Recently, the following message was posted to the GASNET discussion group regarding pacemakers and magnets:

(Do not even think about using email addresses in this page for advertising purposes)

 


>Magnets and pacing isn't simple.

>As I understand it (always a good way to get your misconceptions
>corrected on gas net...)

>Firstly electrical suppression of a modern pacer is almost unheard of
>(and since they have a limited life they are all modern now).

>Secondly it should resolve when the electricity is turned off.

>Thirdly the magnet often puts modern pacers into program mode and
>spurious reprogramming has been reported with magnets and diathermy.
>The exact effect of a magnet is unpredictable. Pacing labs are
>reluctant to let you have a magnet for this very reason.

>Fourthly setting a pacer to asynchronous mode might risk VF (R on T
>pacing), presumably this is why they have a demand pacer in the first
>place.

>The real problem (again very rare) is a myocardial burn at the pacing
>site from induced diathermy currents in the lead(s). This stops the
>pacer working. DC currents in the lead might cause VF, but this seems
>very rare.

>What we do:

>1. Get the pacer tested. Find out what type it is. Ask pacing centre
>about the model if possible.

>2. We have a defib / external pacer unit standing by, and magnet. If
>you haven't got emergency pacing you shouldn’t be doing these patients.

>3. Diathermy should be bipolar or a long way from the pacing circuit.
>Use in short bursts well spaced.

>4. If heart stops (it never has for us so far):
> Stop all electrical interference that can be identified.
> External pacing.
> Magnet if this doesn't work. The only indication for using a magnet
>is that the heart has stopped.
> I suppose isoprenaline (isoproterenol?) if all that fails.

>5. CPR, coroner.


Although this plan is, in general, helpful, some of this information is NOT CORRECT. The original statements are reshown below in italics with comment.

 

>Magnets and pacing isn't simple.

This is true, and misconception and folklore further complicate this subject. Some of it has to do with the fact that more than 26 manufacturers have marketed more than 1440 devices in the US alone. For implantable cardioverter defibrillators, 8 companies have now marketed 122 generators. Further complicating this arena: OLD devices (like the nuclear powered pacemakers from Cordis) still exist in some patients (about 600 are still 'unrecovered').

MAGNETS WERE NEVER INTENDED AS EMERGENCY DEVICES. Instead, reed switches were added to 'sensing' pacers so that the average practitioner could estimate remaining battery life without a device programmer. In some devices, magnet placement initiates a pacing sequence that permits the estimation of pacing threshold without a programmer. Some devices (consider the Guidant/CPI Vigor, Discovery, Pulsar, or Meridian series) can be programmed to completely ignore the magnet. It was a problem with one such Vigor that led me to investigate this whole scene. Finally, some old devices were built with no magnet response.

Most ICDs will disable detection (and therefore therapy) when a magnet is correctly placed. Some devices from Guidant / CPI and Medtronic can be programmed to emit a tone to indicate appropriate magnet recognition. Some ICDs from Guidant/CPI, Angeion, and Ventritex can be programmed to ignore the magnet. In general, antibradycardia pacing is not changed by magnet placement, EXCEPT: Sulzer Intermedics (now Guidant) devices will demonstrate battery life via transient (i.e.; 5-13 beats) pacing rate and/or mode change; devices from ELA Medical will change pace rate (but not mode), and a couple of Telectronics (now St Jude) devices (4202, 4203) will inhibit antibradycardia pacing while a magnet is present.

 

DISCLAIMER: The 'facts' presented below are believed to be true by me. Any evidence (not opinion) to the contrary should be sent to me at mrozner_at_mdanderson.org (change the _at_ to @) so that I can correct my mistakes and provide reference to the appropriate author.

 

>Firstly electrical suppression of a modern pacer is almost unheard of
>(and since they have a limited life they are all modern now).

