Semi-automatic external defibrillation and implanted cardiac pacemakers: understanding the interactions during resuscitation

Resuscitation 30 (1995) 127-131

ELSEVIER

Semi-automatic external defibrillation and implanted cardiac pacemakers: understanding the interactions during resuscitation Koenraad G. Monsieurs* a, Viviane M.A. Conraadsb, Marnix P.N. Goethalsb, Joseph P. Snoeckb, Leo L. Bossaerta ‘Department of Intensive Care, University Hospital Antwerp-UIA, Wilnjkstraat IO, 2650 E&gem, Belgium bDepartment of Cardiology, University Hospital Antwerp-UIA, Wilrijkstraat IO, 2650 Edegem, Belgium

Received 11 February 1995;revision received 5 June 1995;accepted 14 June 1995

Many emergencymedical service (EMS) systemsare currently implementing semi-automatic external defibrillation (AED) by emergencymedical technicians. Surprisingly little information is available on the possible interactions between AEDs and implanted cardiac pacemakers.Therefore, at present there are no clear guidelines for the use of AEDs on patients having a cardiac pacemaker. During resuscitation, multiple interactions between pacemakersand AEDs are possible. External defibrillation can cause damage to several functions of the pacemaker.On the other hand, the presenceof pacemaker spikes during cardiac arrest might prohibit recognition of the ventricular fibrillation by the AED. We report on two resuscitation attempts in which the interaction between the ventricular fibrillation, an implanted dual chamber pacemaker and the AED was decisive for the defibrillation success.A clear understanding of thesepossible interactions is necessaryfor the further refining of diagnostic algorithms and clinical strategiesof prehospita1 defibrillation. Keyword:

Cardiac arrest; Pacemaker; Resuscitation; Semi-automatic defibrillation; Ventricular fibrillation

1. Introduction

Although semi-automatic external defibrillation (AED) by emergency medical technicians (EMTs) has been implemented by the EMS-systems in many countries, little or no published material is Abbreviations: EMS, emergency medical service; AED, (semi-) automatic external defibrillator; EMT, emergencymedical technician: AV, atrio-ventricular; bpm, beats per minute; CPR, cardiopulmonary resuscitation; MICU, Mobile Intensive Care Unit; ICD, implantable cardioverter-defibrillator l Corresponding author, Tel.: (32) 3 821 30 00 ext. 3635; Fax (32) 3 828 48 82.

available about the various possible interactions between semi-AEDs and pacemakers. At present there are no clear guidelines for the use of AEDs on patients with an implanted pacemaker. Becausethe sensitivity of the ventricular fibrillation recognizing algorithm in the presence of pacemakeractivity is considered to be low and becauseof the risk of damage to the pacemaker, the application of AEDs on patients having an implanted pacemaker is often discouraged in the manuals of these devices [ 1,2]. This policy would exclude these patients from the most important therapeutic intervention influencing outcome of

0300-9572/95/.$09.050 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0300-9572(95)00879-X

K.G. Monsieurs et al. /Resuscitation

30 (1995) 127-131

cardiac arrest due to ventricular fibrillation: early electrical defibrillation. We report on two casesin which an interaction occurred between an implanted dual chamber pacemaker and a semi-automatic external defibrillator (Heartstart 2000, Laerdal Medical). In the first case, the AED failed to recognize the ventricular fibrillation becauseof the presenceof pacemaker impulses. In the second case the pacemaker activity was inhibited by the ventricular fibrillation and successful defibrillation was achieved. 2. Case reports 2.1. Case no. I

Patient no. 1 was a Caucasianmale, aged 44. Two months before the cardiac arrest, urgent cardiac pacing was necessary because of a third degree atrio-ventricular (AV) block with a junctional escaperhythm of 45 beatsper minute (bpm). A Vitatron Harmony 703/l dual chamber pacemakerprogrammed in DDD-R mode was implanted at the right pectoral region. Both pacemakerleadswere programmed in the unipolar mode (atria1 sensitivity 0.5 mV; ventricular sensitivity 2.0 mV; AV delay 150ms; lower rate limit 70 bpm; upper rate limit 140 bpm). Atria1 and ventricular capture tresholds were relatively high (0.450 and 0.500 ms, respectively at 5.0 V output). The presenting rhythm on arrival of the EMTs is recorded in Fig. la. A ventricular fibrillation of moderate amplitude can be recognized. Pacemaker spikesare superimposedon the ventricular fibrillation waves. Twelve secondslater, the first analysis was made by the EMTs (Fig. 1b). As prescribed by the protocol, no mechanical movement was present and no cardiopulmonary resuscitation (CPR)

