Management of Implanted Cardiac Defibrillators During Eye Surgery

In Response: Dr. Bayes' statement1 that I wrote that implanted cardiac defibrillators (ICDs) should remain active during ophthalmalogic surgery in the presence of bipolar-only electrosurgery2 is not a true representation of my comments. I believe that, under certain conditions, neither ICD inactivation (by programming) nor magnet placement (on the appropriate ICD) is required. These are different statements. In fact, I went on to say: “…consider that magnet placement on the appropriate ICD might provide an extra margin of safety.” So, if an ophthalmalogic surgery facility or anesthesia provider desires this extra step, I do not judge them to be wrong. However, very few physicians (even cardiologists) understand that there might be problems with magnet placement, since Magnet placement on a current ICD (except Guidant) produces no useful feedback to indicate appropriate positioning, thus the ICD might not actually be disabled; ICDs from St. Jude and Guidant can have their magnet mode disabled by programming (in fact, 46,000 devices from Guidant have their magnet mode permanently disabled due to a problem with the magnet switch3); Some ICDs from Guidant might be unknowingly, permanently disabled after magnet application for more than 30 s4; and Some practitioners often substitute intraoperative magnet placement for appropriate perioperative care, which always includes a preoperative evaluation of the ICD regardless of the proposed surgery. A significant number of patients now undergo ICD implantation for prophylaxis against ventricular arrhythmia without ever demonstrating a propensity for ventricular tachycardia or fibrillation. In their statement regarding driving restrictions, the Heart Rhythm Society estimates discharge while driving for these patients will be 0.15% per year, assuming driving 30 min per day.5 If we extrapolate to one 30-min surgery, then the risk is now 1 per 243,000 cases. These statements do not apply to patients receiving ICDs for a history of ventricular tachycardia or ventricular fibrillation, or for any patient who has had therapy within the last 3–8 mo.6,7 I continue to believe that intraoperative disabling of an ICD is not REQUIRED if monopolar electrosurgery is not used, regardless of the type and duration of anesthesia, if all of the following conditions are met: The ICD undergoes an in-office comprehensive evaluation (not a phone check) shortly (i.e., 1–2 weeks) before the procedure; The ICD lead system has no leads on alert8;—for example, the Medtronic Fidelis lead can create spurious signals resulting in a shock without EKG evidence of a problem9; There is no history of any arrhythmia that will predispose to shock (VT, VF, supraventricular tachycardia, or paroxysmal atrial fibrillation). In addition, I have personally observed many patients who received succinylcholine while connected to pacemaker/ICD programmers without any evidence of oversensing (or I would have reported it). Furthermore, I trust that any anesthesia provider will always place nerve stimulator electrodes in such a way that the current will not cross the ICD or chest (although this statement is now suspect10). Finally, Dr. Bayes is correct when he states that current guidelines (2007) from the American College of Cardiology11 and the ASA Practice advisory12 recommend the disabling of antitachycardia therapy (preferably by programming) during a case, and I cannot argue with any provider who carries out this step. Also, an appropriately placed magnet on the appropriate ICD might prevent an ICD discharge (appropriate or inappropriate). But, provided all of the conditions, described above, are met, I believe this step to be unnecessary and potentially adding cost to, and complicating care of, these patients. My current practice reflects these beliefs. Marc Rozner, PhD, MD The University of Texas MD Anderson Cancer Center Houston, TX [email protected]