Pascal RISK

Transforming malpractice risk management from reactive to proactive: Automated, real-time detection and validation of patient safety events, enabling faster interventions, fewer lawsuits, and dramatically reduced financial exposure.

32d → 18h
Harm Detection
-95%
False Positives
-70%
Claims Reduction
Recognized as Most Innovative Product for Medical Malpractice Market at Lloyd's Lab
Project: Pascal RISK
Role: Principal UX Designer
Timeline: 8 months

The Challenge

Reactive Risk Management

Before Pascal RISK, risk management teams were trapped in a reactive cycle. Critical adverse events often went undetected for weeks, with investigations starting long after patients had left the hospital. This delay increased legal exposure and made proactive mitigation nearly impossible.

  • Delayed Response: Most events discovered after discharge, missing the window for intervention
  • Resource Intensive: Manual review of thousands of reports, most of which were irrelevant
  • Missed Opportunities: Many preventable events weren't caught until after harm occurred
91 days
Median investigation time
96%
Non-risk events reviewed
1000s
Manual reviews per year

Legacy Process Timeline

1
Event Occurs

Adverse event happens

Day 1
32
Med-Mal Claim Received

Notification of lawsuits avg. 32 days post-discharge

Day 32
33
Risk Review

Risk manager reviews claim and assigns to legal team

Day 33
34+
Investigation and Litigation

Opportunity to mitigate harm or legal exposure lost

Day 34+
"We were always fighting fires instead of preventing them. By the time we knew about an issue, the patient was already home." — Risk Management Executive, Client Hospital

Research & Insights

Stakeholder Alignment Challenge

Balancing competing needs across three key stakeholders: Risk managers wanted early detection with high sensitivity, patient safety nurses needed actionable insights without alert fatigue, and malpractice attorneys required defensible documentation. Solution: Configurable sensitivity thresholds with role-based dashboards.

  • Stakeholder Research: 9 risk manager interviews, 6 patient safety nurse sessions, 2 malpractice attorney consultations, process mapping workshops
  • Data Analysis: Thousands of historical events reviewed, 120 candidate triggers identified, 7 critical hand-off points mapped, 3 system integration points
  • Signal Design: Data science workshops, clinical validation sessions, trigger prioritization matrix, false positive reduction strategy
  • User Experience: Journey mapping exercises, interactive prototyping, usability testing sessions
UX Research Journey Map Segment

Design Process

Design Challenge

How do we detect adverse events in real-time without overwhelming risk teams with false alarms? Our approach: Machine learning reduces false positives by 81% while maintaining 93% event capture—prioritizing critical cases for immediate action while filtering out noise.

1

Signal Identification

Collaborated with clinical informaticists to identify the most predictive indicators of adverse events across EMR systems.

  • 120 candidate triggers
  • Clinical workshops
  • 82 validated signals
2

Algorithm Development

Worked with data science team to build machine learning models that reduce false positives while maintaining high sensitivity.

  • Model training & validation
  • False positive reduction
  • Continuous improvement loop
3

Interface Design & UX Decisions

Designed notification hierarchy and dashboards that surface critical events immediately while preventing alert fatigue. Key UX decision: Color-coded severity system (red=critical, yellow=moderate, gray=monitoring) with configurable thresholds per facility.

  • Mobile-first responsive design for on-call risk managers
  • Severity-based color coding with one-tap escalation
  • Interactive timeline showing event progression and interventions
  • Role-based views: Risk managers see legal exposure, nurses see clinical context
4

Pilot & Validation

Conducted phased rollout with continuous monitoring and optimization based on real-world usage patterns.

  • Phased rollout
  • Continuous monitoring
  • Detection time improvement

Solution

Pascal RISK enabled risk management to intervene often, while the patient was still in the hospital, not weeks later. Automated validation and notification reduced manual workload and focused attention on the most critical events. Proactive mitigation steps now routinely prevent lawsuits and reduce financial exposure.

Pascal RISK event dashboard showing real-time adverse event detection and severity-based prioritization
Pascal RISK event investigation interface with timeline, risk assessment, and recommended actions

Real-Time Monitoring Engine

  • • 82 structured data signals
  • • Machine learning validation pipeline
  • • Epic EMR system integration

User Interface

  • • Severity-based prioritization
  • • Interactive timeline visualization
  • • Web dashboard for risk managers

Workflow Automation

  • • Automated event validation
  • • Escalation routing
  • • Documentation generation
  • • Compliance reporting

Impact

9 days

Median Investigation Time

↓90% (from 91 days)

41%

Relevant Risk Events

↑10x improvement

Proactive

Mitigation

Most cases addressed before discharge

174

Interviews in 2024

Key context captured for legal defense

"For the first time, we're preventing lawsuits, not just defending them. Pascal Risk gives us a chance to intervene before harm escalates—and before legal costs spiral." — Director, Risk Services
Industry Recognition 2024

Lloyd's Lab Cohort 12 - Most Innovative Product

Recognized for breakthrough innovation in medical malpractice risk management at the world's leading insurance innovation accelerator

Presented to global insurance leaders at Lloyd's of London, demonstrating how real-time patient harm detection reduces risk exposure and medical malpractice costs for insurers and health systems.