Fig. 1. (A) Rhythm on arrival of the first ambulance (EMTs). Ventricular fibrillation and ventricular pacemaker signals. (B) Rhythm at first analysis. Ventricular fibrillation and temporarily inhibition of the pacemaker. (C) Rhythm 9 min after EMT arrival. Ventricular pacemaker spikes at 140 bpm (the upper rate limit). (D) Rhythm on arrival of the MICU. (E) Rhythm 10 min after MICU arrival. Atrial and ventricular pacemaker signals. (F) Rhythm 20 min after MICU arrival. After the first DC shock: asystole and pacemaker activity.

K.G. Monsieurs et al. / Resuscitarion 30 (1995) 127-131

129

was performed during the registration of this rhythm which is undoubtedly ventricular fibrillation. Nevertheless,after 6 s of analysis, a ‘no shock indicated’ messagewas given (not shown). The rhythm on arrival of the second tier, the Mobile Intensive Care Unit (MICU, physician staffed) can be seen in Fig. Id: ventricular fibrillation waves are less clearly distinguishable. This rhythm was interpreted by the attending physician as electromechanical dissociation and standard advanced cardiac life support was provided. A manual shock of 200 J was fired 20 min later (Fig. If). This strip shows regular atria1 and ventricular pacemaker spikes without electrical capture at a frequency of 70 bpm. This pattern remained unchanged until the end of the resuscitation. There was no return of spontaneous circulation. Ex vivo analysis of the pacemaker confirmed that none of the settings had changed before or during the resuscitation attempt. The charge of the battery was good. 2.2. Caseno. 2 Patient no. 2 was a 74year-old Caucasianmale. One year before the cardiac arrest, a Biotronic Physios 01 DDD pacemaker was implanted because of a complete heart block. The atria1 lead was bipolar, but was programmed in a unipolar pacing mode. A unipolar ventricular lead was implanted (atria1 sensitivity 0.5 mV; ventricular sensitivity 2.5 mV; AV delay 180ms; lower rate limit 60 bpm; upper rate limit 120 bpm). Fig. 2a shows the initial rhythm on arrival of the EMTs. A coarse ventricular fibrillation without pacemaker activity can be seen. After successful defibrillation (Fig. 2b), the pacemaker started liring in an AV-sequential mode at 60 bpm. Electrical capture of the myocardium can be seenin Fig. 2c. At this moment, a pulse was clearly present. Qn arrival at the emergency department, the patient was still unconscious, but had spontaneous movements of his limbs and reacted to pain stimuli. Testing of the pacemaker function showed a normal resistance of the atria1 lead (562 n), but a low ventricular lead resistance(385 B). A high ventricular capture treshold (4.5 V at 0.5 ms) was also observed. The battery of the pacemaker was wellfunctioning. After implantation of a new (this time

Fig. 2. (A) Rhythm on arrival of the first ambulance: coarse ventricular fibrillation. (B) Rhythm after a direct current shock: pacemaker activity without capture. (C) Pacemaker activity with electrical capture.

bipolar) ventricular lead, a good resistance and normal treshold values were obtained. The patient fully recovered. 3. Discussion Since the widespread use of permanent cardiac pacemakers, several reports have been published describing the effects of external defibrillation on pacemaker function. Problems with sensing and capture as well as with generator function are known to occur [3-81. Similar problems after discharge of an automatic implantable cardioverterdefibrillator (ICD) have also been reported [9- 111. After the introduction of ICDs, interactions in the opposite direction (from the pacemaker to the ICD) were described. Dual and triple counting of pacemaker impulses and/or QRS complexes can result in inadvertent ICD discharge [ 12- 141.Failure of recognition of ventricular fibrillation, due to interference of pacemakerspikes can induce im-

