Abstract
Objectives. The aim of this study was to find the electromagnetic interference (EMI) thresholds for several commonly used implantable cardioverter-defibrillators (ICD). Design. Seventeen ICDs were exposed to magnetic fields with different intensities produced by the Helmholtz coil system. Sinusoidal, pulse, ramp, and square-waveforms with a frequency range of 2 Hz to 1 kHz were used. Results. ICD malfunctions occurred in 11 of the 17 ICDs tested. The ICD malfunctions that occurred were false detections of ventricular tachycardia (6/17 ICDs) and ventricular fibrillation (3/17 ICDs), false detection of atrial tachycardia (4/6 dual chamber ICDs) and tachycardia sensing occurring during atrial or ventricular refractory periods (1/17 ICD). In most cases, no interference occurred at magnetic field levels below the occupational safety limits of the International Commission on Non-Ionizing Radiation Protection (ICNIRP). Nevertheless, some frequencies using sine, ramp or square waveforms did interfere with certain ICDs at levels below these limits. No EMI occurred with any of the ICDs below the ICNIRP limits for public exposure. Conclusion. Evaluation of EMI should be part of the risk assessment of an employee returning to work after an ICD implantation. The risk assessment should consider magnetic field intensities, frequencies and waveforms.
Introduction
Implantable cardioverter-defibrillators (ICDs) are used to treat life-threatening ventricular arrhythmias. Pharmaceutical treatment has proven ineffective in treating severe arrhythmias, which has led to the increasing use of ICDs. ICD treatment extends life expectancy and improves the patient's quality of life. As the mean age of patients receiving their first ICD is falling, the number of working- aged people with an ICD is growing, and the concerns regarding electromagnetic interference (EMI) with these devices in work environments is becoming more relevant. The risk probability and the real sources of EMI with ICDs are still not fully known. According to the European EMF directive (2004/40/EC), employers are required to consider the safety of workers at particular risk (Citation1). Workers with a pacemaker or an ICD belong to this group.
Previous studies concerning EMI with ICDs have mostly focused on specific devices or situations. Most of the studies published consider the interference of medical applications with ICDs. Magnetic resonance imaging (MRI) has been a major concern, and although the safety of MRI scans of ICD patients has improved greatly, MRI is still considered a potential risk to ICD patients (Citation2). Other medical applications, such as wireless capsule endoscopy, electrocautery, nerve and muscle stimulation, and left ventricular assist devices have also proven to be possible causes of EMI (Citation3–7). Some common daily used devices may also be sources of EMI in patients with ICDs. Examples of these are electronic article surveillance (EAS) systems, which may cause inappropriate ICD shocks (Citation8,Citation9). Mobile phones were previously a concern as regards EMI with pacemakers, but modern ICDs seem to be immune to today's mobile phones (Citation10,Citation11). Other devices, such as airport metal detectors, personal digital assistants, radiofrequency identification readers (RFID), induction ovens, and magnetically levitated trains are also suspected to cause EMI with ICDs (Citation12–16). Studies concerning low frequency electromagnetic field interference are rarer (Citation17,Citation18).
This study is part of a research project that exposed pacemakers and ICDs in vitro to low frequency magnetic fields. The first part of the project focused on EMI with pacemakers (Citation19). In this study, we exposed 17 ICDs to low frequency magnetic fields with different waveforms and intensities using Helmholtz coil systems.
Material and methods
Magnetic field exposure
The magnetic fields were produced by two Helmholtz coils in an electromagnetically shielded laboratory. The test setup is described in a recent paper and a schematic figure of the test setup is presented in () (Citation19). We used four magnetic field waveforms: sine-, pulse-, ramp-, and square-waves. With sine-, ramp-, and square-waves, we used ten different magnetic field frequencies at a range of 2 Hz–1 kHz, and with pulse-waves we used six frequencies between 2 and 100 Hz. Many devices emitting these kinds of magnetic fields can be found in real life situations (Citation15,Citation21,Citation25). Every exposure period lasted 10 seconds and was followed by a break. This was found to be a sufficient period to trigger possible malfunctioning for the ICDs. For sinusoidal fields, the exposure intensities used started from the reference levels for occupational exposure given by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) (Citation20). For non-sinusoidal waveforms the peak limits were derived from the ICNIRP reference levels for occupational exposure to sine-wave exposure. These peak limits were determined using a magnetic field meter (Narda ELT 400, Narda Safety Test Solutions, Pfullingen, Germany) and were the highest exposure levels for non-sinusoidal fields used in this study. At low frequencies, the highest sine-wave field levels were lower due to the limitations of the exposure equipment. presents the highest magnetic field intensities for every waveform and frequency. We lowered the magnetic field intensities until no interference occurred with the ICD tested. The magnetic field intensities were determined in the middle of the Helmholtz coils by calculations using the analytical formula as well as by measurements made with the magnetic field meter. The ICD under the test was positioned in the middle of the coils.
