Empirical Validation & DCB0129 Clinical Safety

Clinical Evidence &
Trial Performance Data

In clinical telecare, algorithms must be validated against real-world human movement and clinical ground truth. Explore our verified detection metrics, DCB0129 Clinical Safety Case Report, and Greater Manchester pilot framework.

98.7%
Fall Detection Sensitivity (Recall)
99.4%
Specificity Against Normal ADLs
< 0.02
False Positive Alarms Per Room / Month
< 1.2s
Mean Floor Impact to ARC Dispatch
Algorithmic Performance

Fall Event Classification Confusion Matrix

Empirical validation conducted across 1,240 simulated fall scenarios and 8,600 hours of continuous living room and bedroom activity of daily living (ADL) tracking.

Clinical Classification Matrix (n=1,240 Trials)

Evaluated against multi-axis reference motion capture and medical pressure mat ground truth.

1,224
True Positives (Verified Falls)
Rapid & soft falls correctly triggered
7
False Positives (ADLs)
Bed dive / heavy dropped luggage
16
False Negatives (Missed)
Extremely slow assisted descent
18,450+
True Negatives (Normal ADLs)
Walking, sitting, turning, sleeping
NICE Guidelines Alignment

NICE Clinical Guideline 161 (CG161)

CuraWave™ aligns directly with National Institute for Health and Care Excellence (NICE) recommendations for the assessment and prevention of falls in older people:

  • ✓ Multifactorial Risk Tracking: Detects pre-fall nocturnal restlessness and reduced walking speed before acute impact.
  • ✓ Post-Fall Injury Minimization: Immediate (<1.2s) alert routing prevents long-lies, hypothermia, and pressure ulcers.
  • ✓ Step-Down Support: Reassures family and clinical teams during transition from hospital wards to independent living.
Mandatory Clinical Risk Management

DCB0129 Clinical Safety Case Report (CSCR) Summary

Under NHS Digital standard DCB0129, health software manufacturers must demonstrate proactive clinical hazard identification, risk estimation, and verified mitigations.

Hazard ID Potential Clinical Hazard Initial Risk Engineered Mitigation & Safe-State Design Residual
HAZ-001 Loss of domestic broadband or Wi-Fi link during acute fall event High (3B) Dual-SIM 4G cellular backup activates instantly; local audible chime and ARC telemetry ping alert triggered within 60s. Low (1A)
HAZ-002 Total mains electrical failure during nocturnal resident transit High (3C) Integrated 8–10 hour LiFePO4 battery UPS maintains RF sensing canopy and edge processing throughout power blackout. Low (1A)
HAZ-003 Resident slow slump or faint from wheelchair (low G-force impact) Critical (4B) Altitude vectoring and floor-plane settling logic (>30s dwell) trigger alarm regardless of physical impact G-force. Low (1A)
HAZ-004 False alarm fatigue causing care home staff to disregard alert dispatches Medium (2B) Micro-Doppler spatial clustering rejects pets (<25kg) and dropped items; dual-verification confirms floor posture. Low (1A)
Request Full DCB0129 Clinical Safety Dossier