130

K.G. Monsieurs et al. /Resuscitation

proper inhibition of ICD discharge [9,12,15]. Unipolar pacing generates spikes of larger amplitude which increases the probability of misinterpretation as QRS complexes and hence inhibition of the ICD. Therefore, in patients with concomitant pacemaker and an ICD, reprogramming of the pacing mode to a bipolar manner or replacement of unipolar leads is recommended [ 12,15,16]. In both our patients stimulation occurred in the unipolar mode. In the first case described in this report, the presenting rhythm was ventricular fibrillation (Fig. la,b). Superimposed on this rhythm, single pacemaker impulses were present, most likely originating from the ventricular lead. The absence of atria1 pacemaker impulses indicates the presenceof spontaneousatria1 activity. It is indeed well known that at the onset of ventricular fibrillation, atria1 electrical activity can be conserved [ 171. Hence, it is likely that atria1 activity was sensedby the pacemaker followed after a fixed AV delay by a ventricular impulse without capture. This hypothesis is further supported by the observation that 10 min later the rate has increased to 140 bpm which is the preset upper rate limit of the pacemaker (Fig. lc). At this stage, no exogeneous catecholamineswere administered (EMT intervention). The most likely explanation for this increase in rate is the presenceof a spontaneous supraventricular tachycardia, reflecting increased stressand sympathetic stimulation. The variation in intervals between two spikes (Fig. lb) is thought to be due to sensing by the ventricular lead of large amplitude ventricular fibrillation waves resulting in intermittent inhibition of the ventricular stimulation. This inhibition is important since the AED algorithm will be ‘committed to treat’ if ventricular fibrillation is recognized for at least 6 s. Unfortunately in our patient the longest period of ventricular tibrillation without interfering pacemaker spikes lasted for only 1.7 s (Fig. lb). Another 10 min later (Fig. le) regular pacemaker activity was present showing AVsequential pacing at 70 bpm, however without capture. This is most likely due to the spontaneous evolution of ventricular fibrillation into asystole during prolonged resuscitation.

30 (1995) 127-131

In the second patient coarse ventricular tibrillation waves were recorded on arrival of the EMTs (Fig. 2a). These ventricular fibrillation waves resulted in complete inhibition of all pacemaker activity. This inhibition lasted for more than 6 s, allowing the rhythm analysis of the AED to be completed. After successful defibrillation the pacemaker activity resumed. There was a brief period of non-responsivenessof the myocardium, followed by ventricular capture and clear peripheral pulsations. In both casesa Heartstart 2000 AED (Laerdal, Norway) was used in the defibrillation protocol. The ventricular fibrillation recognition algorithm of this device usesseveral parameters to recognize treatable or non-treatable rhythms. This includes QRS peaks, periodicity, frequency, complex amplitude and width. As many other AEDs, the device can not identify pacemaker signals because their specific characteristics are not defined in the AED-algorithm. There are several disadvantages in using the AED on patients with an implanted pacemaker. The pacemaker signal can be interpreted as a genuine QRS complex and this can result in the nondetection of an underlying ventricular fibrillation. The defibrillation current can also cause damage to several functions of the pacemaker, although newer generations of pacemakersare known to be less sensitive to external electrical shocks. Preferential conduction of the defibrillation current through the pacemaker lead can cause defibrillation failure. However, there are also a number of favourable situations in which the defibrillation can indeed be successful.In somepatients the cardiac arrest is related to primary pacemaker malfunction, with disappearanceof the pacemaker signals. It would be wrong to withhold these patients a defibrillation attempt. Secondly, pacemakers programmed in a bipolar stimulation mode generate smaller electrical impulses at the surface electrocardiogram. Therefore, the probability that these signals will interfere with the AED analysing system is low. Thirdly, moderate and coarse ventricular fibrillation waves can inhibit the pacemaker, thus allowing the analysing system of the AED to detect the ventricular fibrillation.

