Introduction
Currently, Point-of-Care (POC) ECG monitoring works either as plot devices or alarms for abnormal cardiac rhythms using predefined normal trigger ranges. Modern devices also incorporate algorithms to analyze arrhythmias improving their specificity. On the other hand, full 12-derivation ECG machines are complex, are not suited to use as simple monitors and are used with strict techniques for formal diagnostics of hearth electric conduction pathologies. The automatic diagnostics are derived from a complete analysis of the 12-dimension data after it is fully and well collected. Both systems do not handle disconnected leads and patient’s motions, being strictly necessary to have a good and stable signal to allow proper diagnosis. These interferences with the data collection frequently originate false alarms increasing both patient and staff’s stress; depending on how it is measured, the rate of false alarms (overall) in ICU is estimated at 65 to 95%1.
Alarm fatigue is a well-known problem that consists of a sensory overload of nurses and clinicians, resulting in desensitization to alarms and missed alarms (the “crying wolf” situation). Patient deaths have been attributed to alarm fatigue2. In 1982 it was recognized the increase in alarms with “no end in sight”; studies have demonstrated that most alarm signals have no clinical relevance and lead to clinical personnel’s delayed response. Ultimately patient deaths were reported related to inappropriate responses to alarms2.
In April of 2013, The Joint Commission3 issued the Sentinel Event Alert4, establishing alarm system safety as a top hospital priority in the National Patient Safety Goal. Nowadays (2021), this subject still on their list, in fourth place of importance5.
In February of 2015, the CinC/Physionet Challenge 2015 was about "Reducing False Arrhythmia Alarms in the ICU6. The introduction article stated that it had been reported that up to 86% resulting of the alarms are false, and this can lead to decreased staff attention and an increase in patients’ delirium7–9.
Due to this matter’s importance, this research aims to identify abnormal hearth electric patterns using streaming data, specifically those who are life-threatening, reducing the false alarms, being a reliable signal for Intensive Care Units to respond quickly to those situations.