OBJECTIVE: This study was undertaken to find out the relationship between the change of QRS duration with pacing and LV systolic function in paced patients. BACKGROUND:Biventricular pacing may lead to functional improvement in selected patients with congestive heart failure. This beneficial effect was associated with narrowing of the QRS complex during biventricular pacing which may correct discoordinate contraction caused by intraventricular conduction delay. Based on the assumption that QRS width itself may be closely related with LV systolic function, we evaluate the relationship between QRS duration and left ventricular ejection fraction in paced patients. MATERIALS AND METHODS: The study subjects were pacemaker dependent patients who had been at least over a year after receiving permanent pacemaker. Those patients who had a history, or symptoms and signs suggestive of congestive heart failure, significant valvular heart diseases, diabetes mellitus, or underlying bundle branch block prior to pacemaker implantation were excluded. The total numbers of subjects were 50 and their mean age was 63.9 years. 37 patients received VVI pacemakers and 13 patients dual-chamber pacemakers. Left ventricular ejection fraction was measured by transthoracic echocardiography. For QRS width analysis, we used the 12 surface ECGs which were recorded right before permanent pacemaker implantation and at the time when measuring the echocardiographic left ventricular ejection fraction during follow-up period (mean F-U period; 55.7+/-37.5 months). Digitizer (Summasketch III, Summagraphics) was used for measuring the QRS duration. The average of 3 times-measurements was used as QRS duration. The QRS width variables were QRSd: QRS duration of pre- implantation, pQRSd; QRS duration of post-implantation, maxpQRSD; maximal post-implantation QRS duration, DQRSd; difference of QRS duration between pre- and post-implantation, maxDQRSd; maximum difference of QRS duration between pre- and post-implantation. RESULTS: The QRSd was 88.09.4 msec and didn't show significant correlation with LV EF (R=0.208, P=.147). Whereas, pQRSd prolonged to160.02.3 msec showed significant negative correlation with LV EF (R=0.545, P<0.001). Each of maxpQRSd (184.023.5 msec), DQRSd (72.020.5 msec), and maxDQRSd (107.723.3msec) also showed negative correlation with LV EF respectively (; R=0.516 & P<0.001, R=0.500 & P<0.001, R=0.513 & P<0.001). The similar correlation was also present in subgroup analysis only for the VVI patients. However, the analysis of dual-chamber pacemaker group failed to show such a relationship. CONCLUSION: This study showed the presence of a negative correlation between prolongation of QRS duration and LV EF in paced patients. In a part of paced patients, the QRS widening itself to a certain extent could associate with worsening of LV systolic function which may occur during the clinical course.