Emergency Medicine
Emergency departments present simultaneously ideal and terrible conditions for AI. Ideal because large volumes of structured data (vitals, labs, imaging) feed clear time-sensitive outcomes. Terrible because time pressure precludes verification, missing data is common, patient heterogeneity is extreme, and consequences of errors are immediate and severe. Stroke AI has proven clinical benefit. Sepsis AI remains oversold despite widespread deployment.
After reading this chapter, you will be able to:
- Identify validated AI applications for emergency departments
- Understand AI tools for ICU monitoring and early warning
- Evaluate sepsis prediction systems critically
- Assess stroke, trauma, and cardiac emergency AI tools
- Navigate implementation challenges specific to acute care settings
- Recognize failure modes in high-stakes environments
- Balance speed, accuracy, and safety in AI-assisted acute care
High-Impact AI Applications in Emergency/Critical Care
1. Stroke Detection and Triage
Large Vessel Occlusion (LVO) Stroke Detection:
Systems: Viz.ai, RapidAI, Brainomix FDA Status: Multiple clearances Function: Detect LVO on head CT angiography, alert stroke team immediately Evidence: Reduces time-to-treatment by 30-50 minutes (Figurelle et al., 2023) Clinical impact: Improved functional outcomes (mRS scores) Deployment: 1700+ hospitals (Viz.ai alone)
How it works: 1. Patient gets head CTA 2. AI analyzes immediately 3. Positive LVO → Automated alerts to stroke team, neuroIR, neurosurgery 4. Team mobilized before radiologist reads 5. Faster door-to-groin time
Performance: 90%+ sensitivity for LVO
Limitations: - False positives (vessel anatomy mimics occlusion) - Small vessel occlusions may be missed - Requires CT angiography (not plain CT)
Intracranial Hemorrhage (ICH) Detection:
Systems: Aidoc, Viz.ai, RapidAI FDA Status: Cleared Function: Flag ICH on non-contrast head CT, prioritize worklist Evidence: Reduces time-to-neurosurgery notification (Arbabshirani et al., 2018) Sensitivity: >95% for most hemorrhage types
Use cases: - ED triage (identify critical studies) - ICU monitoring (post-procedure surveillance) - Trauma (rapid identification)
Limitations: - Small subarachnoid hemorrhages sometimes missed - Calcifications, artifacts can cause false positives - Subtle bleeds in posterior fossa challenging
2. Pulmonary Embolism (PE) Detection
CT Pulmonary Angiography AI:
Systems: Aidoc, Avicenna.AI FDA Status: Cleared Function: Detect PE on CTPA, triage positive studies Sensitivity: 90-95% for proximal PE
Clinical benefit: - Faster identification of time-sensitive PEs - Reduced turnaround time for anticoagulation - Worklist prioritization in busy EDs
Limitations: - Subsegmental PE (clinical significance debated, hard to detect) - Motion artifacts reduce accuracy - Chronic vs. acute PE differentiation imperfect
3. Sepsis Prediction and Early Warning
Epic Sepsis Model (Controversial):
Most widely deployed sepsis AI Evidence: MIXED AND CONCERNING
External validation (Michigan Medicine, Wong et al. 2021): - 33% sensitivity (missed 67% of sepsis cases) (Wong et al., 2021) - 12% PPV (88% of alerts were false positives) - Alert fatigue documented - Clinical benefit unproven
Why it struggles: - Sepsis definition ambiguous (clinical judgment, not algorithmic) - Confounding by treatment (sepsis suspicion → antibiotics → AI detects antibiotics, not sepsis) - Missing data patterns (vital signs checked more frequently in sick patients → AI learns correlation) - Real-time prediction harder than retrospective
Deployment reality: - Widely used despite limited validation - User trust low (many ignore alerts) - Some institutions disabled after poor performance
Alternative approaches with better evidence:
SOFA/qSOFA scores enhanced with ML: - Continuous monitoring models - Better than Epic model in some studies - Still imperfect
Vital sign trajectory analysis: - AI detects subtle trends preceding deterioration - Promising but requires prospective validation
Clinical bottom line on sepsis AI: Demand LOCAL validation before deployment, HIGH false positive rates expected, physician oversight essential, NO AI should replace clinical judgment for sepsis
4. Cardiac Emergency AI
ECG Interpretation AI:
Atrial Fibrillation Detection: - Apple Watch, Kardia, others - High sensitivity for AFib - Problem: Low PPV (many false positives) (Perez et al., 2019) - Flood of patient-reported AFib alerts to EDs
STEMI Detection: - AI analysis of 12-lead ECG - FDA-cleared systems available - Reduces door-to-balloon time - Activates cath lab automatically
Hidden MI patterns: - AI detects subtle STEMI equivalents - Posterior MI, Wellens syndrome - Better than many emergency physicians
Evidence: Improving rapidly, deployment expanding
Cardiac Arrest Prediction:
ICU deterioration models: - Predict cardiac arrest 6-12 hours before event - Allow preemptive interventions (transfer to ICU, advanced monitoring) - Evidence: Some RCTs show benefit
5. Trauma and Critical Injuries
Rib Fracture Detection (CT):
Systems: Aidoc, others Function: Detect rib fractures on chest CT Use case: Trauma, elderly falls Benefit: Identify occult fractures, guide pain management, detect instability
C-Spine Fracture Detection:
Systems: Aidoc FDA cleared Function: Detect cervical spine fractures Use case: Trauma workup, reduce missed fractures
Pneumothorax Detection:
Systems: Oxipit, Lunit, Aidoc Function:** Detect PTX on chest X-ray or CT Sensitivity: >95% Use case: Trauma, post-procedure, ICU monitoring
Clinical impact: Faster decompression, reduced tension PTX complications
Intracranial injury triage (TBI):
AI predicts which patients need neurosurgical intervention Evidence: Can guide transfers from community EDs to trauma centers
6. ICU Early Warning Systems
Patient Deterioration Prediction:
Continuous monitoring AI: - Analyzes vital signs, labs, medications in real-time - Predicts deterioration 6-24 hours ahead - Use cases: Ward→ICU transfer decisions, ICU resource allocation
Evidence: - Some RCTs show reduced code blues, ICU transfers - Other studies show no benefit (alert fatigue) - Variable results depend on implementation
Systems: - Epic Deterioration Index - WAVE Clinical Platform (ExcelMedical) - Various hospital-developed models
Challenges: - False positive rates 90%+ (for rare events like cardiac arrest) - Clinicians ignore frequent alerts - Lack of actionable interventions for many alerts
Mechanical Ventilation AI:
Weaning prediction: - AI predicts readiness for extubation - Reduces ventilator days in some studies - Personalized vent settings
Lung-protective ventilation: - AI optimizes PEEP, tidal volume - Reduces ARDS complications
Status: Research stage mostly, some commercial systems emerging
Acute Kidney Injury (AKI) Prediction:
Real-time AKI risk scores: - Predict AKI 24-48 hours before creatinine rises - Allow preemptive interventions (fluid management, nephrotoxin avoidance)
Evidence: Improving, prospective trials ongoing
7. ED Triage and Workflow Optimization