Figure 1. A schematic figure of the test setup used to exposure. A is Ampere. ICD is an implantable cardioverter-defibrillator. MHz is mega Hertz. PC is a personal computer.
Table I Maximum intensities of magnetic fields used (Citation19).
Tested ICDs
We tested 17 explanted or demo device ICDs from three manufacturers: Medtronic (Medtronic Inc., Minneapolis, MN, USA), Boston Scientific (Boston Scientific, Natick, MA, USA), and St. Jude Medical (St. Jude Medical, Sylmar, CA, USA). The ICDs selected for the tests were explanted devices that were obtained from two Finnish university hospitals and demo devices supplied by the manufacturers. All the devices that had enough battery life left were included in the study. The ICDs were of 13 different models (). All the devices were programmed to operate with the highest programmable sensitivities (0.15–0.2 mV). The devices were programmed in DDD (Citation5), DDDR (Citation1), VVI (Citation10), and VVIR (Citation1) modes (Citation21,Citation22). Ventricular tachycardia detection rates were programmed to lowest programmable detection rate (90–130 bpm) and ventricular fibrillation rates as ≥ 182 or 180 bpm. The detection rates for mode switch and atrial tachycardia were programmed to be as low as possible (120–180 bpm). All of the ICDs tested had lower rate programmed as 40 beats per minute.
Table II Manufacturers and models of the ICDs tested.
Measuring procedure
The ICDs tested were immersed in a plastic box phantom (29 ×22.5 ×5.5 cm) filled with approximately two litres of physiological saline solution (0.9%) to simulate the electrophysiological properties of human tissue. The phantom was placed between the two coils in the centre, where the magnetic field is most homogenous. The exposures were conducted in three orthogonal directions (Citation19).
We attached the appropriate electrode leads to the ICDs to be tested. The defibrillation leads were 64 and 65 cm long, and the atrial and ventricular leads were 54, 58, and 59 cm. The loop formed by the leads in the phantom was 35 cm long and the remaining lengths of the leads were looped under the ICD, as often done in real life implantations. The length of the loop and the lead positions inside the phantom were the same as those used in the previous pacemaker study (Citation19). The electrode tips of the leads were arranged so that they did not touch each other.
After the exposure, we interrogated the ICDs and reviewed the ICD reports for possible EMI and malfunctions. We traced all the possible malfunctions back to the specific exposure situations using time stamps. The possible malfunctions were false detection of ventricular tachycardia (VT) and ventricular fibrillation (VF), false detection of atrial tachycardia resulting in mode switch and tachycardia sensing occurring during atrial or ventricular refractory periods resulting in noise reversion. St. Jude Medical devices, ATLAS +DR, EPIC +VR, EPIC VR, and PHOTON µ VR were not able to record and store bradycardia episodes into the device memory.
Results
Eleven of the 17 ICDs tested showed magnetic field interference. All of the registered malfunctions occurred when the magnetic field was perpendicular to the ICD tested. Ten of the ICDs were interfered with sine-wave fields, two with ramp-wave fields and three with square-wave fields. The pulse-wave fields did not interfere with any of the ICDs tested. The ICDs with interference are presented in with waveforms and frequencies of the interfering magnetic fields.
The malfunctions detected were as follows: false VT detections in 6/17 ICDs, false VF detections in 3/17 ICDs, false atrial tachycardia detections resulting in automatic mode switch in 4/6 dual chamber ICDs and tachycardia sensing occurring during atrial or ventricular refractory periods resulting in noise reversion in 1/17 ICDs. The detected malfunctions of the ICDs as well as the waveforms and frequencies of the interfered magnetic fields are presented in . An example of ICDs false detection of VT caused by external magnetic field is shown in ().
