Hemodynamic benefit of maintaining atrioventricular synchrony during cardiac pacing in critically ill patients.

Abstract

To determine the hemodynamic effects of maintaining atrioventricular synchrony during emergency cardiac pacing in critically ill patients. Prospective, within patient double-blind study. ICU or coronary care unit patients in a university hospital. Forty (23 cardiac surgery, ten acute myocardial infarction, and seven general intensive care) seriously ill patients with severe symptomatic bradycardia. Initial randomization of patients to receive either a pacing mode where atrioventricular synchronization was maintained (atrioventricular pacing: atrial demand, atrioventricular sequential, atrioventricular universal) or a mode of pacing where atrioventricular synchrony was not preserved (ventricular demand pacing). The cardiac output increased from a mean of 4.5 +/- 1.7 L/min (95% confidence intervals: 4.0 to 5.0 L/min) during ventricular demand pacing to 5.3 +/- 1.7 L/min (95% confidence intervals: 4.9 to 5.9 L/min) during atrioventricular pacing (p less than .0001) despite trivial decreases in CVP from 14 +/- 4 mm Hg (95% confidence intervals: 13 to 15 mm Hg) to 13 +/- 5 mm Hg (95% confidence intervals: 12 to 15 mm Hg) and pulmonary artery occlusion pressure from 18 +/- 5 mm Hg (95% confidence intervals: 16 to 20 mm Hg) to 17 +/- 5 mm Hg (95% confidence intervals: 15 to 18 mm Hg). At the same time, mean arterial pressure (MAP) increased from 74 +/- 15 mm Hg (95% confidence intervals: 64 to 79 mm Hg) to 83 +/- 15 mm Hg (95% confidence intervals: 80 to 88 mm Hg) and left ventricular stroke work index from 22 +/- 10 g.m/m2 (95% confidence intervals: 19 to 25 g.m/m2) to 30 +/- 11 g.m/m2 (95% confidence intervals: 26 to 33 g.m/m2). There was no significant change in mean pulmonary artery pressure, pulmonary vascular resistance index, or systemic vascular resistance index. When cardiac pacing is required in critically ill patients, maintaining atrioventricular synchrony increases stroke volume, cardiac output, and MAP apparently with minimal effects on preload and afterload.