Journal Logo

Technical Reports

High-Fidelity Simulation of Transcutaneous Cardiac Pacing

Characteristics and Limitations of Available High-Fidelity Simulators, and Description of an Alternative Two-Mannequin Model

Robitaille, Arnaud MD, FRCPC; Perron, Roger RT; Germain, Jean-François RT; Tanoubi, Issam MD, MA(Ed)c; Georgescu, Mihai MD, FRCPC

Author Information
Simulation in Healthcare: The Journal of the Society for Simulation in Healthcare: April 2015 - Volume 10 - Issue 2 - p 122-127
doi: 10.1097/SIH.0000000000000067


Transcutaneous cardiac pacing (TCP) is a potentially lifesaving technique that is part of the recommended treatment algorithm for symptomatic bradycardia according to the advanced cardiac life support (ACLS) guidelines.1 However, most clinicians use TCP uncommonly, and its successful application is not straightforward, especially in the high-pressure situations where it is usually indicated. Simulation could therefore be especially well suited for the training of clinicians, for the maintenance of competency, and for the assessment of their proficiency in the use of TCP. Unfortunately, as we will discuss, even the highest-fidelity mannequins currently available on the market have important shortcomings that limit the potential of TCP simulation, especially if the objective is the assessment of experienced personnel.

In this report, we will first suggest a list of items defining competency in TCP and then infer what should be the key characteristics of TCP simulation. We will then describe the TCP simulation characteristics of both METI’s Human Patient Simulator (HPS, CAE Healthcare, Montreal, Canada) and Laerdal’s SimMan 3G (Laerdal Medical, Stavanger, Norway) and underline their limitations. Finally, we will describe how a novel combination of the 2 mannequins can be used to overcome some of the available mannequins’ shortcomings.


Ahn et al2 have recently published a list of items defining procedural competency in TCP, which they used to assess and give feedback to emergency medicine residents practicing TCP on custom-made task trainers. They based their list on opinions from faculty members, best practice material, peer-reviewed literature, and a leading textbook. We would like to expand on this list and propose a list of criteria defining clinical competency in TCP, which is the ability to establish effective TCP in the clinical context (as opposed to an isolated procedural scenario). Clinical competency in TCP requires the following:

  1. Recognizing when TCP is indicated
  2. Properly applying the multifunction electrode pads and connecting them to a functioning pacing unit
  3. Properly applying the pacemaker’s (PM) electrocardiography (ECG) leads to the patient
  4. Setting the appropriate pacing mode
  5. Setting the appropriate pacing rate
  6. Setting the appropriate pacing output and verifying ventricular capture

Although establishing these criteria is rather uncontroversial, developing an appropriate model for TCP simulation is much less straightforward in part because there is a paucity of literature to guide us. Higher fidelity is certainly not always necessary.3 However, some specific objectives do justify the added complexity and resources necessary for reproducing key aspects of the clinical experience, and assessing the capacity of clinically experienced personnel to establish TCP is such an objective. Indeed, demonstrating true expertise requires going beyond a technique’s basic steps and demonstrating both the capacity to use the technique in the context where it will be required and the ability to avoid pitfalls. A simulator lacking fidelity in these areas may not yield a sufficiently valid assessment.

Taking the first criterion of clinical competency in TCP as an example, although it might be sufficient for a medical student to demonstrate the ability to recognize when TCP is indicated by reading a clinical vignette, seeing a rhythm strip, and obtaining vital signs from a simulated monitor, this is not a valid way of assessing the competence of a practicing anesthesiologist. Indeed, a simulator able to function within a high-fidelity (HF) environment is required to reproduce the high-pressure, high-cognitive task load context similar to the one where TCP is likely to be needed clinically. In addition, using a versatile HF simulator enables the indication for TCP to occur unexpectedly, as it does clinically most of the time, for example, while managing a different, unrelated problem. This provides added validity to the assessment.

