AI-Driven Prediction Models Accurately Predict Critical Care Patient Deterioration

Image: Novel machine-learning approach provides more time for intervention with fewer disruptions for caregivers (Photo courtesy of 123RF)

An intelligent clinical surveillance platform, powered by AI-driven prediction models for patient deterioration and best practice clinical protocols, provides insights based on real-time patient physiology and allows caregivers the opportunity to offer proactive clinical care.

CLEW Medical, Inc. (Netanya, Israel) has developed the CLEW system, a revolutionary tool designed to alert medical staff to the likelihood of patient deterioration up to eight hours earlier than traditional monitoring systems. This advanced warning system enables timely interventions that can reduce complications and fatalities. The CLEW system is adept at predicting both high and low risks of respiratory failure and hemodynamic instability, which are among the most frequent issues impacting patients in intensive care units (ICUs). By providing an early warning when a patient is at high risk of experiencing one of these critical conditions, the system offers healthcare providers the crucial window they need to act before the patient's condition visibly worsens. Early interventions could mean preventing the need for drastic measures like mechanical ventilation for severe respiratory distress and assisting in managing hospital capacity by identifying potential bottlenecks caused by a sudden decline in the patient’s condition.

A comparative study in ICUs evaluated the effectiveness of the CLEW system against the accuracy and utility of alerts of the most widely used telemedicine and bedside monitoring systems. The findings not only highlighted its superior precision but also revealed that the CLEW system produced significantly fewer alarms—50 times less than its counterparts. In the high-pressure settings of ICUs, where staff are constantly managing critical situations, reducing alarm fatigue and the mental load on healthcare workers is crucial. The reduction in false alarms leads to fewer disruptions, contributing to a quieter, more serene ICU atmosphere. The research indicated that, on average, 98% of alarms from standard bedside monitors (equating to 147 out of 150 per patient per day) were false alarms. With its infrequent yet more accurate alerts, the CLEW system was recognized for its potential to alleviate ICU burnout syndrome by lowering unnecessary stress and tasks for the medical staff.