KG. Monsieurs et al. /Resuscitation

Basedon the current experienceswe propose the following strategy for the use of AEDs on patients with pacemakers. We believe semi-AED should never be withheld in patients with implanted cardiac pacemakers.The possible risk of pacemaker or lead malfunction does not outweigh the chances of successful defibrillation. Obviously, general recommendations for external defibrillation on pacemakers also apply to AEDs. To avoid pacemaker dysfunction, defibrillation electrodes should be placed at a distance of at least 10 cm from the pacemaker. Theoretically, the anteriorposterior electrode position is preferred, but this seemsnot feasible for EMTs becausethe protocol should be kept as simple as possible. After restoration of the spontaneous circulation, pacemakers should always be checked for possible damage. Cardiologists should be aware that unipolar pacemaker signals are likely to inhibit proper semi-automatic defibrillation. Therefore, especially in patients at higher risk for sudden cardiac death, a bipolar pacing mode should be preferred, if possible. Anticipating the possibility of resuscitation is reasonable, since many EMS systemsare now implementing AED at an accelerating rate. Finally, manufacturers should be encouraged to continuously improve the ventricular fibrillation recognition algorithms. The development of pacemakeractivity recognition software can tinally lead to the detection of ventricular fibrillation in the presenceof these confounding signals. Acknowledgments

We thank Mark Berkhof for his technical assistance. References [I] Operating Instructions Heartstart 3000. Laerdal Medical, 1990: 5. [2] Operations Manual model 90510 First Medic 510 Semiautomatic Defibrillator. Spacelabs, Inc., 1992: l-3.

30 (1995) 127-131

131

[31 Zullo M. Function of ventricular pacemakers during re-

suscitation. PACE 1990; 13: 736-744. T, Kupersmith J. Implanted automatic detibrillators: Effect of drugs and pacemakers. PACE 1988; 11: 2250-2262. (51 Levine PA, Barold S, Fletcher RD et al. Adverse acute and chronic effects of electrical defibrillation and cardioversion on implanted unipolar cardiac pacing SYStents. J Am Co11Cardiol 1983;6: 1413-1422. f61 Palac RT, Hwang MH, Klondnycky ML et al. Delayed pulse generator malfunction after D.C. countershock. PACE 1981;4: 163-188. [71 Aylward P, Blood R, Tonkin A. Complications of defibrillation with permanent pacemaker in situ. PACE

141 Singer I, Guamieri

1979; 2: 462-464.

PI Das G, Eaton J. Pacemaker malfunction following transthoracic countershock. PACE 1981;4: 187-190.

191 Calkins H, Brinker J, Veltri EP et al. Clinical interactions

between pacemakers and automatic implantable cardioverterdetibrillators. J Am Co11Cardiol 1990, 16: 666-673.

IlO1 Slepian M, Levine JH, Watkins L Jr., Brinker J,

Guamieri T. Automatic implantable cardioverter defibrillator/ permanent pacemaker interaction: Loss of pacemaker capture following AICD discharge. PACE 1987; lo: 1194-1197. Ahem TS, Nydegger C, McCormick DJ et al. Device interaction-antitachycardia pacemakers and detibrillators for sustainedventricular tachycardia. PACE 1991; 14: 302-307. 1121Cohen AI, Wish MW, Fletcher RD et al. The use and interaction of permanent pacemakers and the automatic implantable cardioverter defibrillator. PACE 1988; 1I: 704-711. 1131 Kim SG, Furman S, Matos JA et al. Automatic implantable cardioverter/detibrillator: Inadvertent discharges during permanent pacemaker magnet tests. PACE 1987; IO: 579-582. 1141 Chapman PD, Troup P. The automatic implantable

cardioverter-defibrillator: Evaluating suspectedinappropriate shocks. J Am Co11Cardiol 1986; 7: 1075-1078. 1151 Kim SG, Furman S, Waspe LE et al. Unipolar pacer artifacts induced failure of an automatic implantable cardioverter/detibrillator to detect ventricular fibrillation. Am J Cardiol 1986; 57: 880-881. (161Roofy R, Lah R, Kouchoukas NT et al. Combined bipolar dual chamber pacing and automatic implantable cardioverter/detibriIlator. J Am Coil Cardiol 1986; 7: 933-937. [I71 Capella G, Brugada P, Wellens HJJ. Ventriculoatrial

conduction and atria1 activation during ventricular flutter and fibrillation. Am J Cardiol 1988;61: 916-918.