NOT TRUE. Consider the following, which we captured just a few months ago in a patient with a current Medtronic dual chamber rate responsive device, placed for AV nodal disease, programmed to DDDR and undergoing some plastic surgery:

As can be seen, pacer artifacts (spikes) 1-4 appropriately 'track' the native atrial event. The fifth pacer artifact was not conducted, and NO artifact follows native atrial events at #6 or #7. The pacer folks call this "oversensing"; i.e.; the device mistakenly believed that a native ventricular event followed the native atrial depolarization and therefore did not emit an artifact. Of greater interest is the behavior at 8-10. This apparent switch to DOO mode at rate=58 (the program rate of this device, hysteresis disabled) reflects "noise reversion." Noise reversion mode is the generator's response to confusion, and it VARIES BY DEVICE AND MANUFACTURER. At #11, the electrocautery was stopped and the generator returned to its appropriate behavior. Oversensing is actually quite common and difficult to prevent. Devices with unipolar lead configuration and low sensitivity (i.e. the patient's depolarization are difficult to detect) are more prone to oversensing than bipolar lead configurations and high sensitivity).

 

>Secondly it should resolve when the electricity is turned off.

Presumably, this statement refers to cessation of the cautery electromagnetic interference (EMI). His statement is NOT ALWAYS, but usually true. What can make it untrue? Pacers (and ICDs) must sample their battery periodically to determine if the stored charge is adequate for the presumed challenge. If the sensing activity takes place during an electromagnetic storm, the generator could set its end-of-life (EOL) or elective-replacement-indicator (ERI) flag. When the electricity (cautery) ends, the device will now be programmed to EOL or ERI behavior, which has a high probability of being different from the prior (and expected) program. Also, this EOL/ERI behavior might not be in the patient's best interest (it could be VOO mode). Thus one might think that the device has been 'reprogrammed by bovie'.

Other things are possible as well. The FDA has reports of more than 200 catastrophic failures of generators (pace or ICD) related to electrocautery use over the past 10-12 years. I believe that this statement means that the device could fail. Given that there are 2+ million devices in the US, and given those newer devices (read this as 'capable of rate responsiveness' manufactured after 1993) have been better 'hardened' against EMI, catastrophic failure is an unlikely event. But it is not impossible, and it still takes place (just ask the pacing engineers).

 

>Thirdly the magnet often puts modern pacers into program mode and
>spurious reprogramming has been reported with magnets and diathermy.
>The exact effect of a magnet is unpredictable. Pacing labs are
>reluctant to let you have a magnet for this very reason.

Three statements of variable degrees of correctness. About 60% of programmable pacers need a programming 'wand' with a magnet. About 40% of programmable pacers (like those from Guidant/CPI) are interrogated and reprogrammed by devices with NO magnet in the wand of the programmer. For a device that needs a wand with a magnet, the first statement is true, and spurious reprogramming has been reported. Newer devices (since about 1993, most are capable of rate responsiveness) have complete, two way communication with their programmer using a complicated scheme of checks (looks like computer network protocol) so that each programming command is confirmed and reconfirmed by the device to prevent spurious reprogramming. Of course, pacemaker reset, failure, or change to ERI or EOL behavior will appear to be a 'reprogramming' (see above).

 

>The exact behavior of a magnet is UNpredictable.

NOT TRUE. Every manufacturer KNOWS what to expect with magnet placement, and this behavior is published in physicians' manuals, generator encyclopedias, etc. The variables, which cannot be predicted entirely, are rate (depends upon battery voltage and remaining life), threshold test results (i.e. failure to capture some artifacts owing to threshold margin testing), and programmable 'ignore' feature. ALL of these questions can be answered by a pre-procedure device interrogation, which I believe should become the standard of care for every patient with a device undergoing an elective procedure wherein some untoward event could take place.

 

>Pacing labs are reluctant to let you have a magnet for this very reason.

A better statement is: "If you don't know what to expect, you shouldn't be using a magnet." If part of the anesthetic plan includes magnet application in the event of a problem, it behooves the practitioner to KNOW what to expect. Anything else is just luck, and luck in the operating room is often bad.

 

>Fourthly setting a pacer to asynchronous mode might risk VF (R on T pacing),
>presumably this is why they have a demand pacer in the first place.

This is TRUE without qualification. Pacing manufacturers have spent a lot of time and money developing complicated sensing algorithms to ensure that generator-initiated depolarizations do not compete with native activity.

