© M. Zhou et al., Published by EDP Sciences 2025

1 Introduction

Human activities nowadays are always accompanied and influenced by the use of electricity. Workers within the electrical industry may be exposed to a combination of magnetic fields (MFs), electric fields (EFs), spark discharges, and contact current. This fact has motivated interest and attention towards the assessment of electromagnetic field (EMF) exposure at the workplace. Since the widespread adoption of electrical pacing in the diagnosis and treatment of heart diseases, electromagnetic interference (EMI) has constantly been recognized as a potential hazard to cardiac implantable electronic devices (CIEDs). Given the omnipresence of EMFs, CIEDs, such as pacemakers (PMs) and implantable cardioverter defibrillators (ICDs), are typically equipped with selective filters that significantly reduce or eliminate interference. However, the power frequencies of 50/60 Hz, prevalent in electrical industry workplaces, may not be filtered by the electronic system in CIEDs, as these frequencies fall within the detection range of human cardiac signals (0.05–150 Hz). International guidelines (ICNIRP, 2010; IEEE, 2019) are widely acknowledged and implemented to protect workers from overexposure, however, individuals with CIEDs are outside of their scope. Therefore, exposure to EMFs in the vicinity of industrial electrical apparatus may be considered as a particular risk that employers are obliged to assess, as indicated in the Directive 2013/35/EU.

Studies have been conducted to ensure the safety of patients with cardiac implants in public (Stunder et al., 2017; Driessen et al., 2019; Magne et al., 2020), and some occupational scenarios (Magne et al., 2012; Tiikkaja et al., 2013; Napp et al., 2015; Mattei et al., 2019, 2021). The safety of magnetic resonance imaging (MRI) for CIEDs has also been a prominent area of focus (Nazarian et al., 2006; Nordbeck et al., 2009; Strom et al., 2017; Zeng et al., 2019). In recent years, studies have been carried out on newly applied technologies, such as wireless communication and charging (Seidman et al., 2014; Seckler et al., 2015; Mattei et al., 2016), electrical vehicle and its charger (Tondato et al., 2017; Lennerz et al., 2023; Wase, 2023), transcutaneous electrical nerve stimulation (Badger et al., 2017; Egger et al., 2019; Suhail Arain et al., 2023), and small rare-earth magnets used inside electronic devices (Seidman et al., 2021). However, few studies have prioritized the assessment for power frequencies. Given the number of workers carrying CIEDs (i.e., CIED-employees) continues to increase and their inevitable presences in the workplaces, concerns have been expressed regarding potential dysfunctions of implanted medical devices that may be associated with exposure to high-intensity electric and magnetic fields generated by the transmission, distribution, and utilization of electricity.

In general, exposure to EMFs is considered safe in environments accessible to the public. However, in the workplace, exposures are various and may be higher. CIED-employees are especially susceptible to EMFs and factors such as device sensitivity, implantation configuration, and pacing technology can all affect the performance of CIEDs. To ensure their safety in the workplace, it is important that occupational physicians evaluate each case individually. The European Standard series EN 50527 has then standardized the process of evaluating the exposure of active implantable medical devices (AIMDs) to EMFs. This standard brings up the need for establishing a non-clinical investigation that is simple, reproducible, and risk-free, ensuring effective and consistent risk assessments in the workplace.

In this paper, we established a setup for evaluating occupational hazards associated with cardiac implants exposed to high-intensity EFs at low frequencies in the workplace. Correspondences were built up between in vitro testing and real-case exposures to conduct non-clinical investigation in the risk assessment. A voltage injection system (VIS) was proposed as an on-site solution in the workplace to conduct effective equivalent tests. The VIS assessment allows us to reduce the complexity of the study under EF exposures, which is restrictive in terms of high-intensity voltage, installations, and safety protocols. By using this system, employers can perform simpler tests with reproducible perturbations, thereby facilitating more straightforward and consistent risk assessments. VIS assessment were conducted on four cardiac implants (two PMs and two ICDs) with the presence of cardiac signals, as a demonstration of its application.

2 Materials and methods

2.1 Assessment process in the framework of EN 50527

For a certain case to be assessed, the European Standard EN 50527 indicates a general risk assessment above all in order to determine whether a specific assessment process is necessary. The manufacturers guarantee the immunity of CIEDs conforming to EN 45502 and ISO 14117, which provide safety requirements and technical electromagnetic compatibility test protocols for manufacturers to follow. CIEDs entering the European market are expected to operate without interference as long as the General Public Reference levels of Council Recommendation 1999/519/EC are not exceeded. If the exposure is higher, no risk assessment is required if a responsible physician has confirmed that a sufficient history of uninfluenced behavior at the workplace exists to exclude clinically significant interaction. Otherwise, a specific risk assessment for the CIED-employee is required.