ESI (Emergency Severity Index) Augmentation:
AI-assisted triage: - Predicts acuity, resource needs - Reduces undertriage - Variable evidence for benefit
Chest Pain Risk Stratification:
AI predicts 30-day MACE (major adverse cardiac events): - Better than HEART score alone in some studies - Reduces unnecessary admissions - Safely identifies low-risk patients for discharge
Wait Time Prediction:
AI forecasts ED volume, wait times: - Staffing optimization - Patient communication
Disposition Prediction:
AI predicts admission vs. discharge: - Bed management - Transfer coordination
What Does NOT Work Well:
Autonomous triage without physician oversight: Too many edge cases and liability concerns
Sepsis AI as standalone diagnostic: High false positives, missed cases, and need for clinical judgment
Alert systems without actionability: Warnings without clear intervention pathways create alert fatigue
Black-box predictions without explanation: Clinicians need to understand WHY a patient was flagged
One-size-fits-all thresholds: Optimal operating points vary by institution and patient population
Implementation Challenges Specific to Emergency/Critical Care:
1. Time Pressure: - AI must be FASTER than current workflow or provide substantial value - No time for complex interactions - Alerts must be actionable immediately
2. Incomplete Data: - ED presentations often lack history, prior records - Missing data common - AI must handle missingness gracefully
3. Heterogeneity: - Extreme patient diversity (age, acuity, comorbidities) - Undifferentiated presentations - AI trained on specific populations may fail
4. Alert Fatigue: - ICUs already have alarm overload - Adding AI alerts risks desensitization - Critical: Optimize thresholds for YOUR false positive tolerance
5. Workflow Integration: - Busy clinicians cannot switch to separate systems - Must integrate into EHR and monitor displays - Mobile alerts must reach the right people at the right time
6. Liability in High-Stakes Settings: - Missed diagnoses in ED/ICU carry high malpractice risk - Over-reliance on AI vs. under-utilization both risky - Documentation of AI use and overrides essential
7. Shift Work and Handoffs: - Multiple clinicians per patient - AI alerts must persist across handoffs - Continuity challenges
Best Practices for Emergency/Critical Care AI:
Pre-Deployment: - LOCAL retrospective validation (YOUR patients, YOUR data) - Prospective silent mode testing (3-6 months minimum) - False positive rate assessment (calculate alerts per shift) - Workflow mapping (where alerts go, who responds, what actions) - Clinical champion identification (ED/ICU physician leader)
Deployment: - Gradual rollout (pilot unit → full ED/ICU) - Threshold optimization (balance sensitivity vs. alert burden) - Mobile alert systems (push notifications to responsible clinicians) - Clear escalation pathways (what to do when AI flags patient) - Override mechanisms (clinicians can dismiss with documentation)
Post-Deployment: - Weekly performance monitoring (initial 3 months) - User feedback collection (alert fatigue assessment) - False positive/negative tracking - Clinical outcome monitoring (does it improve patient outcomes?) - Continuous threshold adjustment (refine based on real-world performance)
Red Flags to Stop/Revise: - Alert response rate below 50% (clinicians ignoring alerts) - More than 10 false positives per shift (unsustainable alert burden) - User complaints escalating - Adverse events potentially related to AI (missed cases, over-reliance) - Performance drift detected (accuracy declining)
Evidence-Based Assessment by Application:
Strong Evidence (Deploy with confidence):
LVO stroke detection (Viz.ai, RapidAI): Multiple prospective studies, proven clinical benefit, widespread deployment
ICH detection (Aidoc, Viz.ai): High sensitivity, reduces notification time, low downside risk
PE detection: Solid validation, workflow benefits
Moderate Evidence (Deploy with caution, monitor closely):
Sepsis prediction: Mixed evidence, high false positives, local validation essential
Deterioration prediction: Variable results, implementation-dependent
Chest pain risk stratification: Promising but needs more validation
Weak Evidence (Pilot only, research stage):
Automated triage: Insufficient validation for autonomous use
Ventilator management: Early stage, more research needed
Most “AI-enhanced” early warning systems: Incremental benefit unclear
Special Considerations:
Pediatric Emergency/Critical Care:
Challenge: Most AI trained on adults Problem: Pediatric physiology, vital sign ranges, disease patterns differ Need: Pediatric-specific AI validation Current state: Limited pediatric AI available
Mass Casualty and Disaster:
Potential: AI-assisted triage, resource allocation Reality: Insufficient validation in disaster scenarios Concern: Undertriage of salvageable patients
Rural/Community EDs:
Challenge: Different patient populations, resources, workflows than academic centers where AI trained Need: External validation in community settings Transfer decisions: AI may help identify patients needing transfer to higher-level care
The Clinical Bottom Line:
Stroke AI has strongest evidence: LVO and ICH detection proven to improve outcomes
Sepsis AI is oversold: Epic model has major limitations, don’t trust blindly
Time savings matter most: AI must be faster or substantially better to justify use
Alert fatigue is real: Optimize thresholds carefully, monitor response rates
Local validation essential: Academic medical center performance ≠ your ED/ICU
False positives are costly: In high-volume EDs, even 1% FPR = dozens of false alerts/day
Clinical judgment irreplaceable: AI assists but doesn’t replace physician assessment
Integration is everything: Standalone systems won’t be used in fast-paced environments
Liability remains with physician: AI doesn’t change your medical-legal responsibility
Continuous monitoring required: Performance drifts, vigilance essential
Communication matters: Alert right person, right time, right information
Evidence hierarchy: Prospective trials > retrospective studies > vendor claims
Future Directions:
Near-term (1-3 years): - More stroke applications (wake-up stroke, hemorrhagic conversion prediction) - Better sepsis models (lower false positives, earlier prediction) - Expanded trauma AI (solid organ injury grading, hemorrhage prediction) - Real-time clinical decision support (integrated into EHR workflows)
Medium-term (3-7 years): - Multimodal AI (vitals + labs + imaging + notes integrated) - Continuous learning systems (improve from local data) - Personalized risk prediction (accounting for individual patient factors) - Closed-loop systems (AI suggests intervention, monitors response)
Long-term (7+ years): - AI co-pilots for emergency/critical care (comprehensive decision support) - Autonomous monitoring systems (with human oversight) - Predictive resource allocation (anticipate surges, optimize staffing)
But always: Human expertise, clinical judgment, and physician responsibility remain central.