Figure 2. An example of an ICD's false detection of ventricular tachycardia. A sine-wave magnetic field (60 Hz, 410 µT) produced an artefact of ventricular tachycardia and made the ICD to switch to noise mode (DDI). The interference lasted 10 seconds which is equal to the time the magnetic field was applied. The ICD in question was St. Jude Medical's PROMOTE RF 3213-36. DDI is the programmed noise mode of the ICD. Hz is Hertz. ICD is an implantable cardioverter-defibrillator. µT is micro Tesla. s is second.
Table III ICDs that experienced interference, the type of interference and the waveform and frequency of the magnetic field used.
All of the malfunctions occurred above the ICNIRP general public exposure reference levels or peak levels derived from them (Citation20). The lowest ICNIRP reference level for occupational exposure to sine waves that did not cause interference with any of the ICDs tested were 23% for the 10 Hz field corresponding to a 570 µT magnetic field flux density. The lowest peak level derived from the ICNIRP reference levels for occupational exposure that did not cause interference with ramp and square wave fields was 33% for the 5 and 10 Hz fields, corresponding to 41 and 45 µT magnetic field flux densities for ramp wave fields and 70 µT flux densities for square wave fields, respectively. The highest magnetic field flux densities that did not cause interference with any of the ICDs tested are presented in .
Table IV The highest magnetic field flux densities and ICNIRP reference levels that did not cause interference with any of the ICDs tested.
Discussion
In this study, we used the well-known ICNIRP reference levels for time varying magnetic fields, given in 1998: 100 µT for the general public and 500 µT for occupational exposure to sinusoidal fields at a frequency of 50 Hz (Citation20). In 2010, ICNIRP published new corresponding guidelines: 200 µT for public and 1000 µT for occupational exposure (Citation23). According to European Norm 50527-1, a magnetic flux density of 100 µT is considered the ‘safety-level’ for pacemakers and ICDs at 50 Hz, and thus we chose the former reference levels to be used in this study (Citation24). As the results indicate, the new reference levels should be used with caution as regards ICDs. This finding is in compliance with the manufacturers’ recommendation which is 100 µT (Citation25).
EMI testing with ICDs has usually focused on radiofrequency electromagnetic fields or on some specific application, such as an induction oven or a metal detector. Low frequency magnetic field studies, excluding power frequency (50/60 Hz), have been rarer. Earlier studies have also often concentrated on sinusoidal waveforms, although many of the fields in industrial workplaces are non-sinusoidal. There is a particular lack of knowledge regarding the EMI of non-sinusoidal low frequency fields with ICDs. The results of this study indicated that in case of non-sinusoidal fields the malfunctions appeared at lower levels compared to sinusoidal fields, which is important to take into account when conducting risk assessments of employees’ ability to return to work after an ICD implantation.
The malfunctions detected during our tests were false VT and VF detections, false detection of atrial tachycardia resulting in automatic mode switch, and tachycardia detection during atrial or ventricular refractory period resulting in noise reversion. In false VT or VF detection, the ICD misinterprets the magnetic field as rapid intrinsic heart beats and activates arrhythmia termination either by anti-tachycardia pacing (ATP) or by a high voltage therapy shock. ATP may generate true arrhythmias, and therapy shocks are usually extremely uncomfortable when received in a conscious state. In addition, inappropriate shock can induce sustained ventricular arrhythmia such as ventricular tachycardia and fibrillation. It is also possible, although we did not test it in this study, that external EMI inhibits the detection and the therapy of true ventricular arrhythmias. This can be life-threatening to the patient. In inappropriate mode switch, the ICD misinterprets the magnetic field as atrial arrhythmias, resulting in asynchronous ventricular or atrial pacing. In noise reversion, the false tachycardia detection during refractory period results also in inappropriate asynchronous pacing. We did not have any CRT-D devices in our study. However, in a CRT-D device, inappropriate ventricular sensing by EMI can impose asynchronous ventricular pacing via sense response algorithm. In addition to discomfort, ICD's false detection can result in asystole in a pacemaker-dependent patient without own intrinsic rhythm.
The malfunctions appeared almost immediately after the exposure started and ended after the exposure stopped. This indicates that even quickly passing an exposure source without staying longer in a magnetic field can pose a danger to a worker with an ICD.
