Simulation of patient-nurse-computer interaction in a cardiovascular intensive care unit



Journal Title

Journal ISSN

Volume Title



An interactive human machine (IBM) system comprised of a patient, a nurse, and a monitoring computer in the post-surgical cardiovascular intensive care unit (ICU) of Methodist Hospital was simulated using the System State Model on the IBM 360 computer. The most critical feature of the IBM system is the occurrence of abnormal heart beats, which are called premature ventricular contractions or PVCs, and their timely detection by the attending nurse. The prompt detection of PVCs is important because PVCs represent a disturbance of the normal heart rhythm and are often a sign of impending fibrillation or heart stoppage. Detection of PVCs necessitates prompt medication by the nurse. PVCs can be detected visually by the nurse from the electrocardiogram (ECG) monitor. The model simulates the heart beat of a patient with PVCs occurring on a random basis throughout based on an empirical statistical distribution. This distribution was determined from an analysis of actual patient data. The model also simulates the detection of these PVCs by the nurse as she proceeds about her various duties and by the monitoring computer. Results are shown comparing the number of PVC alarm conditions the nurse detected with the aid of the ECG alone and with the aid of the monitoring computer. Other results are presented showing the effect of changing the nurse's routine and the computer monitoring system. Conclusions are presented showing the effectiveness of the monitoring computer as an aid to the nurse. Changes to nursing routines and scheduling are suggested as well as modifications to the computer monitoring system. In addition, observations are made about the statistical occurrence of PVCs and their patterns.