Next Chapter: We’ll explore AI in Internal Medicine and Hospital Medicine, where longitudinal data and chronic disease management create different opportunities and challenges.
Professional Society Guidelines on AI in Emergency and Critical Care
The American College of Emergency Physicians has developed AI resources through its Research Committee’s Artificial Intelligence Subcommittee:
JACEP Open Primer (2025): “Artificial Intelligence in Emergency Medicine: A Primer for the Nonexpert” provides foundational guidance on AI applications in emergency medicine, including:
- Triage system enhancement
- Disease-specific risk prediction
- Staffing needs estimation
- Patient decompensation forecasting
- Imaging interpretation assistance
AIIPEM Program: The Artificial Intelligence to Improve Performance in Emergency Medicine (AIIPEM) program focuses on optimizing the ED intake process through AI applications.
Key Guidance: ACEP emphasizes that AI integration should enhance emergency physician decision-making without replacing clinical judgment, particularly for time-critical conditions where AI false negatives could be catastrophic.
Society of Critical Care Medicine (SCCM)
SCCM has published educational content exploring AI applications in the ICU:
Current Applications Under Investigation:
- Replacement of traditional monitoring systems with multidimensional pattern recognition
- Enhanced clinical risk assessment tools
- Efficient extraction and interpretation of clinical information
- Predictive modeling for patient deterioration
Design Considerations (2024 ICU Design Guidelines):
The SCCM 2024 Guidelines on Adult ICU Design note that prior guidelines (1995-2012) did not envision: - Remote manipulation of ventilator settings - Remote infusion pump adjustments - AI-integrated monitoring systems
These capabilities are now being incorporated into modern ICU design standards.
European Society of Intensive Care Medicine (ESICM)
ESICM provides international perspective on critical care AI implementation, emphasizing:
- Multicenter validation requirements for predictive algorithms
- Integration with existing clinical decision support systems
- Attention to alert fatigue in high-acuity environments
- Cross-border data governance considerations
Surviving Sepsis Campaign
The Surviving Sepsis Campaign guidelines (endorsed by SCCM and ESICM) have implications for AI:
- Sepsis screening algorithms should be validated locally
- Electronic alert systems must balance sensitivity with specificity
- AI predictions should support, not replace, clinical assessment of sepsis
- Time-to-treatment metrics remain the focus, regardless of detection method
Check Your Understanding
Scenario 1: Sepsis AI Alert Fatigue and Missed Diagnosis
You’re an emergency medicine physician working a busy night shift at a 400-bed academic medical center. Your ED recently implemented the Epic Sepsis Model as part of your EHR.
Your experience with the system: - First month: 3-5 sepsis alerts per shift - Majority are false positives (patients not septic) - Common false triggers: Febrile URI, dehydration, COPD exacerbation - You and your colleagues increasingly ignore the alerts
11:45 PM - New patient arrival: - Patient: 67-year-old woman - Chief complaint: “Weakness and confusion × 2 days” - Triage vitals: BP 108/62, HR 98, RR 18, T 37.8°C (100.0°F), SpO2 96% on RA - Triage note: Alert and oriented × 3, no acute distress
12:10 AM - Epic Sepsis Alert fires: “HIGH RISK for sepsis. Consider sepsis workup and antibiotics.”
Your assessment: Patient looks okay, vitals not alarming. Probably another false positive. You’re managing 2 critical patients (STEMI, respiratory failure). You acknowledge the alert and plan to see patient when freed up.
1:30 AM - Nurse pages you: “Room 14 (the weakness patient) now BP 88/50, HR 115, more confused. Family says she’s not acting right.”
You reassess immediately: - Vitals: BP 85/48, HR 118, RR 24, T 38.2°C (100.8°F) - Exam: Confused, lethargic, poor skin turgor, no obvious source - Family: “She had UTI symptoms last week, didn’t want to see doctor”
Your workup: - Labs: WBC 18.5, lactate 4.2 mmol/L, Cr 2.1 (baseline 0.9) - Urinalysis: 100+ WBC, nitrite positive, bacteria - Diagnosis: Urosepsis with septic shock
You initiate sepsis bundle: - Fluid resuscitation - Blood cultures - Broad-spectrum antibiotics (2 hours after ED arrival)
ICU course: - Required vasopressors × 48 hours - AKI requiring temporary dialysis - Prolonged ICU stay (7 days) - Eventual recovery but new baseline kidney dysfunction
M&M Review findings: - Epic Sepsis Alert fired at 12:10 AM based on: - Elevated heart rate (98 vs. patient’s baseline 70s) - Subtle temp elevation - Elevated lactate (2.1 mmol/L on triage labs, before you saw patient) - Confusion (documented by triage nurse)
Committee question: “The AI correctly identified sepsis 1 hour 20 minutes before you initiated treatment. Why was the alert ignored?”
Question 1: What factors led to the delayed sepsis recognition despite AI alert?
Root causes of AI alert being ignored:
1. Alert fatigue from high false positive rate - Epic Sepsis Model has documented 33% sensitivity (missed 67% of cases), 12% PPV (88% false positives) - Clinician experience: 3-5 alerts/shift, most not sepsis - Pattern learned: “Sepsis alerts are usually wrong” - Cognitive bias: Automation complacency → ignore frequent incorrect alerts
2. Non-specific presentation - Triage vitals borderline (not meeting classic SIRS criteria) - Initial lactate 2.1 (elevated but not dramatically) - Temperature initially normal-range (100.0°F) - Patient looked “okay” on initial assessment
3. Competing priorities - 2 critical patients requiring immediate attention (STEMI, respiratory failure) - Sepsis alert for stable-appearing patient deprioritized - Time pressure in busy ED
4. Alert design issues - Alert did not convey urgency effectively - No clear action pathway (“Consider sepsis workup” too vague) - Alert easily dismissed without forcing reassessment
5. Lack of trust in AI system - Previous false positives eroded confidence - No explanation provided for WHY patient flagged - “Black box” prediction without clinical reasoning
Question 2: Are you liable for the delayed sepsis treatment?
