RESUMO
BACKGROUND: About 200,000 patients suffer from in-hospital cardiac arrest (IHCA) annually. Identification of at-risk patients is key to improving outcomes. The use of continuous ECG monitoring in identifying patients at risk for developing IHCA has not been studied. OBJECTIVE: To describe the profile and timing of ECG changes prior to IHCA. DESIGN: Retrospective, observational. SETTING: Single 520-bed tertiary care hospital. PATIENTS: IHCA in adults between April 2010 and March 2012 with at least 3 hours of continuous telemetry data immediately prior to IHCA. MEASUREMENTS: We evaluated up to 24 hours of telemetry data preceding IHCA for changes in PR, QRS, ST segment, arrhythmias, and QTc in ventricular tachycardia cases. We determined mechanism and likely clinical cause of the arrest by chart and telemetry review. RESULTS: We studied 81 IHCA patients, in whom the mechanism was ventricular tachycardia/fibrillation in 14 (18%), bradyasystolic in 21 (26%), and pulseless electrical activity (PEA) in 46 (56%). Preceding ECG changes were ST segment changes (31% of cases), atrial tachyarrhythmias (21%), bradyarrhythmias (28%), P wave axis change (21%),QRS prolongation (19%), PR prolongation (17%), isorhythmic dissociation (14%), nonsustained ventricular tachycardia (6%), and PR shortening (5%). At least one of these was present in 77% of all cases, and in 89% of IHCA caused by respiratory or multiorgan failure. Bradyarrhythmias were primarily seen with IHCA in the setting of respiratory or multiorgan failure, and PR and QRS prolongation with IHCA and concomitant multiorgan failure. LIMITATIONS: This is a retrospective study with a limited number of cases; each patient serves as their own control, and a separate control population has not yet been studied. CONCLUSIONS: ECG changes are commonly seen preceding IHCA, and have a pathophysiologic basis. Automated detection methods for ECG changes could potentially be used to better identify patients at risk for IHCA.