Because properly applying the multifunctional electrode pads and connecting them to a functioning pacing unit are also required for clinical competency in TCP, an HF TCP simulator should allow the placement of the pads not only in both appropriate positions (anteroposterior and anterolateral) but also in inappropriate positions. It should also be compatible with a variety of pacing units. Again, this capability is not required for all simulation purposes but could be important for specific scenarios (eg, testing residents with limited resuscitation experience before they begin taking call in the intensive care unit). In addition, when used for assessment, the simulator (and the scenario) must be sufficiently realistic to prevent the participants from using cues specific to the simulation and not available clinically to anticipate the course of events: if specific contacts are required to connect the mannequin to the PM, their presence on the mannequin at the beginning of the scenario, before the bradycardia occurs, might alert some astute participants familiar with the mannequin to the possibility of bradycardia occurring in the future. This could decrease the validity of the assessment.

A high-fidelity TCP simulator should also allow dual monitoring of the cardiac rhythm (simultaneously on the PM unit and on another monitor), as would normally occur in most hospital settings. Indeed, experienced clinicians will expect dual ECG monitoring because it is very easy to connect both the PM’s leads and another monitor’s leads at the same time on a real patient. However, if a simulator has only one set of contacts to connect the ECG leads, it will be necessary to disconnect an “unstable patient” from the vital sign monitor to connect the PM’s leads, which would be highly unrealistic in any hospital setting. Alternatively, the learner would have to use the PM without connecting its ECG leads, which in addition to being unrealistic, will only allow fixed, asynchronous pacing (but not the demand mode).

To allow the fourth and fifth objectives, the TCP simulator should respond appropriately to both pacing modes (fixed and demand) and be able to function across a normal PM’s range of heart rates (normally between 30 and 200 beats per minute).

The sixth objective, setting the appropriate pacing output and verifying ventricular capture, is crucial and, in our experience, often poorly performed in the clinical context. Indeed, in high-pressure high-cognitive task load situations, even experienced clinicians often fail to properly assess ventricular capture and instead rely on inappropriate cues to infer that effective pacing has been established. Some are inappropriately satisfied simply by seeing the patient’s upper body twitch. Others are fooled by a pacing artifact on the ECG tracing produced by the electrical current generated by the PM traveling across the patient’s chest (Fig. 1B, C). This artifact may be especially apparent when the clinical vital signs monitor is used to monitor the ECG instead of the PM’s monitor, as commonly occurs clinically, because more information is generally present in the former. Indeed, the PM’s ECG signal is generally optimized by blanking 40 to 80 milliseconds after the current is delivered, which is generally not the case on the vital signs monitor. Finally, some participants will rely on the characteristic PM spike and/or on the wide QRS complex indicating electrical capture (Fig. 1A) but which are not necessarily associated with effective ventricular contraction and an associated pulse. As discussed earlier, HF TCP simulation should be able to simulate all of these misleading cues so that participants can demonstrate that they will not rely on them in the clinical setting and instead check for a pulse and verify that it corresponds to the set pacing rate. Again, this capacity might not be necessary for all teaching purposes, but it is certainly desirable if one is to assess the expertise of individuals likely to be exposed to TCP and all of its pitfalls while resuscitating a real patient. It is interesting to note that Ahn et al,2 in their study of procedural proficiency in TCP, found that the step most frequently omitted by their learners was checking for a pulse to confirm capture. They hypothesized but could not verify that this was because their simulators lacked tactile pulse feedback. This illustrates how using a simulation model with limited fidelity can hamper one’s ability to distinguish between behaviors that can be attributed to the particularities of the simulation model itself from behaviors that will actually be reproduced in a high-stakes clinical situation, with potential consequences for the patient.

Different tracings obtained during transcutaneous pacing. A, Typical paced ECG (electrical capture). B, Pacing artifact that could be confused with capture (low current). C, Pacing artifact that could be confused with capture (high current). D, Simulated tracing used as both TCP artifact and paced ECG in the TMM. Tracings A, B, and C adapted from Medtronic6 with permission.

When dealing with experienced participants who have often already been exposed to real-life TCP, a simulator that does not reproduce patient twitching or ECG pacing artifact will sometimes elicit an undesired response: some clinicians will start troubleshooting the PM machine and infer appropriately that the absence of patient twitching and/or pacing artifact despite a high-energy setting signifies the absence of electrical output. This constitutes another reason why HF TCP should reproduce the patient twitching and ECG artifact occurring clinically with electrical stimulation.