Legal analysis:
Standard of care for sepsis: - Surviving Sepsis Campaign guidelines: Antibiotics within 1 hour of recognition - CMS SEP-1 core measure: Antibiotics within 3 hours for severe sepsis, 1 hour for septic shock - Emphasis: “Recognition” is key. Clock starts when clinician suspects sepsis
Plaintiff’s argument:
- “The AI system correctly identified sepsis at 12:10 AM”
- “Dr. Smith ignored the AI alert for 1 hour 20 minutes”
- “If antibiotics had been given at 12:10 AM instead of 2:00 AM, patient would not have required dialysis”
- “Hospital implemented AI system but physician failed to act on alerts”
- Damages: AKI requiring dialysis, prolonged ICU stay, permanent kidney dysfunction, pain and suffering
Defense arguments:
1. Standard of care is clinical judgment, not AI compliance: - AI is decision support, not diagnostic certainty - Physician must assess patient, not blindly follow algorithm - Initial presentation did not meet clinical criteria for sepsis (SIRS, qSOFA)
2. Recognition at reassessment (1:30 AM) was appropriate: - Initial vitals borderline, patient stable-appearing - When clinical deterioration occurred (hypotension, worsening mental status), sepsis recognized immediately - Treatment initiated within 30 minutes of deterioration
3. Competing priorities justified: - STEMI and respiratory failure patients were higher acuity - Resource allocation appropriate for ED triage
4. Causation uncertain: - AKI may have been present on arrival (Cr already elevated) - Earlier antibiotics may not have prevented dialysis need - Sepsis progression can be rapid despite treatment
Plaintiff’s rebuttal:
Epic Sepsis Alert is hospital’s chosen standard: - Hospital spent millions implementing system - Training emphasized following AI recommendations - Other EDs using system successfully - Hospital policy may state: “Respond to all sepsis alerts”
Competing priorities don’t excuse delayed assessment: - Could have delegated initial assessment to resident, PA, or advanced practice provider - Could have re-triaged patient higher after alert - 1 hour 20 minutes too long to defer assessment
Likely outcome:
Defensible but risky:
- If hospital policy REQUIRES response to sepsis alerts: Stronger plaintiff case (policy violation)
- If policy states alerts are “advisory only”: Stronger defense, clinical judgment prevails
- Key factor: Was initial assessment reasonable given presentation?
- Vitals not meeting sepsis criteria → defense stronger
- Lactate 2.1 visible on chart → should have prompted earlier assessment
Settlement likely given: - Sympathetic plaintiff (permanent kidney damage) - AI correctly identified sepsis - 1 hour 20 minute delay documented - Adverse outcome
Lessons for risk management: - Acknowledge AND assess all high-risk alerts within defined timeframe - Document rationale if disagreeing with AI (e.g., “Assessed patient, does not meet sepsis criteria, will monitor closely”) - Re-triage patients when AI flags high-risk conditions
Question 3: How should sepsis AI be implemented to prevent this scenario?
Best practices for sepsis AI implementation:
1. Pre-Implementation Validation
LOCAL performance assessment (mandatory):
Run sepsis AI in silent mode for 3-6 months on YOUR patient population:
| Metric | Target | Unacceptable |
|---|---|---|
| Sensitivity | >80% | <70% |
| Specificity | >70% | <60% |
| Positive Predictive Value | >30% | <20% |
| Alert rate | <5 per shift | >10 per shift |
If performance unacceptable: Do not deploy OR adjust thresholds
2. Alert Design to Reduce Fatigue
Tiered alert system:
HIGH PRIORITY (immediate assessment required): - Septic shock criteria (SBP <90 + 2 SIRS criteria + suspected infection) - Lactate >4 mmol/L - Alert: Page physician immediately, cannot dismiss without assessment
MODERATE PRIORITY (assess within 30-60 minutes): - 2 SIRS criteria + lactate 2-4 - Suspected infection + organ dysfunction - Alert: Task in EHR, reminder if not addressed
LOW PRIORITY (monitor, no immediate action): - 1 SIRS criterion - Borderline labs - Alert: Passive flag in chart, no interruption
3. Actionable Guidance (Not Just Warning)
Poor alert: “HIGH RISK for sepsis. Consider sepsis workup.”
Better alert:
SEPSIS ALERT - Patient meets predictive criteria
Risk Score: 78% probability of sepsis
Key factors: Lactate 2.1, HR 98 (baseline 70), confusion, suspected UTI
RECOMMENDED ACTIONS:
☐ Reassess patient within 30 minutes
☐ Order sepsis labs if not done: CBC, CMP, lactate, blood cultures
☐ Consider empiric antibiotics if sepsis confirmed
☐ Acknowledge alert and document assessment4. Clinical Decision Support Integration
Order set auto-population: - If sepsis alert fires, pre-populate sepsis workup orders (pending physician review) - One-click order placement (don’t make physician manually enter 10 orders)
Documentation template: - Auto-generated sepsis assessment template in chart - Forces structured evaluation
5. Feedback Loop for Learning
Alert outcome tracking:
Every sepsis alert should be reviewed: - Was patient septic? (gold standard: physician diagnosis + antibiotics given) - If yes, was treatment timely? - If no, why false positive?
Share performance data with clinicians: - “Last month: 45 sepsis alerts, 18 true positives (PPV 40%)” - “Top false positive triggers: COPD exacerbation, dehydration” - Goal: Help clinicians calibrate trust in system
6. Threshold Optimization
Adjustable sensitivity:
Different EDs may prefer different operating points:
High-volume academic ED: Lower sensitivity (fewer alerts), higher PPV to reduce fatigue Community ED with limited backup: Higher sensitivity (catch more cases), accept lower PPV
Allow customization based on local performance and preferences
7. Escalation Pathway for Ignored Alerts
If alert not acknowledged within 1 hour: - Escalate to charge nurse - Charge nurse assesses patient or ensures physician has seen - Prevents alerts from being lost
8. User Training (Essential)
All clinicians must understand: - How sepsis AI works (what inputs, what it’s predicting) - What to do when alert fires (assessment, workup, documentation) - AI is adjunct, not diagnostic truth - Clinical judgment can override AI (with documentation) - PPV expectations (e.g., “30% of alerts will be true sepsis”)
Key message: “AI helps you NOT MISS sepsis, but YOU decide if patient is septic”
9. Audit and Accountability
Monthly review: - Sepsis cases missed by AI (false negatives) → Why? - Sepsis alerts ignored that were true sepsis → Why? - Trends in alert response rates
Individual feedback: - If physician repeatedly ignores alerts without documentation → coaching - If physician has better sepsis recognition than AI → learn from their practice
10. Vendor Accountability Questions
Before purchasing sepsis AI:
MUST ANSWER: 1. “What is PPV at 10% sepsis prevalence in ED population?” (not just AUC) 2. “Provide data from 3+ external validation sites (not just your development site)” 3. “What is alert rate per 100 ED patients?” 4. “How often do clinicians dismiss alerts at your deployment sites?” 5. “Provide prospective trial data showing improved outcomes (not just retrospective prediction)” 6. “What happens when your AI misses sepsis? Will you share liability?”