Although VF is rather unusual (pacing engineers and cardiologists tell me that it will happen in less than 5% of paced patients), many patients will not tolerate (in hemodynamic terms) low rate pacemaking depolarizations with higher rate native contractions (called competition by the pacing folks). Such activity is VERY uncomfortable and patients can tell when competition is occurring (see literature about pacemaker syndrome).

Some pacers go to asynchronous pacing at program rate with placement of a magnet (about 25-30% of implanted devices). For patients whose devices are merely listening (i.e. the patient has a native rate greater than the program rate), these devices then appear to compete for myocardial depolarizations with the native impulse. When I ask for a device to be reprogrammed to asynchronous mode, I administer drugs to ensure that the patient's native depolarization rate remains below the programmed (pacer) rate.

 

>The real problem (again very rare) is a myocardial burn at the pacing
>site from induced diathermy currents in the lead(s). This stops the
>pacer working. DC currents in the lead might cause VF, but this seems
>very rare.

Myocardial burns are VERY RARE, and generally require a direct electrocautery strike at the pacer housing or along a pacer electrode. The real problem is lack of knowledge surrounded by the abundance of myth and folklore.

Other pearls:

1) Only about 60% of all implanted devices have an xray code that identifies the device. Many of the devices with codes are cryptic and can be identified only by experts in cryptology. I recently sent a scan of an xray to a company since I knew the code belonged to them but I could not identify the device. Turns out that this particular xray code had never been registered (a mistake on their part). Most Medtronic and Guidant/CPI devices (about 85% of the implanted marketplace in the US) have xray codes. I wonder how many folks can identify this device:

 

2) Reprogramming to asynchronous mode does not protect the device from the ravages of EMI (like device reset, EOL/ERI conversion, or failure); however, it does prevent non-pacing from oversensing or over-pacing from undersensing. When is reprogramming indicated? Tough to say. Clearly, devices with minute ventilation sensors SHOULD NOT BE connected to any device that might inject electricity on the chest wall (called bioimpedence measurement). Last July (98), Richard Nishman and I alerted the FDA Center for Devices and Radiologic Health about a serious problem with minute-ventilation rate responsive devices. Our work led to an FDA alert (see URL=http://www.fda.gov/cdrh/safety/minutevent.html) and some published stuff (Anesthesia and Analgesia, ASA Newsletter). The list in the FDA alert is not complete, as a few manufacturers have introduced new models.

 

Currently, four pacemaker manufacturers market 22 devices with minute ventilation sensors. They are:

  • Guidant /CPI Pulsar Family 1170, 1171, 1172, 1270, 1272
  • Telectronics/St Jude Meta Series 1202, 1204, 1206, 1230, 1250, 1254, 1256
  • Medtronic Kappa Series Model 400 (but not Kappa 700)
  • ELA Medical
    • Brio Series 212, 220, 222
    • Talent Series 130, 213, 223
    • Opus RM Series 4534
    • Chorus RM Series 7034, 7134

____________________________

The GASNET author has offered the following steps (italics), I have added to them:

 

>1. Get the pacer tested. Find out what type it is. Ask pacing centre
>about the model if possible.

This is not enough. Get the pacemaker implantation record to find out why the thing was implanted in the first place. I have a healthy respect for a patient with no underlying rhythm. Get the device interrogated prior to the procedure, and reprogram devices where appropriate (this is a tough issue, but I reprogram any device in a patient who must have monopolar electrocautery, a patient with a minute ventilation sensor, a chronotropically incompetent patient with expected need for higher cardiac output [like a retroperitoneal exploration, whipple, splenectomy]). Whenever monopolar electrocautery is used, the device should be tested after the procedure is complete. The threshold margin test with magnet placement can be used for this function.

 

>2. We have a defib/external pacer unit standing by, and magnet. If
>you haven't got emergency pacing you shouldn’t be doing these patients.

These statements are self-evident. I would take it one step further: Implantable intrathoracic gadgets are complicated (but generally robust) devices and should be well understood prior to care of a patient possessing one (or two) or two of them like the patient shown below. Magnets rarely work in the setting of a true failure.

