Current issues of ACP Journal Club are published in Annals of Internal Medicine


Atrial pacing reduced thromboembolic events in the sick sinus syndrome

ACP J Club. 1995 May-June;122:60. doi:10.7326/ACPJC-1995-122-3-060

Source Citation

Andersen HR, Thuesen L, Bagger JP, Vesterlund T, Thomsen PE. Prospective randomised trial of atrial versus ventricular pacing in sick-sinus syndrome. Lancet. 1994 Dec 3;344: 1523-8.



To compare ventricular pacing with atrial pacing in patients with the sick sinus syndrome.


5-year randomized controlled trial.


University hospital in Denmark.


225 patients (mean age, 76 y; 142 women) referred for treatment of their first pacemaker between May 1988 and December 1991 who had symptomatic bradycardia < 50 beats/min or QRS pauses of > 2 seconds. Exclusion criteria were age < 50 years; grade 1, 2, or 3 atrioventricular block; chronic atrial fibrillation (AF), AF > 50% of the time, AF with RR interval > 3 seconds, and AF with QRS rate < 40 beats/ min; bifascicular bundle-branch block; hypertension; cancer; cerebral disease; stroke within the previous 3 months; cardiac or other surgery planned; and Wenckebach block < 100 beats/min.


Stratified by age group, patients received single-chamber atrial pacing (n = 110) or single-chamber ventricular pacing (n = 115).

Main Outcome Measures

Mortality, AF (diagnosed by electrocardiogram), thromboembolism (persistence of neurologic symptoms > 24 hours, death within 24 hours from an acute cerebrovascular event, or embolectomy or necropsy-verified peripheral embolus), and heart failure (New York Heart Association criteria).

Main Results

During follow-up, thromboembolic events occurred in 6 patients (5%) in the atrial group compared with 20 patients (17%) in the ventricular group {95% CI for the 12% absolute risk reduction, 4% to 20%, P = 0.005; relative risk reduction, 71%; number needed to treat, 8; CI 5 to 25} (numbers calculated from data in article). After 3 months, AF was always more common in the ventricular pacing group, but the difference did not reach statistical significance. The groups did not differ for mortality or heart failure.


Compared with ventricular pacing, single-chamber atrial pacing reduced the number of thromboembolic events in patients with the sick sinus syndrome.

Sources of funding: Danish Heart Foundation and Sygekassernes Helsefond.

For article reprint: Dr. H.R. Andersen, Department of Cardiology, Skejby Sygehus, Aarhus University Hospital, 8200 Aarhus N, Denmark. FAX 45-8969-6002.


The optimal choice of pacer therapy for symptomatic sick sinus syndrome is controversial. Although retrospective studies have suggested that atrial-based pacing is associated with a lower risk for AF and thromboembolism (1), concern exists that nonrandomized studies may have substantial selection bias favoring atrial pacing. This study by Andersen and colleagues is the first randomized trial comparing atrial with ventricular pacing in patients with the sick sinus syndrome, and it provides convincing data that atrial pacing substantially decreases the risk for thromboembolic events.

The benefits of maintaining atrioventricular synchrony are increasingly appreciated. The mechanism of the benefit of atrial compared with ventricular pacing presumably relates to the absence of retrograde ventriculoatrial conduction, which is associated with development of atrial arrhythmias. Additional mechanisms include atrial distention that results from alterations of ventricular hemodynamics caused by ventricular pacing. It is of interest that although progressive left atrial dilation was noted in both study groups, substantially more atrial dilation occurred in the ventricular-paced group, even in patients remaining in sinus rhythm. The atrial contribution to ventricular diastolic filling is especially important in elderly patients with the sick sinus syndrome in whom diastolic function often is abnormal. Although the groups did not differ for heart failure symptoms in this study, the detection of a more subtle difference in heart failure symptoms or exercise tolerance related to diastolic ventricular filling would require a larger sample.

The importance of selecting patients carefully for atrial pacing is emphasized by this study. Of the 1052 consecutive patients referred for pacer therapy for all kinds of bradycardia, 79% were excluded before randomization. In patients with conduction abnormalities and regular atrial activity, a logical hypothesis (as suggested by retrospective studies) is that maintenance of atrioventricular synchrony (through dual-chamber pacing) also will be beneficial.

Catherine M. Otto, MD
University of Washington Seattle, Washington