> ## Documentation Index > Fetch the complete documentation index at: https://omniloy.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Patient Data Structure > JSON format to configure SofIA chat with the patient's clinical history and complete context The `patientdata` property allows you to provide the patient's clinical history and context to SofIA. While optional, it is **strongly recommended** — it enables SofIA to reference the patient's background when generating reports and answering clinical questions. ## Data Structure The `patientdata` prop accepts a JSON object with the following structure: | Field | Type | Description | | ----------------- | ------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **fullName** | `string` | Patient's full name | | **birthDate** | `string` | Date of birth (any format: `YYYY-MM-DD`, `DD/MM/YYYY`, etc.) | | **phone** | `string` | Contact phone number | | **address** | `string` | Patient's address | | **extraData** | `object` | Flexible key-value object for any additional clinical data | | **signedConsent** | `{ signed: boolean, date?: string } or undefined` | Indicates whether the patient has signed the consent form to be recorded. Optional. The consent indicator must be enabled in your Omniloy account settings for this to be shown (contact [support@omniloy.com](mailto:support@omniloy.com)). | The fields `fullName`, `birthDate`, `phone`, and `address` are automatically anonymized in AI processing. SofIA replaces these values with placeholders before sending data to the AI engine, then restores them in the final output. This ensures HIPAA/GDPR compliance. All patient data shown in the examples below is entirely fictional. Never use real patient information in configuration examples, testing environments, or source code. ## Usage Examples ### Minimal Example The simplest usage with just the patient's identity: ```json theme={null} { "fullName": "Jane Doe", "birthDate": "1985-05-15" } ``` ### Standard Example A typical usage including contact information and medical history: ```json theme={null} { "fullName": "Jane Doe", "birthDate": "1985-05-15", "phone": "6425482447", "address": "123 Main Street, City", "extraData": { "medical_practice": "Cardiology", "patient_medical_notes": { "notes": [ { "consultation": "Cardiology follow-up", "date": "2024-10-25", "summary": "Stable blood pressure with current medication. ECG normal." }, { "consultation": "General checkup", "date": "2024-08-12", "summary": "Routine annual exam. All vitals within normal range." } ] } } } ``` ### Comprehensive Example Full usage with multiple clinical data sources: ```json theme={null} { "fullName": "Jane Doe", "birthDate": "1985-05-15", "phone": "6425482447", "address": "123 Main Street, City", "signedConsent": { "signed": true, "date": "2025-08-01" }, "extraData": { "medical_practice": "Internal Medicine", "patient_medical_notes": { "notes": [ { "consultation": "Endocrinology", "date": "2024-11-01", "url": "https://ehr.example.com/notes/12345", "summary": "HbA1c at 7.2%. Adjusted metformin dosage." } ] }, "allergies": ["Penicillin", "Sulfonamides"], "active_medications": [ {"name": "Metformin", "dose": "1000mg", "frequency": "twice daily"}, {"name": "Lisinopril", "dose": "10mg", "frequency": "once daily"} ], "lab_results": { "date": "2024-10-28", "values": { "glucose_fasting": "142 mg/dL", "HbA1c": "7.2%", "creatinine": "0.9 mg/dL" } }, "chronic_conditions": ["Type 2 Diabetes", "Hypertension"] } } ``` ## The `extraData` Field The `extraData` field is a flexible `Record` that accepts any key-value structure. This is where you provide the clinical context that SofIA will use during the consultation. **Common fields used in `extraData`:** | Field | Description | | ----------------------- | --------------------------------------------------------------------------------------------------------------- | | `medical_practice` | The doctor's specialty (e.g., "Cardiology", "Pediatrics"). Helps SofIA tailor responses to the clinical context | | `patient_medical_notes` | Previous consultation notes. Include `url` fields if you want SofIA to cite the source | | `allergies` | Known allergies for safety cross-referencing | | `active_medications` | Current medications for interaction checking | | `lab_results` | Recent laboratory values | | `chronic_conditions` | Ongoing medical conditions | Include a `url` field in medical notes if you want SofIA to cite the data source when guiding the doctor. SofIA will reference the URL in its responses so the doctor can verify the original record. ## Passing `patientdata` to the Component ### As HTML Attribute Since `patientdata` is a JSON type, it is automatically parsed by the SDK: ```html theme={null} ``` ### Programmatically (JavaScript) ```javascript theme={null} const sofiaElement = document.querySelector('sofia-sdk'); const patientData = { fullName: "Jane Doe", birthDate: "1985-05-15", extraData: { medical_practice: "Cardiology", allergies: ["Penicillin"] } }; sofiaElement.setAttribute('patientdata', JSON.stringify(patientData)); ``` ### React ```jsx theme={null} const [patientData, setPatientData] = useState({ fullName: "Jane Doe", birthDate: "1985-05-15", extraData: { medical_practice: "Cardiology" } }); return ( ); ``` ## Limits and Validation | Rule | Details | | ------------------- | ------------------------------------------------------------------------ | | **Maximum size** | 100 KB | | **Error on exceed** | The component rejects with error: `"patientData exceeds 100 kB"` | | **Format** | Must be valid JSON (or a JavaScript object when passed programmatically) | | **Required fields** | None — all fields are optional | ## Automatic Anonymization SofIA automatically anonymizes the following personal fields before sending data to the AI engine: * `fullName` * `birthDate` * `phone` * `address` These values are replaced with neutral placeholders during AI processing and restored in the final output. This ensures that personally identifiable information (PII) never reaches the AI model directly. Data inside `extraData` is **not** automatically anonymized. Avoid placing raw PII (social security numbers, government IDs, etc.) in `extraData` fields. Use anonymized identifiers when possible. ## Next Steps 1. [Integrate with your framework](/sofia/en/sdk/vanilla) 2. [Design clinical data schemas](/sofia/en/sdk/templates) 3. [Test your integration](/sofia/en/sdk/testing)