Table 1 summarizes some of the TCP simulation characteristics of the SimMan 3G and the HPS, and compares them to the characteristics of a proposed alternative, a 2-mannequin model (TMM). Videos further illustrate pacing characteristics of the SimMan 3G (see Video, Supplemental Digital Content 1,, which illustrates pacing with increasing current intensity of the SimMan 3G); of the HPS (see Video, Supplemental Digital Content 2,, which illustrates pacing with increasing current intensity of the HPS); and of the TMM (see Video, Supplemental Digital Content 3,, which illustrates pacing with increasing current intensity of the TMM).

Characteristics of Different HF Transcutaneous Pacing Simulators

Although both mannequins are quite versatile and can be integrated into an HF environment approaching clinical reality, some shortcomings limit their usefulness. Indeed, with both mannequins, special cables, which are different from the multifunction sticky electrode pads used clinically, are generally required to connect the PM to the mannequins. Using these special cables does not allow the evaluation of participants’ knowledge of which material is used clinically for TCP or if proper pad placement will be used because in both cases, the special cables can only be connected to the mannequin in one specific manner. With the HPS, one configuration does allow pacing using the sticky electrode pads. However, this entails using metallic discs on the mannequin’s skin to connect the pads to the mannequin’s pacing circuit.4 These discs are identical to the ones used to create contact between the mannequin’s defibrillation circuit and the defibrillation paddles, which poses a significant risk to the mannequin’s electrical system because participants can thus mistakenly deliver a high-energy shock via the metallic discs to the mannequin’s pacing circuit, which is not designed to withstand it.

Dual ECG monitoring is possible with the SimMan 3G because its custom laptop monitor will continue to display the ECG tracing (taken directly from the physiologic model) despite mannequin disconnection, and the PM’s monitor will display the ECG tracing because the mannequin’s ECG contacts emit it. This is not possible however with the HPS because it uses a clinical monitor instead of a custom monitor: because only one set of ECG contacts are available on the mannequin’s skin in its original configuration, one can only connect one monitor at a time.

Both mannequins do respond appropriately to both fixed and demand pacing modes, although for the HPS to be paced in the demand mode, it must be disconnected from the clinical vital sign monitor’s ECG, which is highly unrealistic in the hospital setting. Finally, both mannequins are able to function across a normal PM’s range of heart rates.

When paced with increasing current intensity, the HPS does not twitch. When stimulated below ventricular capture threshold, the ECG on the clinical monitor does not display typical pacer spikes, which, in addition to the absence of twitching, may lead experienced users to question whether any electrical output is being delivered. Furthermore, it is not possible to simulate any pacing artifact or dissociate electrical capture (wide QRS complex) from ventricular capture (wide QRS complex + pulse) because any pacing above threshold will necessarily produce ventricular capture (ie, no possibility of dissociating the ECG rate from the pulse rate).

When paced with increasing current intensity, the SimMan 3G does not twitch either. When stimulated below ventricular capture threshold, the ECG on the clinical monitor does display pacer spikes. However, as with the HPS, it is not possible to simulate any pacing artifact or dissociate electrical capture from ventricular capture because any pacing above threshold will necessarily produce ventricular capture.

Although both mannequins’ capabilities might be suitable for many simulation purposes, in our experience, their limitations significantly hamper their use with more experienced clinicians, and as we have alluded to previously, these limitations might be especially bothersome if an objective of the simulation is assessment and not only teaching.


The TMM was developed to improve our ability to use simulation to assess the clinical competency of senior residents and attending staff to treat severe bradycardia with hemodynamic instability. It was developed with the 6 proposed TCP clinical competency criteria in mind, hopefully as a temporary fix until improved simulators become commercially available.