Major differences on the ICDs’ susceptibility against EMI were found between the models of different device manufactures. However, the devices of one manufacturer (Medtronic) were the most immune against non-sinusoidal magnetic fields, while the one of the Boston Scientific's ICDs that experienced EMI was susceptible only to non-sinusoidal fields.
Magnetic field interference with ICDs may be caused by the loop formed by the electrodes of the ICD leads and the voltages induced between them. In an ICD lead, the sensing/pacing electrodes are at the tip of the lead, 2–3 cm from each other, and the high voltage electrodes (defibrillating electrodes) more proximal, 8–18 cm from the tip of the lead. In case of malfunction in ICD's sensing feature, interference is likely to be occurring in the sensing/pacing electrodes.
The malfunctions were reviewed from the ICDs’ memory. Because of the lack of continuous monitoring, possible malfunctions that were not recorded are missing in this study. If the pacing was inhibited because of EMI, it resulted in the loss of pacing information. Only the sensing feature of the bradycardia function could be recorded. We were also unable to investigate any possible interference with real time arrhythmia detection and treatment. These points can be considered study limitations, especially because some of the ICDs tested were not able to record any bradycardia events. Some of the ICD units tested are older models, but continue to be implanted and are used by patients in many countries. Another obvious limitation of our study was that the response of ICDs to ventricular arrhythmia detection could not be tested (because the therapies were suspended). This was necessary because most of the ICDs used were explanted devices, with their batteries close to end of life status.
This kind of in vitro study can only be indicative of real interference between ICDs and external magnetic fields. In order to get sufficient proof of the possible interference, in vivo studies with human volunteers are necessary. Therefore we are planning to proceed to interference studies with human volunteers with a widespread number of ICDs. The results of the in vivo study aim to help conducting a worker's risk assessment after an ICD implantation.
Conclusions
The results of this study show that magnetic field interference with an ICD is possible below the 1998 ICNIRP reference levels for occupational exposure or under the peak values derived from them (Citation20). Hence one needs to be cautious, when applying the new 2010 ICNIRP reference levels to workers with cardiac pacemaker devices (Citation23). However, no interference occurred below the 1998 ICNIRP reference levels for public exposure which indicates proper function of the ICDs in public spaces and in most occupational environments (Citation20). EMI evaluation should be part of the risk assessment of a worker returning to work after an ICD implantation. The assessment should consider magnetic field intensities, frequencies and waveforms, which all influence the occurrence of interference. However, most workplaces have low electromagnetic fields, and the problems of EMI with ICDs only exist in specific occupational environments. The further information on the occurrence of EMI with ICDs in occupational environments is essential as the amount of electrical equipment emitting EMFs and workers with an ICD are growing.
Acknowledgements
The authors wish to thank Medtronic, St. Jude Medical and WL-Medical (representing Boston Scientific) for consultation and equipment resources.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
This work was supported by the Finnish Work Environment Fund [grant number 107236].
References
- Directive 2004/40/EC of the European Parliament and of the Council on the minimum health. and safety requirements regarding exposure of workers to the risks arising from physical agents (electromagnetic fields). Brussels: EC, 2004.
- Roguin A, Schwitter J, Vahlhaus C, Lombardi M, Brugada J, Vardas P, . Magnetic resonance imaging in individuals with cardiovascular implantable electronic devices. Europace. 2008;10:336–46.
- Dubner S, Dubner Y, Rubio H, Goldin E. Electromagnetic interference from wireless video-capsule endoscopy on implantable cardioverter-defibrillators. Pacing Clin Electrophysiol. 2007;30:472–5.
- Cheng A, Nazarian S, Spragg DD, Bilchick K, Tandri H, Mark L, . Effects of surgical and endoscopic electrocautery on modern-day permanent pacemaker and implantable cardioverter-defibrillator systems. Pacing Clin Electrophysiol. 2008;31:344–50.
- Holmgren C, Carlsson T, Mannheimer C, Edvardsson N. Risk of interference from transcutaneous electrical nerve stimulation on the sensing function on implantable defibrillators. Pacing Clin Electrophysiol. 2008;31:151–8.
- Siu CW, Tse HF, Lau CP. Inappropriate implantable cardioverter defibrillator shock from a transcutaneous muscle stimulation device therapy. J Interv Card Electrophysiol. 2005;13:73–5.
