heart rate or in impulse duration of the spike potentials of each of the standard limb leads. The radiologic position of the electrode indicated that it was not in close proximity to the phrenic nerve. The fact that diaphragmatic pacing was abolished by deep inspiration further suggests that it followed direct contact between the perforated pacing electrode and the diaphragm, and was not due to phrenic nerve stimulation. The important complications of myocardial perforation are failed pacing and hemorrhage with subsequent cardiac tamponade. Although the latter is frequently mentioned in the literature as a possible complication, it rarely It has been recommended that a perforated electrode should be retracted and repositioned, or left in situ and a new one inserted.* Since our patient had paced normally with myocardial penetration probably of more than one year's duration, we elected to leave the electrode where it was. Although this was a departure from conventional management, we made this decision in view of the radiologic stability of the electrode, and because of the only slight rise in the pacing threshold. Because the electrode remained in a radiologically stable position with uninterrupted cardiac pacing for more than one year, we considered further change unlikely. The rise in pacing threshold to 2.9 volts was well within the capabilities of the power unit. Furthermore, the diaphragmatic pacing did not trouble her, and repositioning of this electrode might thus have been simply meddlesome. Indeed, one might argue that after a year a track may well have formed around the perforated electrode and remanipulation in these circumstances would be more likely to cause bleeding than to prevent it. This is supported by the view of Kramer et a12 who prefer to leave a perforated electrode in situ rather than to reposition it. Careful close monitoring of the patient over one month, following discovery of perforation, revealed no abnormality in pacing. Although we only saw her at four to six month intervals at our clinic because she lived 1,000 miles distant, she was under the care of an internist in her home town, who was fully aware of her condition and who examined her regularly. Follow-up of this patient, once the diagnosis was known, was thus frequent, with due attention being given to all possible criteria of electrode movement and pacing failure. Any interruption in cardiac pacing would have been rapidly noted, and have been regarded as an indication to urgently change the pacing electrode. We have thus shown with follow-up of our patient that cardiac pacing had not failed over a period of at least 19 months and that the electrode was unchanged in position during this time. This case of uninterrupted cardiac pacing together with intermittent diaphragmatic pacing following perforation of a pacing catheter is unique in our experience.