In the TMM, participants interact with a modified HPS that provides all physical findings and vital signs except for the ECG signal (Fig. 2). Another mannequin (a SimMan 3G), located in an adjacent room, generates the ECG signal seen by the participants, which is sent lead by lead to the modified HPS using a custom-made cable (see Figure, Supplemental Digital Content 4,, which illustrates the custom-made cable). The cable is made by soldering sequentially:

Diagram of the TMM. A SimMan 3G located in a room adjacent to the simulation suite generates the ECG signal. It is picked up and carried, lead by lead, by a modified cable to the simulation suite. Inside the HPS, a finger latching connector located at the end of the modified cable connects to a corresponding connector within the HPS, sending the ECG signal to the 5 ECG contacts on the surface of the mannequin (to create a signal for the fifth, right leg lead, the signal from the right arm lead of the SimMan 3G is doubled within the modified cable). For clarity, the skin of the HPS has been pulled back to reveal the connection.
  • A. The “patient” extremity of a Single-Patient Use IEC2 5-lead ECG cable (Dräger, Telford, PA), whose other extremity (which normally connects to the monitoring unit) has been cut off.
  • B. Twenty feet of RJ-45- 5E network cable.
  • C. A finger latching connector (like the one normally relaying the signal from the HPS rack to the HPS’ ECG contacts, within the mannequin (see Figure, Supplemental Digital Content 5,, which illustrates the Finger latching connectors enabling the ECG signal of the SimMan 3G to be sent to the HPS’ ECG contacts, replacing the signal normally originating from the HPS rack).

To send the ECG signal lead by lead through the cable, each lead of the IEC2 ECG cable (A) is soldered individually to one of the wires in the RJ-45 cable (B). At the other extremity of the RJ-45 cable, each wire is then soldered to the corresponding wire at the level of the finger latching connector (C). To identify correctly how to solder the wires, continuity testing with a multimeter is used. Inside the HPS mannequin, the wire from the HPS rack carrying its ECG signal is disconnected, and instead, the finger latching connector of the custom-made wire is connected to its corresponding connector in the HPS (see Figure, Supplemental Digital Content 5,, which illustrates the Finger latching connectors enabling the ECG signal of the SimMan 3G to be sent to the HPS’ ECG contacts, replacing the signal normally originating from the HPS rack), finally sending the signal to all 5 ECG contacts at the level of the mannequin’s skin. To make the custom-made wire inconspicuous, it is made to follow the rest of the cables entering the HPS between its legs. Because the SimMan 3G has no right leg lead, the wire carrying the right arm lead signal is split and sent to both the HPS’ right arm and right leg leads. Because the right leg lead is always used as a reference electrode and does not contribute to the ECG signal per se, any lead from the SimMan 3G could be split and sent to the HPS’ right leg lead.

The ability to dissociate the ECG tracing from the rest of the HPS’ physiology is critical: when pacing the TMM below pacing threshold, it is now possible to set the (SimMan 3G’s) ECG rate to match the paced rate (eg, 100 per minute) while keeping the (HPS’) pulse slow (eg, 35 per minute), with all its accompanying physiology. It is also possible to modify the shape of the QRS complex to change it from a narrow complex to a large complex preceded by a PM spike (using the “ventricular pacemaker rhythm” with the “muscular artifact” option on the SimMan 3G, Fig. 1D), without having an automatic increase in the pulse. This allows us to assess whether our participants will be wrongly satisfied with the increase in the ECG rate and the changes in the ECG tracing or will verify ventricular capture by checking the pulse.

The TMM uses an HPS mannequin in the simulation suite because it is an HF mannequin whose vital signs are measured by a real clinical monitor, allowing the signals from 2 different physiologic models to be fed into the mannequin, picked up by the monitoring equipment, and displayed simultaneously on the screen. Mannequins such as the SimMan 3G however, which use a specific (laptop) monitor instead of a clinically used one, where vitals signs are taken directly from the physiologic model without passing through the mannequin, are therefore not suitable for the TMM because it is impossible to incorporate a tracing generated by another model on the monitor. Regarding which mannequin can be used to generate the ECG tracing, a SimMan 3G was chosen in our TMM, but any mannequin or simulator able to generate an ECG signal with the appropriate tracings could do (including, for example, a second HPS).