RED FLAGS: - Vendor cannot provide external validation data - Only reports AUC, not PPV/alert rate - Claims “90%+ accuracy” without defining what that means - Resists local validation period - No mechanism for threshold adjustment
Lesson: Sepsis AI with high false positive rates creates alert fatigue, leading to ignored alerts and missed diagnoses. Implementation must include local validation, tiered alerts, actionable guidance, threshold optimization, and continuous monitoring of both AI performance and clinician response rates. Documentation of alert acknowledgment and clinical reasoning is essential for liability protection.
Scenario 2: Stroke AI False Positive and Unnecessary Thrombectomy
You’re an emergency medicine physician at a comprehensive stroke center. Your hospital uses Viz.ai for automated large vessel occlusion (LVO) detection on CT angiography.
System track record: - Deployed 18 months ago - Generally excellent performance - Reduced door-to-groin time by 40 minutes on average - High staff satisfaction
2:30 AM - Patient arrival: - Patient: 58-year-old man - EMS report: Found by wife at 11 PM with slurred speech, right arm weakness - Last known well: 10 PM (4.5 hours ago) - NIHSS: 6 (moderate stroke severity)
2:35 AM - Head CT non-contrast: No hemorrhage
2:40 AM - CTA head and neck ordered
2:43 AM - Viz.ai ALERT fires:
LARGE VESSEL OCCLUSION DETECTED
Vessel: Left M1 MCA occlusion
Confidence: HIGH
IMMEDIATE THROMBECTOMY CANDIDATEAlert simultaneously sent to: - You (ED physician) - Stroke neurologist (Dr. Lopez) - Neurointerventional radiologist (Dr. Chen) - OR team
2:45 AM - Stroke team assembles
Dr. Lopez (neurologist) reviews patient: - NIHSS now 5 (mild improvement) - Right arm drift, mild dysarthria - Alert and cooperative
Dr. Lopez: “Viz.ai says M1 occlusion. Let’s get him to angio suite.”
You: “Should we wait for official radiology read?”
Dr. Lopez: “Viz.ai is 95% accurate. We’ve done 30 cases with it, never been wrong. Time is brain. Let’s go.”
Dr. Chen (neuroIR) reviews CTA images on mobile device: “I see the M1 cutoff. Looks like LVO. Let’s take him.”
3:00 AM - Patient to angio suite
3:15 AM - Groin access, catheter advanced
3:25 AM - Dr. Chen performs angiogram:
Finding: Left M1 appears patent on angiogram. No occlusion visible.
Dr. Chen: “This is strange. The CTA definitely showed cutoff, but angiogram shows flow. Maybe it recanalized?”
Dr. Chen performs thrombectomy attempt anyway (already committed, patient under anesthesia):
Result: No clot retrieved. Vessel appears normal.
3:45 AM - Procedure concluded
4:00 AM - Overnight neuroradiologist (Dr. Patel) reads CTA (official report):
IMPRESSION:
1. No large vessel occlusion identified
2. Left M1 segment demonstrates atherosclerotic narrowing but patent
3. Apparent "cutoff" on CTA likely artifact from patient motion + atherosclerotic calcification
4. Small lacunar infarct left corona radiata (chronic, not acute)
CONCLUSION: No acute LVO. CTA findings likely motion artifact.Patient outcome: - Thrombectomy complications: Groin hematoma requiring compression, contrast-induced AKI (Cr 1.1 → 2.4) - NIHSS improved to 2 by morning (likely TIA or minor stroke, not LVO) - Discharged day 3 with residual mild weakness - Follow-up: Angry about “unnecessary procedure,” considering legal action
Question 1: What went wrong in this case?
Root causes of Viz.ai false positive leading to unnecessary thrombectomy:
1. AI misclassification - CTA artifact: Patient motion + atherosclerotic calcification mimicked M1 occlusion - Viz.ai trained on: Large datasets but motion artifacts can fool algorithm - False positive: System flagged stenosis + artifact as complete occlusion
2. Over-reliance on AI without independent verification - “Viz.ai is 95% accurate”: True overall, but 5% error rate means 1/20 cases wrong - No independent radiology confirmation before thrombectomy - Neuroradiologist read would have identified artifact (did identify, but after procedure)
3. Cognitive biases - Automation bias: Trusting AI over clinical judgment - Confirmation bias: Dr. Chen “saw” occlusion on CTA because AI said it was there - Sunk cost fallacy: Once in angio suite, proceeded with thrombectomy despite normal angiogram
4. Time pressure overriding verification - “Time is brain” urgency led to skipping official radiology read - Valid concern for true LVOs, but prevented error detection
5. Lack of protocol for AI-physician discordance - What if angiogram doesn’t match CTA? No clear pathway for this scenario - Should have aborted thrombectomy when angiogram showed patent vessel
Question 2: Who is liable for the unnecessary thrombectomy?
Legal analysis:
Standard of care for LVO stroke: - Mechanical thrombectomy proven for LVO strokes up to 24 hours (DAWN, DEFUSE-3 trials) - CTA is standard imaging for LVO detection - Thrombectomy should be performed rapidly when LVO confirmed
Plaintiff’s argument:
- “Doctors performed invasive procedure based solely on AI, without radiologist confirmation”
- “Angiogram showed NO occlusion, yet they attempted thrombectomy anyway”
- “If they had waited 15 minutes for official read, would have avoided unnecessary procedure”
- “Resulted in groin hematoma, kidney injury, unnecessary anesthesia risk”
- Damages: Procedural complications, AKI, emotional distress, medical bills
Defendants:
- Dr. Lopez (neurologist) - Ordered thrombectomy based on AI
- Dr. Chen (neuroIR) - Performed procedure, continued despite normal angiogram
- You (ED physician) - Raised concern but deferred to specialists
- Hospital - Implemented AI system, protocols
Defense arguments:
1. Standard of care supports rapid intervention: - “Time is brain.” Every minute delay causes more infarction - Waiting for official read would delay treatment - CTA showed apparent occlusion (artifact mimicked occlusion convincingly)
2. AI system generally accurate: - Viz.ai FDA-cleared, widely deployed, strong track record - Prior 30 cases at this hospital were all correct - Reasonable to trust system
3. Angiogram discordance addressed appropriately: - When angiogram showed no occlusion, Dr. Chen did NOT force stent retriever - “Thrombectomy attempt” was diagnostic angiography - Procedure aborted when no clot found
4. Complications minor and resolved: - Groin hematoma treated conservatively - AKI resolved (Cr returned to normal) - No permanent harm
Plaintiff’s rebuttal:
AI is adjunct, not diagnostic gold standard: - Standard of care requires physician interpretation, not blind AI adherence - Neuroradiologist should review CTA before irreversible intervention - 15-minute delay for official read is reasonable
Angiogram showed NO occlusion, yet procedure continued: - “Thrombectomy attempt” on normal vessel is below standard of care - Should have aborted immediately when angiogram normal
Informed consent inadequate: - Patient not told “AI detected occlusion but not confirmed by radiologist” - Patient consent assumed confirmed diagnosis
Likely outcome:
Moderate liability risk:
- For proceeding without radiology confirmation: Defensible if time-critical, but risky
- For continuing procedure after normal angiogram: Harder to defend
- Plaintiff sympathetic: Unnecessary procedure with complications
Settlement possible depending on: - Hospital’s AI protocols (do they require radiology confirmation?) - Severity of AKI and whether permanent - Patient’s residual deficits from stroke itself
Expert testimony critical: Did Dr. Chen breach standard by “attempting thrombectomy” on patent vessel?
