Electrocardiographic Artifacts During Shoulder Arthroscopy Using a Pressure-Controlled Irrigation Pump

Electrocardiographic interference which mimics cardiac arrhythmias is reportedly caused by a variety of medical equipment (1–4). Recently, we encountered puzzling baseline electrocardiographic artifacts that simulated atrial flutter or atrial fibrillation during shoulder arthroscopy using a pressure-controlled irrigation pump (Fig. 1). We concluded that a new technique or equipment using a roller pump can create electrocardiographic artifacts that can mimic atrial flutter or atrial fibrillation.Figure 1: A pressure-controlled irrigation pump (The Apex Universal Irrigation System 1996, Linvatec, Largo, FL). The arrow indicates the tubing and the roller pump head.Case Report A 62-yr-old man with a rotator cuff tear of the left shoulder was scheduled to undergo shoulder arthroscopic repair. Physical examination results were unremarkable except for severe pain in the left shoulder. The patient’s blood pressure was 140/80 mm Hg and his heart rate 50 bpm and regular. A preoperative electrocardiogram (ECG) revealed left ventricular hypertrophy and bradycardia. In the operating room, electrodes were carefully attached and covered with sealing tape, with special attention given to achieve good skin-to-electrode contact. The patient was monitored with an ECG, a noninvasive blood pressure device, and a pulse oximeter. General anesthesia was induced with thiopental and maintained with sevoflurane and N2O/O2. After the patient was placed in the right lateral position, shoulder arthroscopy was started by using an automatic fluid pump (Fig. 1) at a pressure of approximately 70 mm Hg with a D-sorbitol solution. The patient remained hemodynamically stable during the first stage, but 45 min after the start of the procedure, the ECG revealed sudden onset of what appeared to be atrial flutter or coarse atrial fibrillation (Fig. 2). This series of waves then changed in shape and frequency, subsequently disappearing spontaneously (Fig. 2), even though the pressure setting of the irrigation pump was maintained at a constant 70 mm Hg. After 190 min, the operation was finished, and the patient’s trachea was extubated uneventfully. No pseudoarrhythmias were observed while he was in the hospital postoperatively.Figure 2: A, Electrocardiogram tracing of lead II at the lateral position before operation. B, Artifacts which simulated atrial flutter or atrial fibrillation occurring during arthroscopy.Because the blood pressure and the plethysmographic signal did not change, we suspected that what appeared to be atrial flutter or atrial fibrillation on the ECG was an artifact related to the infusion device. We subsequently recorded the ECG tracings in another case, both while the speed of the pump was increasing and while the pump was turned off (Fig. 3). This revealed that ECG artifacts did not appear until the pump speed was 250 mL/min and that their shape or frequency depended on the pump speed.Figure 3: Real-time rhythm tracings while the speed of the pump was increasing and while the pump was turned off. A, At a speed of 0 mL/min. B, At a speed of 250 mL/min. C, At a speed of 500 mL/min. D, At a speed of 750 mL/min. The arrow indicates the instant when the pump was turned off.Discussion ECG artifacts have been reported to be caused by a variety of sources, including poor electrode contact (5), an extracorporeal roller pump (1,2), a hemodialysis blood pump (3), an infusion pump (4), and 60-Hz interference (6). If the pseudoarrhythmias in our case had appeared immediately after the pump started and had continued while it was running, we would have quickly realized that these waves were the result of ECG interference. In our case, however, the ECG showed a regular rhythm in the first stage of the arthroscopy and then a sudden onset of what appeared to be pseudoatrial flutter or atrial fibrillation, depending on pump speed (from Fig. 3), even though the irrigation system was kept at a constant pressure setting. The shape or frequency of these waves changed depending on the pump speed, and the artifacts themselves did not appear until the speed had reached 250 mL/min. Although the ECG was equipped with a noise filter, this did not protect it from the effects of electrical interference. These factors made correct diagnosis of the arrhythmias and ECG artifacts difficult. In conclusion, we suggest that a new technique or equipment using a roller pump can be a potential source of electrical interference and ECG pseudoarrhythmias. Changes in blood pressure or plethysmographic signal must be considered to confirm the existence of true arrhythmias.