In the specific assessment process, information relevant to the equipment producing EMFs, information collected from the responsible physician, and clinical details about the CIED-employee, such as their dependency on the CIED, are taken into consideration.. If this information fails to exclude risk to the CIED-employee of the EMI at the workplace, an additional investigation is necessary. The additional investigation can be conducted in either clinical method (or in vivo) or non-clinical methods (in vitro or comparative study). Considering complex environments and individual situations, the clinical method might be contraindicated in some circumstances. In consequence, non-clinical investigation shall be carried out. At the end, a final report of the investigation shall be written, containing all the methods applied, the findings and the conclusions from the process.

2.2 Non-clinical investigation method

The EN 50527-2-1 for pacemakers and the EN 50527-2-2 for ICDs proposed in vitro testing as one among the investigation approaches. Interference thresholds may be obtained through provocative studies (Gerçek et al., 2020; Zhou et al., 2022). Due to the inhomogeneity of the human body, reproducing the complex variations of induction in the EF exposure using homogeneous phantom material is challenging. To address this issue, a funnel-shaped phantom was used to replicate the induction in the chest and in the heart (where cardiac implants are typically installed) due to its geometric design. This phantom has been employed in the interference investigation for cardiac implants by in vitro testing under laboratory EF exposures, with equivalence factors of 2.39 for unipolar sensing and 3.64 for bipolar sensing. These equivalence factors describe the ability to substitute the real case while retaining important characteristics. This laboratory EF exposure system was established in the laboratory capable of producing exposure up to 50 kV/m (equivalent to 119.5 kV/m in unipolar sensing and 182 kV/m in bipolar sensing in real case), however, requiring strict experimental environment and security protections (Zhou, 2023).

To provide an accessible approach for employers in the workplace, we have established a voltage injection system (VIS) using the same funnel-shaped phantom (Fig. 1A). This setup is expected to produce similar induction in the phantom without a complex configuration as for the laboratory EF exposure. A metal grid with a diameter of 280 mm was placed on top of the phantom solution, with a banana socket on the top to fix the injection cable (Fig. 1B). Voltages were injected into the grid from a generator to produce EF induction in the phantom, while the phantom base was connected to the common ground of the setup. We used the digital lock-in amplifier HF2IS from Zurich Instruments™ (Zurich, Switzerland) to generate small-scale voltage injections while simultaneously measuring the voltage applied to the phantom. Its high-impedance input ensures measurement fidelity and minimizes loading effect. The device’s supporting software LabOne^®, enables direct adjustment of field levels and real time monitoring on a PC.

This setup was simulated in the CST Studio^®, showing that same EF induction distribution may be obtained as in laboratory EF exposure (Zhou et al., 2024). Under exposure of 1 kV/m at 50 Hz, the induction gradually decreases upwards in the upper part of the phantom (funnel shape), with an average induced EF of 0.8 mV/m, same to that in the thorax where impulse generators are installed; the induction remains constant in the lower part of the phantom (cylindrical shape), with an induced EF of 4 mV/m as in the heart, where the sensing lead tip is located (Fig. 1C). In vitro testing may be carried out using it, as long as the same configuration of the device is applied as in laboratory exposure. To simulate the real-case implantation environment and to observe the dysfunctions caused by EF exposures, heart signals of 1 Hz (60 bpm) were continuously sent to cardiac implants via optic fiber during the assessment using a cardiac signal generation unit (ECG Unit) applied in the laboratory exposure system (Zhou, 2023). Typically, physicians set the sensitivity of cardiac implants to be 3–4 times lower than the cardiac signal amplitude. In this study, the amplitude of cardiac signals was configured to be 4 times the sensitivity setting. Therefore, VIS assessment is validated for reproducing the same exposure conditions to the cardiac implant and allowing observation of real-case interference without an exceptional experimental environment.