To permit twitching of the mannequin in response to electrical simulation, a SimDefib (SimAction LLC, San José, CA) is used. This commercially available system is a motorized actuator that briefly lifts the mannequin’s chest in response to electrical stimulation (from a defibrillator or a PM).5 By design, the SimDefib uses a sensing clamp placed on the cable of the multifunction electrode pads to detect the pacing signal, which is not very realistic. To increase realism and robustness, we permanently attached the SimDefib’s sensor directly on the HPS’ internal PM sensing circuit (and defibrillation sensing circuit), within the mannequin (see Figure, Supplemental Digital Content 6,, which illustrates the SimDefib sensor permanently within the HPS). Furthermore, to lower the usual threshold needed to induce twitching from 120 mA to a more realistic 70 mA, 3 turns around the sensor were made. Having the mannequin twitch with pacing below threshold allows us to assess whether our participants wrongly interpret patient twitching as a sign of effective TCP or will check for ventricular capture by checking the pulse. In addition, for experienced participants who have already used TCP, seeing the mannequin twitch allows them to confirm that electrical current is actually being delivered. An alternative to using the SimDefib that might be equivalent for many purposes (although perhaps not as realistic) would be to project a video of a patient twitching.

To allow the use of the multifunction sticky electrode pads used clinically for pacing (and defibrillation) instead of dedicated custom cables that need to be connected to specific contacts on the mannequin and to do this without exposing the mannequin’s pacing circuit to a high-energy shock (if a defibrillation attempt was made, eg, via the multifunction pads), further modification of the HPS is needed. A dedicated multifunction electrode wire whose sticky pads have been cut off and whose wires have been soldered directly to the mannequin’s internal pacing wires is used to connect the PM to the mannequin (see Figure, Supplemental Digital Content 7,, which illustrates the pacing and double ECG monitoring of the TMM). During the simulation, sticky pads can be placed anywhere on the mannequin but are not connected to the PM, which is instead discretely connected by a confederate directly to the mannequin’s internal pacing circuitry. Our participants are instructed to do “as they would clinically,” and this allows us to assess whether they choose the right material and place the multifunction electrode pads appropriately. This solution represents a tradeoff: the HPS mannequin is better protected from a potentially harmful high-energy shock, sticky pads can be used instead of custom cables, and the pads can be placed anywhere on the mannequin, but it is not possible to connect the PM to the mannequin exactly as one would clinically.

Further modification of the HPS used in the TMM is necessary to allow dual monitoring of the ECG (ie, on the vital signs monitor and the PM unit simultaneously). As illustrated (see Figure, Supplemental Digital Content 8,, which illustrates the doubling of the ECG contacts to allow simultaneous monitoring of the ECG on the PM and on the vital sign monitor), the ECG contacts on our METI HPS have been doubled: a 20-gauge copper wire secured on the electrode with the factory nut is used to make the bridge between the original HPS electrode and the doubled one, making the reading of the ECG signal the same on both electrodes. This doubling is permanent on our HPS, allowing dual ECG monitoring whenever the defibrillator/PM is used. In addition, that the doubling is permanent makes it less conspicuous.

The TMM has been used reliably in numerous HF scenarios involving junior residents, senior residents, and attending staff. The TMM has been used both to teach and to assess TCP competency at our institution. As hypothesized when the TMM was developed, a vast majority of participants (all ACLS certified but most having never or seldom used TCP clinically), when placed in a high-stress high-cognitive task load situation requiring TCP, at least initially interpret the combination of patient twitching and changed ECG tracing as evidence of effective pacing. Most fail to take a pulse to check for ventricular capture, and in the minority of those who do check, most do not realize the pulse is still slow (this is consistent with the findings of Ahn et al2 but cannot be attributed to lack of tactile pulse feedback in our model). Only after a significant amount of time has lapsed with the patient remaining hemodynamically unstable despite pacing, or after significant prompting, do participants generally realize that ventricular capture has not been achieved. During debriefing, this important pitfall of TCP is systematically addressed, and participants often report having seen the same problem occur clinically. Participants also often report having felt quite confident with their TCP abilities after ACLS training and are grateful for having had the opportunity to commit some mistakes during simulation rather than during a situation involving a real critically ill patient. This leads us to hypothesize that the training of clinicians on lower-fidelity, often part-task simulators, might lead to overconfidence and perhaps even to negative training.

