Pogosh CDS — Structured Clinical Decision Support

What it does

Pogosh is a structured clinical decision support engine covering 413 conditions across 40+ specialties. When a clinician opens a chart, Pogosh surfaces the high-yield labs that need to be ordered, flags the ones that are missing, identifies escalation criteria based on current vitals, recommends imaging with patient-specific alternatives, and tells the team exactly which specialist to call and why — all in under 200 milliseconds.

The problem it solves

Clinicians managing six patients at once miss things. Not because they don’t know the medicine — because cognitive load under time pressure creates gaps. Blood cultures that should be drawn before antibiotics in a high-PSI pneumonia patient. Urine antigens in severe presentations. Admit criteria triggered by SpO2 + respiratory rate that nobody caught because the data came in between handoffs. The information exists in guidelines, but nobody has time to look it up for every patient, every interaction, every shift.

What the engine provides

• High-yield labs — not a generic lab panel, but the specific tests that change decision-making for this condition. LOINC codes, rationale, and critical thresholds. For pneumonia: procalcitonin for bacterial vs. viral differentiation, blood cultures required before antibiotics if PSI Class III–V.
• Imaging guidance — first-line studies and alternatives based on patient-specific factors. Shellfish allergy flags CTPA, recommends V/Q scan. CKD present? Contrast exposure noted.
• Specialist routing — when to call pulmonology vs. infectious disease vs. send to the ICU, with the specific indication to give them. Actual routing criteria, not “consider specialist referral.”
• Escalation triggers — specific vitals and lab thresholds that indicate deterioration. SpO2 30 in a pneumonia patient = admit criteria met. Flagged before the team catches it.
• Missing order alerts — cross-references what should be ordered against what has already been ordered or resulted. CBC back but blood cultures not sent in a high-PSI patient? Surfaced immediately.

How it works

• REST API deployed on Cloudflare’s global edge network
• Your EMR, clinical app, or AI assistant calls the API with a condition ID and patient context
• Engine returns structured JSON with patient-specific recommendations in under 200ms
• Runs as a copilot — called on every chart interaction, not just once per encounter
• For OpenEMR: ready-to-install module adds a copilot panel directly in the patient chart view

Key features

• 413 clinical conditions spanning 40+ specialties
• Thousands of traceable clinical rules built from peer-reviewed guidelines
• REST API with <200ms global response time
• OpenEMR module — ready to deploy
• Every recommendation tied to a condition, guideline, and clinical rationale — not a black box
• Patient-specific adjustments for comorbidities, allergies, and existing orders
• Not an AI chatbot — a structured rule engine with zero hallucination risk

Specialty coverage

Emergency Medicine (STEMI, PE, sepsis, stroke, anaphylaxis, DKA), Internal Medicine (pneumonia, CHF, AKI, hepatic encephalopathy), Interventional Radiology (ablation planning, anticoagulation bridging, post-procedure care), Gastroenterology (GI bleed, Crohn’s, bowel obstruction, cholangitis), Pulmonology, Nephrology, Neurology, Cardiology, Ophthalmology, Orthopedics, Urology, OB/GYN, Hematology/Oncology, Endocrinology, Infectious Disease, Rheumatology, Dermatology, ENT, and more.

Who it’s for

• Urgent care and primary care — copilot for every encounter: chart open, labs resulted, disposition decision
• Emergency departments — fast, specialty-aware, patient-specific. Distinguishes first PE in a 30-year-old from prior DVT on anticoagulation in a 70-year-old
• Specialty practices — GI nurses managing IBD, ophthalmology triage, orthopedic pre-op planning
• EMR developers — clean JSON, REST API, OpenEMR module ready to integrate
• Health systems — reduce cognitive load on nursing staff, standardize workup, catch missed orders during busy shifts

Clinical Validation Partner Program

We are selecting 10 clinical sites to deploy Pogosh CDS in real patient care workflows. Partners receive 3 months of full access at no cost, the complete condition library, OpenEMR module or API access, and a direct line to the development team. In exchange, we ask for structured monthly feedback on clinical accuracy, workflow integration, and edge cases. This is not a free trial — it’s a collaboration.

Pricing

Clinic: $149/month (up to 5 providers). Practice: $299/month (up to 15 providers). Enterprise: custom. All tiers include the full condition library, unlimited API calls, and the OpenEMR module. No per-call fees.

Technical notes

REST API on Cloudflare’s global edge. Structured JSON responses. LOINC-coded lab recommendations. Designed for structured CDS, not PHI storage — the API accepts condition IDs and clinical parameters, not patient identifiers. OpenEMR module deployments keep all data within existing EMR infrastructure.

Privacy / data handling

Pogosh processes clinical parameters and condition identifiers — not patient names, MRNs, or other PHI. For OpenEMR deployments, the module runs within your existing infrastructure. Do not submit patient identifiers to the standalone API. This tool provides clinical decision support and is not a substitute for clinical judgment.

Reviewed by Pouyan Golshani, MD, Interventional Radiologist — March 1, 2026