At low frequencies, the dominant effect is the induction of electric current density in the body, which is considered an assessable quantity due to its direct association with biological effects. Therefore the interference investigation may be conducted also based on the study of the induced voltages on the leads. A measuring system dedicated to evaluating the induced voltages on cardiac implants under EF at ELFs was designed and applied for the study of laboratory EF exposure (Zhou et al., 2024). The findings enabled the determination of equivalence factors between real case exposure and laboratory EF exposure. In this work, the measuring system was designed to conduct measurements of induced voltages on cardiac implants for VIS, while maintaining the same device implantation configuration. Consequently, the equivalence for VIS to the real case may be determined accordingly, along with the association of exposures.

Fig. 1 Voltage injection system configuration.

3 Results

3.1 Establishment of VIS assessment

The measuring system and the same device implantation configuration in the study of laboratory EF exposure (Zhou et al., 2024), were used in the VIS to evaluate the induced voltages at the input of cardiac implants under EF exposure at power frequency (50 Hz). The induced voltages were measured with increasing voltage injection levels. The results were postprocessed to obtain the component at 50 Hz in order to eliminate the potential external inference from other frequencies. Figure 2 shows the measurement results in VIS for bipolar and unipolar sensing. The blue dots represent the induced voltages obtained with the measuring system, while the black lines show the simulation results. Linearity between the injected voltage and the induced voltage can be observed in both bipolar and unipolar sensing mode. The gradients of the trend lines from the results indicate the induced voltage for normalized injection (1 mV). We measured an induced voltage of 772 μV and obtained 668 μV by simulation for unipolar sensing, with an injection of 1 mV. For bipolar sensing, we measured 121 μV and obtained 117 μV by simulation for bipolar sensing, with an injection of 1 mV as well.

Previous study has performed such measurements for laboratory EF exposure (Zhou et al., 2024). The induced voltages for a normalized electric field exposure of 1 kV/m were 438 μV/80 μV (unipolar/bipolar), corresponding to induced voltages of 183 μV/22 μV (unipolar/bipolar) in real case. Integrating with the findings in VIS, the association of two exposure systems and real case may be established based on the induced voltage quantities. Figure 3 demonstrates this association for both unipolar sensing and bipolar sensing. For a given exposure in real case (horizontal axis), the EF levels required to conduct an assessment in the laboratory EF exposure system can be found on the left axis; the injected voltages that are required in the VIS assessment can be found on the right axis. The exposure levels are limited to real-case exposure of up to 35 kV/m, complying with the exposure limit value (ELV) of EN 50647. The exposure references given by the ICNIRP guidelines and the IEEE standard C95.1-2019 were labeled in the figures, along with the Action Levels (ALs) indicated in the Directive 2013/35/EU. These figures may work as lookup tables for non-clinical investigation of an employee bearing a CIED in the specific assessment process. By checking the voltage injection dose required to reproduce the exposure in a real-case scenario, VIS assessment may be performed on the device under test.

Fig. 2 Comparison of measured and simulated induced voltage versus voltage injection for unipolar and bipolar configurations.

Fig. 3 Comparison of two exposure systems and real case based on the induced voltage quantities for unipolar sensing and bipolar sensing.

3.2 Specific assessment process

In this paper, we intend to interpret the application of VIS assessment in the specific assessment process on four CIEDs (two PMs and two ICDs) from Medtronic™ (Minnesota, U.S.). We assume that the CIED-employees bearing these devices may be exposed higher than 5 kV/m. In consequence, additional investigations are required. Considering the various features of natural cardiac signals that different patients may possess and different clinical situations, each cardiac implant was inspected with three sensitivities: the nominal sensitivity (Nom.), which is the default setting and most commonly used; the maximum sensitivity (Max.) of the device, representing patients with very low-amplitude cardiac signals or high dependency on pacing; and a third sensitivity (3rd Sen.), selected randomly between the nominal and the minimum sensitivity at 1.5 to 2 times the nominal setting, representing patients with high-amplitude cardiac signals and stable heart rhythms, or a clinical chose to prevent EMI. To conduct non-clinical investigations in these cases, the employers may ensure CIED-employees’ safety by following processes in the EF exposed workplace:

1. Survey on implanted cardiac device and its setting

   Multiple factors may affect the severity of EMI, such as the sensing lead, sensing configuration, and sensitivity setting of the device, etc. We chose the following settings :

   • Implantation method: right pectoral implantation and vertical sensing lead end (Tip-Ring pair) for all devices (worst-case scenario among pectoral implantation (Gustrau et al., 2002));

   • Sensing method: unipolar and bipolar sensing for pacemakers, bipolar sensing for ICDs (unipolar sensing is not applicable for ICDs);

   • Device information (Tab. 1).

