Loading documentation…
Loading documentation…
PearMDx is the same clinical intelligence engine that powers the PearCare platform — available to your application as a clean REST API. Post symptoms, get back a structured differential diagnosis, triage level, and follow-up questions. This is everything you need to integrate it: API documentation, code examples, and an interactive sandbox.
Hospitals run the PearMDx engine through the PearCare EHR, with no code required. Developers reach the very same engine through this API. One core, two front doors.
The PearMDx clinical engine reaches developers through this REST API and hospitals through the PearCare platform. Same reasoning, same knowledge base, same safety net — exposed two ways.
Post symptoms and demographics, get back a differential diagnosis, a triage level, follow-up questions, red flags, and recommended investigations — structured JSON your application can act on directly.
The engine is tuned for African disease prevalence and local symptom terminology, with location-aware weighting — not a generic Western model wrapped in an endpoint.
Integrate in under 5 minutes with JavaScript, Python, or cURL.
Complete documentation for all 13 documented v1 endpoints, including FHIR R4.
Copy-paste examples for assessment, parsing, and conversational flow.
Test API calls live in your browser with pre-filled requests.
Behind every PearMedica API call is PearMDx — a retrieval-augmented clinical decision support engine grounded in a curated African knowledge base, with multiple independent safety layers ensuring clinical reliability.
A rules-based safety layer flags life-threatening symptom combinations immediately — before the full assessment completes — so red flags surface first. It is designed to minimise missed critical presentations, with ongoing validation under clinical governance.
Assessments are grounded in a curated knowledge base built for conditions prevalent across Africa — 30 conditions fully modelled for MVP, with the knowledge base expanding toward broader African coverage. Not generic Western training data.
Every AI output passes through a 6-layer safety net of independent checks — including hallucination detection, emergency pattern matching, and WHO-guideline compliance — before reaching your application. No single point of failure.
4-level urgency scoring — Emergency, Urgent, Routine or Self-Care — with nearest-facility recommendations based on patient location.
If the AI engine is temporarily unavailable, PearMDx serves safe rule-based responses and reports its status via the health endpoint — your integration degrades gracefully instead of failing hard.
Built-in AI clinical assistant available to staff via PearCare. Answers questions about patients, assessments, and clinical protocols using real facility data.
The API is language-agnostic — any stack that can make an HTTP request works. Here is what teams build on it.
Pre-assess patients at registration or in a 24/7 mobile symptom checker, then route them to the right level of care.
Capture and structure the patient story before the clinician joins, so consultations start with a differential already in hand.
Surface differentials, red flags and recommended investigations inside your existing EHR or nursing-intake workflow.
Send and receive standard FHIR R4 Bundles against /v1/fhir/assess, with a published CapabilityStatement for auto-discovery.
// Assess patient symptoms with one API callconst response = await fetch('https://api.pearmedica.com/v1/assess', {method: 'POST',headers: {'Authorization': 'Bearer sk_live_your_api_key','Content-Type': 'application/json',},body: JSON.stringify({patient: { age: 28, sex: 'female', location: 'lagos_nigeria' },evidence: [{ id: 's_fever', choice_id: 'present', duration_days: 3 },{ id: 's_headache', choice_id: 'present', severity: 'severe' },],}),});const { conditions, triage, next_question } = await response.json();// → conditions: [{ name: "Malaria", probability: "high", ... }]// → triage: { level: "urgent", recommended_action: "..." }
The same engine ships fully wired into PearCare — a complete EHR with outpatient records, billing, inpatient, pharmacy and a patient portal — no code required.