Efficiency of arrhythmia detection by nurses in a coronary care unit using a decentralised monitoring system.

Abstract
The efficiency of nurses in detecting arrhythmias was studied in a coronary care unit (CCU) with 6 beds, equipped with a system for decentralized ECG display and with well-defined routines for arrhythmia documentation. Arrhythmias were classified and reported at 1/2 h intervals. Patients (50) were included. A continuous ECG recording was subsequently analyzed beat by beat and used as reference. The arrhythmias were classified into: benign arrhythmias, warning arrhythmias, malignant arrhythmias and ventricular tachycardia. The detection rate for different arrhythmias varied from 96% for patients with benign arrhythmias occurring in at least three 1 min periods per 1/2 h and 86% for patients with warning arrhythmias to 42% for patients with ventricular tachycardia. Special attention was paid to infrequently occurring arrhythmias which were analyzed in relation to 1/2 h periods. Of warning arrhythmias occurring during only one 1 min period per 1/2 h, 48% were detected, but the detection rate increased to 71% when the same type of arrhythmia occurred during 2 or more 1 min periods per 1/2 h. The mean delay time for detection of warning and malignant arrhythmias was less than 1/2 h. The exact delay time could not be evaluated as the study was based on the 1/2 h reports of the nurses. The fact that the most infrequent warning arrhythmias were detected in 48% during the first possible minute indicates that the true delay time is much shorter. The efficiency of arrhythmia detection is higher in this than in other published studies, both for detection rate and delay time. Possible explanations for this are that well-defined routines for arrhythmia documentation are established and that a system for decentralized ECG display is used.