2. Survey on field source at workplace

   General information on the measurement of the fields is provided in IEC 61786-2. Field levels at the workplace can be determined either by measurement or by modeling. Here we assume that the four employees are exposed to exposure levels at the limits specified in the Directive and standards.

   • Exposure field type: EF;

   • Characteristics of exposure: power frequency (50 Hz), homogeneous;

   • Direction: Vertical (from exposed object’s head to bottom);

   • Values (Tab. 2).

3. VIS assessment

   Possessing the information collected from the cardiac devices and the field sources, we carried out the VIS assessment by applying the corresponding voltage injection doses in the immunity test for the four devices under test with three sensitivities (Max., Nom, 3rd Sen.) in unipolar and bipolar sensing mode (if applicable).

   In an assessment, a certain amount of voltage is applied to the phantom to generate expected exposure for the cardiac implant under test while cardiac signals are continuously sent to this device. This cardiac implant is examined by its programmer after exposure of 1 minute. Missed cardiac signal detection, erroneous cardiac signal detection, and incorrect ventricular fibrillation (VF) detection are considered as device dysfunctions, which indicate that the cardiac implant is subjected to significant interference and not able to function properly. This procedure takes time but it was always performed three times to guarantee reproducibility.

   According to the lookup table in Figure 3, an injection of 1.19 mV was applied for unipolar setting and 0.91 mV for bipolar setting in the assessment of Source-I (5 kV/m); an injection of 2.37 mV was applied for unipolar setting and 1.82 mV for bipolar setting in the assessment of Source-II (10 kV/m); An injection of 4.74 mV was applied for unipolar setting and 3.64 mV for bipolar setting in the assessment of Source-III (20 kV/m).

   No abnormality was found on PM-I and PM-II in terms of Source-I and Source-II. Certain cardiac signals were missing when PM-I and PM-II in unipolar setting with the maximum sensitivity (1 mV) were subjected to Source-III. As for ICD-I and ICD-II (possess only bipolar sensing), erroneous cardiac signal detections and incorrect VF detections were observed when they are subjected to Source-II with maximum sensitivity, plus ICD-II with nominal sensitivity. Except for ICD-I with maximum sensitivity, no abnormality was found in terms of Source-I. In terms of Source-III, dysfunctions were observed in all the cases of ICDs, except ICD-I with 3rd sensitivity (0.6 mV).

4. Further investigation

   Additional tests were carried out to obtain the interference thresholds in these cases. By increasing the injected voltage in VIS, the maximum exposure level at which no dysfunction was observed, was identified as the interference threshold. Each threshold was tested three times as well to ensure reproducibility. Figure 4 illustrates the thresholds identified through the immunity tests conducted for all the cases. The sources are represented by three planes. Interference threshold surpassing the plane indicates that no dysfunction occurred for such source.

5. Documentation

   As stated in EN 50527, a final report of the investigation shall be completed and in the possession of the employer to evaluate the situation of the CIED-employees, including overall risk assessment process, method chosen, rationale for the choice, findings, and conclusions.

Fig. 4 Interference thresholds (in kV/m) identified through VIS assessment conducted for all the cases (Yellow bars: maximum sensitivity; Purple bars: nominal sensitivity; Green bars: the third sensitivity; Grey plane: Source-I; Blue plane: Source- II; Red plane: Source-III).

4 Discussion

In this work, a risk assessment system VIS was designed based on the experimental investigation of the induced voltages on the cardiac implants under EF exposures. This system requires only a phantom setup and a voltage source. By controlling the voltage injection toward the phantom, the EF exposure in the implantation area may be characteristically reproduced. For many enterprises, due to their diversity, conducting immunity tests in laboratory EF exposure, which requires rigorous experimental configuration and maintenance, is neither practical nor efficient. In contrast, in the VIS assessment, the phantom ensures the reproduction of exposure characteristics, while its easy setup ensures that the process is portable and accessible. It presents an ideal alternative for conducting tests for cardiac implants, being simple and not requiring complex materials. Therefore, VIS assessment is proposed for the non-clinical investigation in the process of specific assessing for CIED-employees in the workplace.