Despite our best efforts in developing the TMM, certain limitations remain. First, our model uses the same SimMan 3G tracing for both the pacing artifact and electrical capture (with or without ventricular capture). Ideally, these 2 tracings should be distinct:

  • Electrical capture should be characterized by a widening of the QRS complex and a tall broad T wave, changes typical of ventricular complexes (Fig. 1A)
  • Pacing artifacts should resemble a ventricular complex after the pacing spike but with a somewhat less typical shape; in addition, the amplitude of the complex should increase with an increase in pacing output (Fig. 1B, C).

A second limitation occurs when a plethysmography tracing or an arterial line tracing is displayed: at slow rates, a lack of synchrony between the QRS complexes (generated by the SimMan 3G) and the pulse (generated by the HPS) can become apparent. When this occurs, it is possible to restart the rhythm stage in the SimMan 3G model and try to obtain a better synchronism between the 2 models. A third limitation is that with the current version of the SimDefib, the amplitude of the mannequin’s movement is fixed and not proportional to the amount of electrical current used, as occurs clinically. A fourth limitation is TCP pacing with the demand mode: it is not possible with our model to alternate realistically from “patient” generated beats to paced beats, as would normally occur clinically with this mode. Nevertheless, because our scenarios requiring TCP often involve a very unstable patient, fixed rate pacing is most commonly chosen. A fifth limitation occurs when visualizing the PM’s monitor during pacing (which participants seldom do in the hospital context, as most instinctively look at the larger clinical vital signs monitor where more of the clinically relevant information is concentrated): there is a doubling of the PM spikes on PM unit’s monitor, as the device’s ECG not only picks up the spikes generated by the SimMan’s signal but also adds its own. A sixth limitation is that the dissociation between the HPS’ physiologic model and its ECG tracing complicates the simulation of other unstable cardiac rhythms (eg, ventricular tachycardia) with the TMM. The TMM is therefore used in scenarios designed to involve bradycardia as the main unstable rhythm, and other arrhythmias are simulated in separate scenarios, with the mannequins in their original configuration (switching between both configurations can be easily done during debriefing). Finally, the complexity of our model, which involves 2 mannequins, 2 controlling computers, and additional cables and connections, increases the risk of a malfunction occurring. However, setting up the TMM with our modified mannequins takes only minutes, and the TMM has performed very reliably in our institution.

In conclusion, we have proposed a set of criteria defining clinical competency in TCP, reviewed how commercially available HF mannequins perform, pointed out their shortcomings, and proposed a novel TMM solution by combining a modified HPS and a SimMan 3G. Our hope is that newer generations of HF mannequins will be upgraded to offer truly HF TCP simulation, which should not be difficult technically. These improved mannequins would provide a simpler way to teach and to assess this potentially lifesaving technique to participants across the range of clinical experience. In addition, further research is required to determine which aspect of TCP simulation is required depending on the level of experience of the user and the objective of the activity.


1. Neumar RW, Otto CW, Link MS, et al. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010; 122: S729–S767.
2. Ahn J, Kharasch M, Aronwald R, et al. Assessing the accreditation council for graduate medical education requirement for temporary cardiac pacing procedural competency through simulation. Simul Healthc 2013; 8: 78–83.
3. Norman G, Dore K, Grierson L. The minimal relationship between simulation fidelity and transfer of learning. Med Educ 2012; 46: 636–647.
4. Medical Education Technologies Inc. METI HPS User Guide, Version 5; Sarasota, FL: 2009.
5. SimAction Medical Simulation Solutions. SimDefibTM Defibrillation Simulator. Available at: Accessed September 23, 2014.
6. Medtronic. Clinical Information: Pacing Artifact May Masquerade as Capture. Available at: Accessed September 23, 2014.

    Transcutaneous cardiac pacing; Simulation; Assessment

    Supplemental Digital Content

    © 2015 Society for Simulation in Healthcare