Question 3: How should stroke AI be implemented to prevent false positive procedures?
Best practices for LVO detection AI:
1. Understand AI as Triage Tool, Not Diagnostic Certainty
Viz.ai role: - Triage: Prioritize worklist, mobilize team - Notification: Alert stroke team rapidly - Time-saving: Reduce door-to-groin time
Viz.ai is NOT: - Diagnostic confirmation - Replacement for radiologist interpretation - 100% accurate
Performance expectations: - Sensitivity ~90-95% (misses 5-10% of LVOs) - Specificity ~85-90% (5-15% false positives) - At 10% LVO prevalence: PPV ~60-70% (30-40% of positive alerts are false)
Clinical implication: 1 in 3 to 1 in 4 Viz.ai positive alerts may be false positives
2. Verification Protocol Before Thrombectomy
Recommended workflow:
STEP 1: Viz.ai fires LVO alert
↓
STEP 2: Stroke team mobilizes (appropriate, saves time for true LVOs)
↓
STEP 3: While patient moved to angio suite, SIMULTANEOUS:
- Neurologist examines patient
- Neuroradiologist reviews CTA (STAT read, 5-10 minutes)
- Anesthesia preps patient
↓
STEP 4: CONFIRMATION REQUIRED before groin puncture:
☐ Neuroradiologist confirms LVO on CTA (not just AI)
☐ Neurologist confirms clinical syndrome consistent
☐ Time window appropriate (within 24 hours for confirmed LVO)
↓
STEP 5: Proceed with thrombectomyKey principle: Mobilization based on AI, but INTERVENTION based on physician confirmation
3. Neuroradiology Confirmation Protocol
STAT CTA reads for stroke: - Neuroradiologist contacted immediately when Viz.ai alert fires - Target: Official read within 10-15 minutes - Neuroradiologist reviews images BEFORE thrombectomy start - Can be done while patient transported, prepped (minimal delay)
Confirmation checklist:
Neuroradiologist must confirm:
☐ Large vessel occlusion present (not artifact)
☐ Vessel identity correct (M1 vs M2 vs ICA vs basilar)
☐ No contraindications visible (hemorrhage, mass, old infarct)
☐ Collateral flow assessment
☐ Clot burden estimation
Neuroradiologist signs off: "Confirmed LVO, safe to proceed"4. Handling AI-Angiogram Discordance
Protocol for discrepant findings:
If CTA (confirmed by radiologist) shows LVO, but angiogram shows patent vessel:
Possible explanations: 1. Spontaneous recanalization (happens in ~20% of LVOs) 2. CTA artifact (false positive) 3. Technical issue with angiogram
Decision tree:
Angiogram shows NO occlusion:
→ If neurologic improvement (NIHSS decreased): STOP procedure
- Likely spontaneous recanalization or false positive
- No benefit to thrombectomy if vessel already open
→ If neurologic stable/worsening: Repeat angiography, different angles
- May be technical miss
- If still no occlusion visible: STOP procedure
→ NEVER perform thrombectomy on angiographically patent vessel5. Informed Consent Specific to AI-Detected LVO
Consent discussion should include:
“Imaging shows what appears to be a blocked blood vessel in your brain. An AI system detected this and alerted our team. Our radiologist is reviewing the images now to confirm. If confirmed, we recommend a procedure to remove the clot. This can significantly improve outcomes, but carries risks including bleeding, stroke, and groin complications. Do you have questions?”
Key elements: - AI detected, physician confirming - Procedure risks explained - Time-sensitive decision
6. Audit and Feedback
Track all Viz.ai alerts:
| Month | Alerts | Confirmed LVO | False Positives | PPV | Thrombectomies | Clot Retrieved |
|---|---|---|---|---|---|---|
| Jan | 12 | 9 | 3 | 75% | 9 | 8 (89%) |
| Feb | 10 | 7 | 3 | 70% | 7 | 7 (100%) |
Review false positives: - Why did AI misclassify? - Common artifacts causing false positives - Share with team to improve recognition
Review false negatives (missed LVOs): - Were there clinical clues AI missed? - Should have been escalated despite negative AI?
7. Vendor Accountability
Questions for Viz.ai (or any LVO AI vendor):
- “What is PPV at 10% LVO prevalence?” (not just sensitivity)
- “What percentage of your alerts at other sites are false positives?”
- “What are most common causes of false positives?” (motion artifact, atherosclerosis, etc.)
- “Do you recommend radiologist confirmation before thrombectomy, or is AI alone sufficient?”
- “What is your liability if AI false positive leads to unnecessary procedure?”
- “Provide data on thrombectomies performed that retrieved no clot (suggests false positive)”
RED FLAGS: - Vendor claims “no need for radiologist confirmation” - Vendor cannot provide PPV data - Vendor dismisses false positives as “rare” - Vendor resists post-market surveillance audits
8. Team Training
All stroke team members must understand:
Viz.ai is highly sensitive triage tool, not diagnostic certainty 1 in 3 to 1 in 4 alerts may be false positives Radiologist confirmation required before thrombectomy Angiogram overrides CTA if discordant Clinical judgment can override AI (document reasoning)
Scenario-based training: - “Viz.ai says M1 occlusion, radiologist sees artifact. What do you do?” - “Angiogram shows patent M1 despite CTA occlusion. Proceed or stop?”
Lesson: Stroke AI tools like Viz.ai improve outcomes by reducing time-to-treatment, but false positives occur (1 in 3-4 alerts). Standard of care requires neuroradiologist confirmation before irreversible interventions like thrombectomy. When angiogram contradicts CTA, clinical judgment (neurologic improvement vs. worsening) and technical considerations (spontaneous recanalization vs. artifact) should guide decision-making. Never perform thrombectomy on angiographically patent vessel solely because AI flagged it.
Scenario 3: ICU Early Warning System and Code Blue
You’re an intensivist at a 24-bed medical ICU at a large academic hospital. Six months ago, your hospital deployed the Epic Deterioration Index, an AI-based early warning system that predicts patient deterioration, cardiac arrest, and ICU transfer needs.
