Smart Alarm Management System for Patient Monitoring in Intensive Care

Abstract

This paper examines the alarm load on nurses in the Intensive Care Unit (ICU), focusing on the issue of alarm fatigue caused by often unnecessary alarms. Existing research indicates that 80% to 99% of ICU alarms are false or clinically insignificant, contributing to stress and reduced productivity among nurses. This study uses alarm data from the Pediatric Intensive Care Unit (PICU) of the Wilhelmina Kinderziekenhuis (WKZ) to visualize and analyze alarm loads, aiming to provide insights into how comparing different filtering methods can reduce the alarm load for nurses. A web application was developed to visualize alarms per bed and to summarize alarm loads per nurse, allowing for comparison of filtering methods. The validation of this application demonstrated that applying duration filters can significantly reduce alarm loads. However, further research is needed to ensure medically relevant alarms are not filtered out. This application serves as a tool to support development of solutions to alarm fatigue, enabling detailed analysis and validation of proposed filtering methods.