Factors affecting bubble formation during delivery of defibrillator pulses to arrhythmogenic cardiac tissue via a catheter are unknown. We investigated the role of energy, electrode surface area, interelectrode distance, and electrode polarity on bubble formation and on current and voltage waveforms during delivery of damped sinusoidal discharges from a standard defibrillator to anticoagulated bovine blood. Gas composition was studied with mass spectrometry. Defibrillator energy settings were varied between 5 and 360 J. The principal catheter used for study was a Medtronic 6992A lead. Additional electrodes tested included 2, 5, and 10 mm long No. 6F, 7F, and 8F copper electrodes. Interelectrode distances used to assess the effect of anode-cathode spacing were 1, 5, 10, and 20 cm. Bubble volume increased linearly from 0.043 to 0.134 ml per cathodal pulse and from 0.030 to 3.50 ml per anodal pulse as energy settings were increased from 5 to 360 J (r = .99). Typical smooth waveforms for both current and voltage were seen only in the absence of bubbles. The voltage waveform was distorted for each cathodal pulse of 100 J or more and for each anodal pulse of 10 J or more only if bubbles were present. The effect of electrode surface area on bubble formation was tested at a 200 J energy setting and at a 10 cm interelectrode distance with the use of cathodal pulses. Bubble formation varied inversely with electrode surface area (r = .876). Bubble formation, however, varied minimally as interelectrode spacing was changed from 1 to 20 cm. The effect of polarity on bubble formation when the Medtronic 6992A distal electrode and an 8.5 cm disk electrode separated by 10 cm were used was highly significant. For a 200 J pulse, bubble formation with the catheter as anode was 3.30 +/- 0.10 ml and with the catheter as cathode it was 0.070 +/- 0.002 ml (p less than .001). Mass spectrometry of both anodal and cathodal gas samples demonstrated the constituents of the gas bubble to include a variety of gases, which is inconsistent with simple electrolytic production of the bubbles observed. The predominance of nitrogen in either polarity sample suggested that the principal source of the bubble was dissolved air. In summary, bubble formation at an electrode receiving damped sinusoidal outputs from a standard defibrillator does not vary significantly with varying interelectrode distance. However, it is directly proportional to energy and inversely proportional to electrode surface area. Anodal catheter discharges produce considerably more bubbles than do cathodal discharges.(ABSTRACT TRUNCATED AT 400 WORDS)