System description: - Analyzes vital signs, labs, medications, nursing assessments in real-time - Generates risk score 0-100 (higher = greater risk) - Alerts when score crosses thresholds: 50 (moderate), 70 (high), 90 (critical)
Your experience: - 10-15 alerts per shift (24-bed ICU) - Most alerts are patients you’re already managing (already in ICU, on pressors, etc.) - Rarely actionable (patient already receiving maximum care) - You and ICU team have learned to mostly ignore alerts
2:00 PM - You’re managing: - 3 post-op patients on ventilators - 2 septic shock patients on 3 pressors each - 1 ARDS patient on ECMO - Multiple floor patients awaiting ICU transfer (no beds available)
2:15 PM - Epic Deterioration Alert:
PATIENT: Jackson, Robert (ICU Bed 12)
AGE: 72
DETERIORATION INDEX: 78 (HIGH RISK)
PREDICTED RISK: Cardiac arrest within 6 hours
RECOMMENDATION: Assess patient urgentlyYou review chart: - Patient: 72-year-old man, post-op day 3 after colectomy for colon cancer - Current status: Extubated yesterday, doing well, off pressors - Vitals (last 2 hours): BP 118/70, HR 88-95, RR 16-20, SpO2 96-98% on 2L NC - Labs (this morning): WBC 11.5, Hgb 9.2 (stable post-op), Cr 1.1, K 3.8 - Nurse note (1 hour ago): “Patient comfortable, tolerating clear liquids, ambulated to chair”
Your assessment: Looks fine, probably false positive. Patient clearly improving, not deteriorating.
You acknowledge alert, no action taken.
4:45 PM - You’re in family meeting for ECMO patient
4:50 PM - Overhead page: “CODE BLUE, ICU BED 12”
You run to Bed 12:
Finding: Mr. Jackson unresponsive, pulseless
Nurse: “I was checking on him, he looked fine 10 minutes ago. Then alarms went off. V-fib on monitor!”
Code Blue team initiates ACLS: - CPR started - Defibrillation × 2 - Epinephrine, amiodarone given
5:05 PM - ROSC achieved (return of spontaneous circulation)
Post-code workup: - Stat labs: K 6.9 mmol/L, Mg 1.2, pH 7.18, lactate 8.2 - ECG: Peaked T waves (hyperkalemia) - Review of vitals trend: - 2:00 PM: HR 88 - 2:30 PM: HR 92 - 3:00 PM: HR 95 - 3:30 PM: HR 102 - 4:00 PM: HR 108 - 4:30 PM: HR 118 - 4:45 PM: V-fib arrest
Cause identified: Hyperkalemic cardiac arrest
Root cause investigation: - Patient has chronic kidney disease (baseline Cr 1.4) - Post-op, placed on IV fluids containing potassium - Morning labs: K 3.8 (low-normal) - Replacement order: KCl 40 mEq IV × 2 doses (given at 8 AM, 12 PM) - Renal function declined: Post-op AKI (Cr 1.1 → 1.8 by afternoon, not yet resulted in chart) - Potassium accumulated: K 3.8 → 6.9 over 6 hours
Epic Deterioration Index retrospective analysis: - Why did AI flag patient at 2:15 PM? - Subtle upward trend in heart rate (88 → 95) - Decreased urine output (30 mL/hr last 2 hours) - Potassium replacement orders in chart - Post-op patient with CKD - AI predicted deterioration 2.5 hours before cardiac arrest
Patient outcome: - Survived cardiac arrest - Post-arrest care in ICU - Anoxic brain injury (prolonged downtime before code called) - Neurologic prognosis uncertain - Family considering withdrawal of care
M&M Committee Review:
Committee: “The AI correctly predicted cardiac arrest 2.5 hours in advance. Why was the alert ignored? If potassium had been checked at 2:15 PM when alert fired, hyperkalemia would have been detected and treated, preventing the arrest.”
Question 1: What factors led to the ignored early warning alert and subsequent cardiac arrest?
Root causes:
1. Alert fatigue from poor PPV - 10-15 deterioration alerts per shift in 24-bed ICU - Most alerts for patients already critically ill (already maximal care) - Alert system “crying wolf.” Clinicians habituated to ignore
2. Alert timing and context - Alert fired for post-op patient who appeared stable - Recent vitals reassuring (BP 118/70, SpO2 96%) - Nurse assessment 1 hour ago: “comfortable, improving” - Cognitive dissonance: Alert says “high risk,” eyes say “patient looks fine”
3. Lack of actionable guidance - Alert said “Assess patient urgently” but didn’t suggest WHAT to assess - No specific recommendation (e.g., “Check potassium level”) - Unclear what intervention would address “cardiac arrest risk”
4. Competing priorities - ICU at capacity, multiple critical patients requiring attention - Family meeting in progress when alert fired - Stable-appearing patient deprioritized
5. System limitations not well understood - Clinicians didn’t understand WHY patient flagged - “Black box” prediction without explanation - If alert had said “Risk factors: K replacement + declining UOP + CKD → check K level,” might have prompted action
Question 2: Are you liable for failing to act on the AI alert?
Legal analysis:
Standard of care for ICU monitoring: - Intensivists must monitor for patient deterioration - Timely response to changes in clinical status - Electrolyte monitoring for at-risk patients (CKD, K replacement)
Plaintiff’s argument:
- “Hospital deployed AI early warning system to prevent exactly this type of event”
- “AI correctly predicted cardiac arrest 2.5 hours early”
- “Dr. Anderson acknowledged alert but took no action”
- “If potassium level had been checked at 2:15 PM, hyperkalemia would have been identified and treated”
- “Cardiac arrest and brain injury were preventable”
- Damages: Anoxic brain injury, prolonged ICU stay, likely death or severe disability, pain and suffering
Defense arguments:
1. Clinical assessment at 2:15 PM was reasonable: - Patient appeared stable (normal vitals, comfortable, improving post-op course) - No clinical signs of hyperkalemia at that time - Physician assessed risk and determined patient stable
2. AI early warning systems have high false positive rates: - Most alerts do not result in deterioration - Physician must exercise clinical judgment, not blindly follow algorithm - Standard of care is clinical assessment, not AI compliance
3. Hyperkalemia was unpredictable: - Morning K level was low-normal (3.8), appropriately repleted - Renal function decline not yet evident (afternoon Cr not resulted) - Rapid K accumulation unusual
4. Cardiac arrest survival achieved: - Code team responded appropriately - ROSC achieved within 15 minutes - Standard of care met for code response
Plaintiff’s rebuttal:
Hospital chose to deploy this AI system: - Hospital invested in Epic Deterioration Index for early intervention - Training emphasized following AI recommendations - If physician routinely ignores alerts, why have system?
