Dual Clinical Detection (Sepsis & UTI)
Lucton House Room 2A — Predictive AI Monitoring Across the Patient Journey
Lucton House, Birmingham
The Challenge
The resident in Room 2A presented a complex dependency profile: immobile, requiring assistance of two for transfers, and living with advanced dementia. Critically, this resident was completely non-verbal and unable to use a standard call bell.
Detection of clinical decline (Sepsis/UTI) in non-verbal residents historically relies on physical symptoms which often only manifest once the condition has reached a crisis point.
Our Solution
Deployment of Earzz AI-powered acoustic monitoring. By utilising non-intrusive sound recognition, the system monitors the resident by tracking non-verbal distress cues (shouting, restlessness, and movement).
This allows for continuous proactive resident monitoring without the privacy concerns of cameras or the dependency on resident-led alerts.
Behavioural Impact Timeline
Longitudinal analysis of non-verbal distress cues leading to clinical diagnoses
Pre-Hospital Peaks
System identified three distinct rising patterns of restlessness (shouting/movement) leading to Apr 5 diagnosis.
Hospitalisation (Apr 6-22)
Clear absence of data confirms hospitalisation period following Earzz-triggered intervention.
UTI Warning (Apr 26)
Sharp, anomalous spike detected within 72 hours of hospital return, identifying a secondary UTI.
Detailed Clinical Outcomes
1. Proactive Sepsis Detection
The Earzz AI visualised a clear rising pattern of restlessness in Room 2A. This behavioural data empowered the management team to proactively contact the resident's doctor for investigation.
RESULT: Early sepsis diagnosis prevented emergency readmission and potentially saved the resident's life.
2. Secondary UTI Identification
Following return from hospital, Earzz data showed a "notably restless night" on April 26th with a clear spike in vocalisation.
RESULT: This data-driven flag triggered an assessment that confirmed a UTI, allowing for immediate treatment.
Impact Summary
- Paradigm Shift: Moved care from reactive (symptom-based) to predictive (behaviour-based).
- Enhanced Monitoring: Provided continuous oversight during the high-vulnerability period post-hospital discharge.
- Staff Empowerment: Gave staff objective data to justify clinical escalations to physicians.
Clinical Evidence
Management teams reported that the behavioural insights provided by Earzz were the catalyst for clinical investigation. The correlation between shouting events and infection onset was validated by healthcare professionals during the trial.
Key Takeaway
AI monitoring provides continuous clinical oversight. It serves as an early warning system for non-verbal residents, catching clinical deterioration days before physical signs manifest to human eyes.
