packets, so a single template never holds across practices. Talonic classifies the form and maps it to the patient-intake schema in the Field Registry, which separates demographics, insurance, clinical history, medications, allergies, and consent. Checkbox grids for the medical history are captured as selected and unselected items rather than flattened to text, and handwritten free-text fields such as medications and allergies are returned with per-field confidence so a low-confidence entry is flagged. The insurance block maps the carrier, member ID, group number, and policy-holder relationship to the billing fields a practice management system expects. Consent and HIPAA acknowledgment sections are checked for a signature and date. Every value returns with a confidence score and pixel-region provenance under DIN SPEC 91491 conformity, so staff verify a captured allergy or policy number against the source form.
Sample extraction
A scanned two-page new-patient intake packet
{
"patient_name": "Helen Park",
"date_of_birth": "1968-02-14",
"address": "4120 Cedar Lane, Austin, TX 78704",
"phone": "+1 512 555 0147",
"insurance": {
"carrier": "Aetna",
"member_id": "AET558210934",
"group_number": "GRP-88231",
"policy_holder": "self"
},
"reason_for_visit": "Annual physical and blood pressure follow-up",
"medical_history": [
"hypertension",
"type 2 diabetes"
],
"medications": [
"Lisinopril 10 mg",
"Metformin 500 mg"
],
"allergies": [
"penicillin"
],
"consent": {
"signed": true,
"signed_date": "2026-05-30",
"hipaa_acknowledged": true
}
}Frequently asked
Does it capture checkbox medical-history grids?
Yes. A history grid is returned as selected and unselected conditions rather than a single text blob, so a practice can load discrete problems such as hypertension and Type 2 diabetes into the record.
How does it handle handwriting?
Handwritten fields such as medications and allergies are run through OCR with per-field confidence scoring, so a low-confidence allergy entry is flagged for staff to confirm rather than entered silently, which matters because a missed allergy is a safety issue.
Does it check that consents are signed?
Yes. Consent and HIPAA acknowledgment sections are checked for both a signature and a date, so an unsigned or undated consent is caught at intake instead of surfacing later during billing.
What fields are extracted from patient intake forms?
Talonic returns patient intake forms as schema-validated, typed fields. Common fields include Patient Name, Date of Birth, Address, Insurance Carrier, 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 patient intake forms, 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 patient intake forms 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 patient intake forms?
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 patient intake forms?
Author note
Reviewed by Talonic engineering, schema review · last reviewed 2026-06-09