AI identified specific risk 2.5 hours early: - Alert was NOT for already-critical patient (patient was stable post-op) - AI detected subtle pattern (trending HR, UOP decline, K replacement orders) - Reasonable physician would have checked K level in this context
Standard of care for post-op CKD patient receiving K: - Should monitor K levels closely - 6 hours between K repletion and next K check too long for CKD patient
Likely outcome:
High liability risk:
- Strong plaintiff case: Preventable arrest, severe injury, AI correctly predicted
- Sympathetic plaintiff: Brain injury from preventable arrest
- Hospital policy question: Did hospital policy require response to high deterioration alerts?
- If YES → stronger plaintiff case (policy violation)
- If NO → stronger defense (alerts advisory only)
Key expert testimony: Would reasonable intensivist check K level for post-op CKD patient with deterioration alert + K repletion orders? - Likely answer: YES, checking K level is low-cost, high-value, should have been done
Settlement very likely given severity of injury and clear prevention opportunity
Question 3: How should ICU early warning systems be implemented to be useful, not just noisy?
Best practices for ICU deterioration AI:
1. Optimize Thresholds to Reduce Alert Fatigue
Problem: Default Epic thresholds generate too many alerts
Solution: Site-specific threshold tuning
Run AI in silent mode for 3 months, track:
| Threshold | Alerts/Day | True Deteriorations | PPV | Alert Fatigue Risk |
|---|---|---|---|---|
| Score >50 | 45 | 8 | 18% | VERY HIGH |
| Score >70 | 18 | 7 | 39% | HIGH |
| Score >85 | 6 | 5 | 83% | MODERATE |
Choose threshold that balances: - Sensitivity (catch deteriorations) - PPV (avoid alert fatigue)
For ICUs: Higher threshold (80-90) may be appropriate. ICU patients already closely monitored, want HIGH specificity
2. Actionable, Specific Alerts (Not Generic Warnings)
Poor alert:
DETERIORATION INDEX: 78
Cardiac arrest risk high
Assess patient urgentlyBetter alert:
DETERIORATION INDEX: 78
Cardiac arrest risk: 15% within 6 hours
KEY RISK FACTORS:
• Heart rate trending up (88 → 102 over 2 hours)
• Urine output declining (30 mL/hr × 2 hours)
• Potassium replacement orders + CKD history
• Post-op day 3 (risk period)
SUGGESTED ASSESSMENTS:
☐ Check stat basic metabolic panel (K, Cr, Mg)
☐ Review fluid balance and UOP trend
☐ Assess for occult bleeding (post-op)
☐ Consider EKG if electrolyte abnormalitiesKey improvements: - Quantified risk (15% not just “high”) - Explanation (why flagged) - Specific actions (check K level, not just “assess”)
3. Integrate Alerts into Workflow (Not Separate System)
Alert delivery: - In-basket task in EHR (not just pop-up that can be dismissed) - Cannot be cleared without documentation: “Assessed patient, [findings], [plan]” - Escalation: If not addressed in 1 hour, alert charge nurse
Order set integration: - Alert includes one-click order for suggested workup - Example: “Order stat BMP for Deterioration Alert” (pre-populated order)
4. Contextualize Alerts (Filter Out Already-Managed Patients)
Avoid alerting for: - Patients already on maximum ICU care (3 pressors, ECMO, etc.). You already know they’re high-risk - Patients with comfort-measures-only status - Patients actively being managed for deterioration
DO alert for: - Stable-appearing patients with subtle trends - Post-op/post-procedure patients (often lower acuity but can deteriorate suddenly) - Patients on general ICU monitoring (not already high-intensity care)
5. Feedback Loop and Continuous Learning
Track all alerts:
| Patient | Alert Time | Score | Assessed? | Action Taken | Outcome |
|---|---|---|---|---|---|
| Jackson, R | 2:15 PM | 78 | No | None | Arrest 4:50 PM |
| Smith, J | 2:30 PM | 72 | Yes | Checked labs, normal | No deterioration |
| Lee, K | 3:00 PM | 81 | Yes | Transfused, transferred | Stabilized |
Monthly review: - True positives: Alerts that preceded deterioration → Learn what worked - False positives: Alerts that didn’t lead to deterioration → Adjust thresholds - False negatives: Deteriorations not predicted → Improve model
Share with team: - “Last month: 42 alerts, 18 true deteriorations (PPV 43%)” - “Top reasons for true positives: post-op AKI, sepsis, arrhythmia” - Goal: Help clinicians calibrate when to trust vs. question alerts
6. User Training and Expectations
All ICU clinicians must understand:
What deterioration AI predicts (arrest, transfer, mortality) How it works (what inputs, what patterns) What to do when alert fires (specific assessments, not just “look at patient”) Expected PPV (e.g., “40% of alerts will be true deteriorations, 60% false”) Physician judgment overrides AI (but must document rationale)
Key principle: “AI helps you catch SUBTLE trends you might miss, but YOU decide what to do”
7. Protocol for High-Risk Alerts
When Deterioration Index >85:
MANDATORY ACTIONS (within 30 minutes): 1. Bedside assessment by physician or advanced practice provider 2. Vital signs recheck 3. Review I/O, medications, recent labs 4. Stat labs if risk factors suggest (e.g., K replacement + CKD → check K) 5. EKG if cardiac arrest risk 6. Document findings and plan in chart
If alert seems inappropriate: - Document why (e.g., “Patient extubated, ambulating, tolerating diet. Alert appears false positive, will monitor”) - DO NOT simply dismiss without assessment
8. Vendor Accountability
Questions for Epic (or any deterioration AI vendor):
- “What is PPV for cardiac arrest prediction at 1% base rate?” (not just AUC)
- “What alert rate per 100 ICU patient-days do you recommend?”
- “How many sites have reported alert fatigue and stopped using the system?”
- “Provide prospective trial data showing reduced code blues or mortality” (not just retrospective prediction)
- “Can thresholds be customized per institution?”
- “What explanations does system provide for WHY patient flagged?”
RED FLAGS: - Vendor cannot provide site-level PPV data - One-size-fits-all thresholds (no customization) - No prospective outcome trials - Black-box predictions without explanations - High alert rates at deployed sites (>5 per 100 patient-days)
Lesson: ICU early warning AI can detect subtle deterioration patterns, but high false positive rates create alert fatigue leading to ignored alerts and preventable adverse events. Effective implementation requires threshold optimization for local population, actionable and specific alert content with suggested interventions, workflow integration, user training on expected PPV, and protocols for mandatory assessment of high-risk alerts. Explanations of WHY patient flagged are critical for clinical decision-making.