- Oswald H, Klein G, Strüber M, Gardiwal A. Implantable defibrillator with left ventricular assist device compatibility. Interact Cardiovasc Thorac Surg. 2009;8:579–80.
- Gimbel JR, Cox JW. Electronic article surveillance systems and interactions with implantable cardiac devices: Risk of adverse interactions in public and commercial spaces. Mayo Clin Proc. 2007;82:318–22.
- Groh WJ, Boschee SA, Engelstein ED, Miles WM, Burton ME, Foster PR, . Interactions between electronic article surveillance systems and implantable cardioverter-defibrillators. Circulation. 1999;100:387–92.
- Bassen HI, Moore HJ, Ruggera PS. Cellular phone interference testing of implantable cardiac defibrillators in vitro. Pacing Clin Electrophysiol1998;21:1709–15.
- Tandogan I, Ozin B, Bozbas H, Turhan S, Ozdemir R, Yetkin E, . Effects of mobile telephones on the function of implantable cardioverter defibrillators. Ann Noninvasive Electrocardiol. 2005;10:409–13.
- Kolb C, Schmieder S, Lehmann G, Zrenner B, Karch MR, Plewan A, . Do airport metal detectors interfere with implantable pacemakers or cardioverter-defibrillators?J Am Coll Cardiol. 2003;41:2054–9.
- Tri JL, Trusty JM, Hayes DL. Potential for personal digital assistant interference with implantable cardiac devices. Mayo Clin Proc. 2004;79:1527–30.
- Seidman SJ, Brockman R, Lewis BM, Guag J, Shein MJ, Clement WJ, . In vitro tests reveal sample radiofrequency identification readers inducing clinically significant electromagnetic interference to implantable pacemakers and implantable cardioverter-defibrillators. Heart Rhythm. 2010; 7:99–107.
- Binggeli C, Rickli H, Ammann P, Brunckhorst C, Hufschmid U, Luechinger R, . Induction ovens and electromagnetic interference: What is the risk for patients with implantable cardioverter defibrillators?J Cardiovasc Electrophysiol. 2005;16:399–401.
- Fukuta M, Mizutani N, Waseda K. Influence of electromagnetic interference on implanted cardiac arrhythmia devices in and around a magnetically levitated linear motor car. J Artif Organs. 2005;8:154–60.
- Souques M, Magne I, Lambrozo J. Implantable cardioverter defibrillator and 50-Hz electric and magnetic fields exposure in the workplace. Int Arch Occup Environ Health. 2011; 84:1–6.
- Gurevitz O, Fogel RI, Herner ME, Sample R, Strickberger AS, Daoud EG, . Patients with an ICD can safely resume work in industrial facilities following simple screening for electromagnetic interference. Pacing Clin Electrophysiol. 2003;26:1675–8.
- Tiikkaja M, Alanko T, Lindholm H, Hietanen M, Hartikainen J, Toivonen L. Experimental study on pacemakers’ malfunction due to exposure to different external magnetic fields. J Interv Card Electrophysiol. 2012;34: 19–27.
- International Commission on Non-Ionizing Radiation Protection. Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). Health Phys. 1998;74:494–522.
- Bernstein AD, Camm AJ, Fletcher RD, Gold RD, Rickards AF, Smyth NP, . The NASPE/BPEG generic pacemaker code for antibradyarrhythmia and adaptive-rate pacing and antitachyarrhythmia devices. Pacing Clin Electrophysiol. 1987;10:794–9.
- Bernstein AD, Daubert JC, Fletcher RD, Hayes DL, Lüderitz B, Reynolds DW, . The revised NASPE/BPEG generic code for antibradycardia, adaptive-rate, and multisite pacing. Pacing Clin Electrophysiol. 2002;25:260–4.
- International Commission on Non-Ionizing Radiation Protection. Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz to 100 kHz). Health Phys. 2010;99:818–36.
- European Committee for Electrotechnical Standardization. Procedure for the assessment of the exposure to electromagnetic fields of workers bearing active implantable medical devices - Part 1: General. 2010 Brussels, Belgium.
- Personal information from the technical services of Medtronic, Boston Scientific, and St. Jude Medical concerning electromagnetic compatibility of implantable cardioverter-defibrillators. 2012.