Improving Patient Surveillance in Telemetry: Don’t Just Rely on the Monitor
September 16, 2015 | Evaluations & Guidance
Clinicians routinely rely on automated arrhythmia detection algorithms in physiologic monitoring systems to help identify critical and abnormal changes to a patient's ECG. While these algorithms are a valuable tool, clinicians must understand that they cannot detect 100% of arrhythmias, including lethal or potentially lethal arrhythmias—asystole, ventricular fibrillation, and ventricular tachycardia. This can be due to a variety of reasons, which we'll discuss.
ECRI Institute recently published a hazard report to raise awareness about the dangers of overreliance on arrhythmia detection algorithms in telemetry care settings. ECRI Institute is aware that physiologic monitoring telemetry systems have failed to detect lethal arrhythmias, in some cases leading to patient deaths. Similar reports were also found in FDA's MedSun database. While these types of "missed" events are a relatively rare occurrence (compared to the high incidence of false alarms and resulting alarm fatigue), they may be harder to identify and are a serious safety concern that ECRI Institute believes needs to be addressed. Automated arrhythmia detection does not replace the need for close, direct surveillance of the patient. The purpose of this article is to provide guidance to hospitals on improving surveillance of the patient, specifically in their telemetry care areas, to avoid missed arrhythmia events. The sections below describe the challenges with arrhythmia detection algorithms and elaborate on the specific risks associated with a telemetry care setting.
As part of our investigation of this issue, ECRI Institute contacted several patient monitoring manufacturers, and in our discussions, they concurred that no currently available arrhythmia detection algorithm can detect 100% of arrhythmias. Arrhythmias may be missed for many reasons, including:
The fact that algorithms may not be designed to detect specific types of arrhythmias (e.g., some monitors cannot detect some atrial arrhythmias)
Noise in the signal (e.g., due to...