Experimental measurements of induced voltages at the input of cardiac implants in VIS were carried out using the measuring system designed previously. The bipolar sensing mode demonstrates a significant reduction of induced voltage, which is consistent with previous studies [29]. Comparing the measurement and simulation results of induced voltage on the lead, there is a difference of 15.6% for unipolar sensing and a difference of 3.4% for bipolar sensing. Considering the small scale of the voltage, the results are in reasonable agreement. Two lookup tables (Fig. 3) were established, one for unipolar sensing and another for bipolar sensing, to determine the exposure dose for conducting equivalent tests of real-case exposure in the laboratory EF exposure system and VIS. The associations in the lookup tables are based on induced voltage’s studies conducted previously (Zhou et al., 2024) and in this paper. According to the lookup tables, microvolt-level voltages can reproduce equivalent exposures up to 35 kV/m (the exposure limit) in VIS. Its efficiency and easy setup facilitate the experimental study of CIEDs under EF exposures.

A thorough non-clinical investigation for four CIEDs (two PMs and two ICDs) in specific risk assessment process was carried out as the application illustration of VIS assessment. The process utilized VIS assessment to determine whether the devices are subjected to severe interference from the assigned exposures. Considering the diversity of human cardiac signals, we tested the devices at maximum, nominal, and an additional higher sensitivity (3rd Sen.) representing patients with higher amplitude cardiac signals. The amplitude of cardiac signals delivered from the ECG Unit varies accordingly. The findings indicate that cardiac implants at maximum sensitivity are more susceptible to EF exposures than at other settings. Interference was observed in all devices for Source-III (High AL in Directive 2013/35/EU), except for pacemakers in bipolar sensing mode. We may notice the influence of the variety of sensitivity, but it is important to note that the maximum sensitivity represents an extreme scenario. At nominal sensitivity, the most common setting, no abnormalities were found in any of the devices for Source-I (public exposure); both ICDs with nominal settings were perturbed before Source-III, indicating they are more susceptible than pacemakers. Immunity tests for these cases were conducted under laboratory EF exposure. Approximate thresholds were obtained (Fig. 5), further validating the performance of VIS. Despite the cohort size is small (N = 4), these results highlight the importance of conducting assessments for specific cases.

Fig. 5 Interference thresholds (in kV/m) identified through laboratory exposure system (Yellow bars: maximum sensitivity; Purple bars: nominal sensitivity; Green bars: the third sensitivity; Grey plane: Source-I; Blue plane: Source-II; Red plane: Source-III).

5 Conclusion

In this paper, we established an exposure system VIS for evaluating occupational hazards associated with cardiac implants exposed to high-intensity EFs at power frequency in the workplace. In this system, we conducted investigations on the interference thresholds of cardiac implants and experimental measurements of induced voltages. Measurement and numerical results showed good agreement. The correspondences between in vitro testing and real-case exposures based on the findings of measurement were built up in two lookup tables. Equivalent exposure to real case may be produced by applying the exposure dose indicated in the lookup tables in in vitro testing. This system may be used in non-clinical investigation during specific assessing of an occupational exposure case. To illustrate the application of VIS, we conducted a complete process of specific assessing for four CIEDs. An assessment approach, along with the analysis method, was provided for further applications in the workplaces.

Acknowledgments

This project was supported by Réseau de Transport d’Électricité.

Funding

This research did not receive any specific funding.

Conflicts of interest

The authors declare that they have no conflict of interest.

Data availability statement

The research data associated with this article are included within the article.

Author contribution statement

All authors have read and agreed to the published version of the manuscript.

Ethics approval

Ethical approval was not required.

Informed consent

Not applicable.

Reference

All Tables

All Figures

	Fig. 1 Voltage injection system configuration.
In the text

	Fig. 2 Comparison of measured and simulated induced voltage versus voltage injection for unipolar and bipolar configurations.
In the text

	Fig. 3 Comparison of two exposure systems and real case based on the induced voltage quantities for unipolar sensing and bipolar sensing.
In the text

	Fig. 4 Interference thresholds (in kV/m) identified through VIS assessment conducted for all the cases (Yellow bars: maximum sensitivity; Purple bars: nominal sensitivity; Green bars: the third sensitivity; Grey plane: Source-I; Blue plane: Source- II; Red plane: Source-III).
In the text

	Fig. 5 Interference thresholds (in kV/m) identified through laboratory exposure system (Yellow bars: maximum sensitivity; Purple bars: nominal sensitivity; Green bars: the third sensitivity; Grey plane: Source-I; Blue plane: Source-II; Red plane: Source-III).
In the text