Method for identifying and eliminating ventilator false alarms of ventilator based on clinical data analysis

Abstract

Ventilator is a kind of critical medical equipment with the highest clinical risk, and it plays an essential role in intensive care and maintaining patient lives. Identifying and eliminating ventilator false alarms is one of the most critical issues in the clinical treatment process. A considerable number of false-positive alarms may happen because of inaccurate parameter alarm threshold setting and inappropriate alarm rule application. This study proposes a method for identifying and reducing the false alarms of the ventilator based on clinical data analysis. It firstly establishes a real-time monitoring system for the ventilator. A wireless network module can be used to transmit data, including parameter data and alarm data, to the server. Then, the data changing range for one parameter can be calculated and determined. The change range of one parameter can be divided into 10 sub-ranges. The frequency of each parameter monitoring value presented in each sub-range can be calculated. The parameter alarm thresholds can be set according to the frequencies and the value distribution in different sub-ranges. The alarm times for one or more parameters in a specified period can be acquired. The clinical data can be utilized to verify the alarm thresholds. The method has been applied to identify and eliminate false alarms for ventilators in a hospital. The application effect shows that this method can help set the parameter alarm thresholds and identify and eliminate most false alarms.