Clinical Decision Support · Research Preview

PenuX in the Clinical Workflow

From emergency admission to ICU triage — how AI-assisted risk stratification fits the acute pancreatitis patient journey.

The AP Patient Pathway

Six key steps from ED presentation to early targeted intervention. PenuX integrates at step 3, adding an automated risk score with no additional tests required.

Step 1
ED Arrival
Patient presents with epigastric pain, nausea, vomiting. Serum amylase/lipase >3× ULN. AP diagnosis established per 2012 Revised Atlanta Classification.
Step 2
Routine Labs Drawn
Standard admission panel: WBC, CRP, LDH, creatinine, BUN, calcium, glucose, hematocrit, AST, ALT, albumin, bilirubin, lipase, amylase… No extra tests needed.
Step 3 · PenuX
PenuX Scores Risk
59 lab features fed to the Random Forest model via FHIR R4 or HL7 v2. Severe-AP probability returned in <200 ms. Risk tier assigned automatically.
AUROC 0.877
Step 4
Clinical Alert
If P(severe) ≥ 0.535: alert fires to attending physician and ICU team. If Low risk: standard ward pathway proceeds.
Threshold 0.535 · Sensitivity 96.8%
Step 5
Triage Decision
Physician reviews PenuX score alongside clinical context and imaging. High → ICU/HDU. Moderate → enhanced monitoring, repeat labs at 6 h. Low → gastro ward.
Step 6
Early Intervention
Aggressive IV fluid resuscitation (Ringer's lactate 250–500 mL/hr), NPO, analgesia, continuous organ-function monitoring, early enteral nutrition when tolerated.

PenuX result available within 4 hours of admission — before Ranson (48 h), BISAP (24 h), and APACHE-II (24 h). No CT scan, no mental-status exam, no second blood draw required.

Risk Tier Definitions

PenuX returns one of three risk tiers based on the predicted probability of severe AP, aligned with the 2012 Revised Atlanta Classification.

🟩
Low Risk
P(severe) < 0.35

Mild AP likely

  • Standard gastroenterology ward admission
  • Oral fluids when clinically tolerated
  • Routine observations every 4–6 hours
  • Clinical reassessment at 24 hours
  • Early mobilisation encouraged
🟡
Moderate Risk
P(severe) 0.35 – 0.535

Moderately severe AP possible

  • Enhanced observation frequency (2-hourly)
  • Repeat labs at 6 h (CRP, creatinine, WBC)
  • Consider Intermediate Care Unit (IMC/HDU)
  • IV fluid support with strict fluid balance
  • Gastroenterology / HPB consultant review
🔴
High Risk
P(severe) ≥ 0.535

Severe AP likely — act immediately

  • ICU / HDU admission — activate care pathway
  • Ringer's lactate 250–500 mL/hr IV
  • Continuous organ-function monitoring
  • Early HPB and intensive-care specialist activation
  • Early enteral nutrition when haemodynamically stable

Threshold 0.535 achieves 96.8% sensitivity (TP=566, FN=19) with 38.7% specificity on 5-fold CV (n=722). Clinical context always takes precedence over the score.

PenuX vs Existing Severity Scores

Comparing timing, data requirements, and key differentiators of PenuX against established clinical scoring systems for acute pancreatitis.

Score When Available Data Required Key Limitation PenuX Advantage
Ranson Criteria 48 hours 11 criteria across two time-points (admission + 48 h) Not usable at presentation; requires a second data set at 48 h PenuX result available at admission
BISAP 24 hours BUN, SIRS criteria, mental status, age, pleural effusion Subjective mental-status component; 24 h wait PenuX needs no clinical exam findings
APACHE-II 24 hours 12 physiological parameters + Glasgow Coma Scale Complex, time-consuming; GCS unreliable at admission PenuX uses routine labs only; fully automated
CTSI CT required Contrast-enhanced CT (Balthazar grade + % necrosis) Radiation, contrast risk, costly, delayed availability PenuX requires no imaging
Marshall Score 24 hours Respiratory, renal, cardiovascular parameters Organ-failure focus; less predictive at early presentation PenuX predicts severity, not only organ failure
PenuX-AP-Severity This work < 4 hours Routine admission labs (59 features from 19 parameters) Research prototype; single Chinese cohort n=722; external validation pending Earliest available · Fully automated · No imaging · No extra tests

* Comparative figures are from published meta-analyses. PenuX AUROC 0.877 from 5-fold stratified cross-validation on n=722; no external test cohort has been evaluated.

Privacy & Clinical Safety

PenuX is designed to augment clinical judgment — never to replace it. These principles are non-negotiable by design.

🔒

De-identified at Source

Patient date-of-birth, MRN, and name are stripped before the API call is formed. The PenuX server never receives or retains identifiable patient data.

🏗

No Data Retention

Lab values transmitted to the API are used solely for inference. No request payload is persisted. GDPR Article 5(1)(e) storage-limitation principle is upheld by design.

🧍

Physician Retains Full Authority

PenuX provides a probability score and a suggested risk tier. All clinical decisions — admission, treatment, escalation — remain exclusively with the treating physician.

📋

Transparent Uncertainty

The API response lists which fields were used and which are missing, so clinicians can gauge confidence. Missing critical labs are flagged explicitly in every response.

🚧

Research Prototype — Not Approved

PenuX has not received CE mark, FDA 510(k) clearance, or any other regulatory approval. It must not be used as the sole basis for any clinical decision.

📈

Cohort Bias Disclosed

Training data is a single-centre Chinese cohort (n=722). Performance may differ across other ethnic or geographic populations. External validation is actively sought.

⚠️

"The model augments clinical judgment; it does not replace it. A high PenuX score is a prompt to look harder — not a substitute for the physician's examination, imaging, and holistic evaluation of the patient."

EHR Integration Data Flow

PenuX slots into your existing clinical infrastructure via FHIR R4 or HL7 v2 — no proprietary connector required.

🏠
EHR / LIS
Camelion · Epic · Cerner
FHIR R4 Bundle
or HL7 v2 ORU
🔄
Lab Results
LOINC-coded values
HTTPS POST
/fhir/predict
PenuX API
Random Forest · <200 ms
JSON response
P(severe) + tier
🔔
Risk Score
Low / Moderate / High
Alert if
P ≥ 0.535
👁️
Physician Dashboard
EHR alert · Mobile pager
Example API Response
{
  "severe_ap_probability": 0.782,
  "risk_group":            "High",
  "recommendation":       "Consider ICU/HDU admission. Early aggressive fluid resuscitation.",
  "atlanta_classification": "Potentially Severe",
  "fields_used":          ["wbc", "crp", "creatinine", "glucose", "ldh", "ast"],
  "missing_fields":       ["bun", "calcium", "albumin"],
  "model_version":        "1.0.0"
}
View Swagger UI GitHub Repository

Stage 0 — Seeking Hospital Data Partners

We are currently in Stage 0 — building hospital data partnerships before formal validation begins. We are looking for gastroenterology or acute-medicine departments willing to provide de-identified retrospective SAP data, co-develop an IRB protocol, and serve as the first external validation site. Co-authorship offered. Data stays on-site — PenuX supplies the model, pipeline, and analysis.

📋 n=220 validation cohort#129309; Data partnership MOU
🕐 12-month enrolment
🖊 Co-authorship offered
⚖ Shared IRB framework

PI: Netanel Stern — PenuX Research Initiative