Performance of a computerized CCU arrhythmia monitor system

Experience gained during clinical trials at Johns Hopkins University Hospital has led to refinements in a commercial arrhythmia monitoring system. The performance capabilities and limitations are described for normal Coronary Care Unit use. The system evaluation spanned a time period of over 16 months and included 3,176 monitored patient days. The majority of both false and true arrhythmia alarms was found to be produced by two small, statistically independent segments of the patient population.