 

>3. Diathermy should be bipolar or a long way from the pacing circuit. Use in short bursts well spaced.

Right out of the mouth of the engineers. No spurious use of cautery. Limit the current. Attach the current return pad (often misnamed the 'grounding pad') far away from the pacing circuit, and keep the flow of cautery current perpendicular to the pacing circuit. Do not cross the pacing circuit with electrocautery current. Make sure that the mechanical rhythm of the patient can be monitored (auscultation, palpation of the pulse [imagine actually touching the patient], pulse oximetry plethysmography, arterial invasive line) during those periods when the electrocardiogram monitor will be confused by the electrocautery.

 

>4. If heart stops (it never has for us so far):
>Stop all electrical interference that can be identified.
>External pacing.
>Magnet if this doesn't work. The only indication for using a magnet
>is that the heart has stopped.
>I suppose isoprenaline (isoproterenol?) if all that fails.

This is more difficult. Although true failures are rare, they exist. If the problem is related to oversensing, cessation of the EMI storm and/or a magnet applied to the appropriate device likely will help. A magnet could hurt, though, as one might learn that the device has unusual magnet behavior (like cessation of output in the event of a marginal battery). Chest compressions must be started in the absence of a pulse. I usually also tell the surgeon (as in: "...ah, Houston, we have a problem here"). I believe that epinephrine is a better agent than isoproterenol. Anticholinergics are rarely indicated, and glycopyrrolate is not indicated as an emergency drug. Sometimes, despite the best of preparation, something happens, like this:

This patient was put to sleep for a wide local excision of a melanoma with intraoperative lymphatic mapping. His device was placed for sinus node disease, and his program rate was 60 / min. Shortly after the induction of anesthesia, we turned him left side down. Immediately, his device developed an absolute failure to pace, with a partial failure to sense, which was related to a lead fracture. What should be noted: NO pacer artifact becomes a depolarization. Fortunately, the patient developed a junctional escape rhythm at about 47 / min and remained hemodynamically stable (he was awakened without procedure completion). There was some variability (for example, his shortest R-R interval is #7-8 which corresponds to a rate of 55 / min. Sometimes his device sensed the ventricular depolarization and delayed the atrial artifact an appropriate amount of time (atrial artifact after #s 2,3,5,6,8,9,11). At QRS #s 1,4,7,10,and 12, the atrial timer was not reset by the patient's native ventricular depolarization and appears in timely fashion for a program rate of 60 / min. Magnet placement converted this Medtronic device to DOO at 85/min and confirmed failure to capture:

Re-interrogation in our recovery room (his last interrogation was only two weeks prior) confirmed a problem with his leads.

Providing appropriate circulatory support for the patient should be the primary focus of attention followed closely by determining the reason(s) for the failure.

Additionally, the statement about the only indication for a magnet is incorrect. Consider the patient who is undergoing urgent moonlight AAA repair, has a pacer (like Medtronic) programmed to 75 / min (magnet rate 85 DOO), and no programmer available. Patient might be pacer dependent, so every beat is paced. At unclamping, one might put on a magnet to increase heart rate to 85 and therefore increase cardiac output. Alternatively, one might tap on the device, since they usually have piezo-electric rate responsive devices, and get the heart rate up toward the upper sensor rate.

 

>5. CPR, coroner.

An unhappy outcome that might have been preventable with appropriate prior planning.


What he didn't say: Make sure that the patient monitor will show (and not filter) the pacer artifacts. There is a report (Wallden et al, Anesthesia and Analgesia, 86:1339, 1998) in which the paced rhythm was misinterpreted owing to pacing filter in a Datex monitor (see Rozner and Nishman, Anesthesia and Analgesia, 88:965, 1999).

If I say or show any more, I won't have any new material for my Refresher Course at the ASA meeting in Dallas (1999). We will cover Implantable Cardioverter-Defibrillator myths and folklore there as well. That subject is somewhat easier: the anesthetic plan for those patients includes the disabling (via programming) of tachycardia detection.

 

Marc Rozner
Division of Anesthesiology and Critical Care
The University of Texas M D Anderson Cancer Center

August, 1999