Article Text

Utility and acceptability of remote 6-lead electrocardiographic monitoring in children with inherited cardiac conditions
  1. Claire Margaret Lawley1,2,3,
  2. Katarzyna Luczak-Wozniak1,4,
  3. Sheng-Chia Chung3,5,
  4. Ella Field1,3,
  5. Annabelle Barnes1,3,
  6. Luke Starling1,3,
  7. Elena Cervi1,3,
  8. Juan Pablo Kaski1,3
  1. 1Centre for Inherited Cardiovascular Diseases, Great Ormond Street Hospital for Children NHS Foundation Trust, London, UK
  2. 2The University of Sydney Children’s Hospital Westmead Clinical School, Faculty of Medicine and Health, The University of Sydney, Sydney, New South Wales, Australia
  3. 3Centre for Paediatric Inherited and Rare Cardiovascular Disease, Institute of Cardiovascular Sciences, University College London, London, UK
  4. 4Department of Pediatric Cardiology and General Pediatrics, Medical University of Warsaw, Warszawa, Poland
  5. 5Health Data Research UK, Institute of Health Informatics, University College London, London, UK
  1. Correspondence to Professor Juan Pablo Kaski; j.kaski{at}ucl.ac.uk

Abstract

Objective This pilot study sought to investigate the utility and acceptability of the KardiaMobile 6-lead ECG (KM6LECG) as a tool for remote monitoring in children with inherited cardiac conditions.

Design A single-centre prospective cohort study. Children underwent standard clinical evaluation including a 12-lead ECG and a KM6LECG in the clinic. Participants recorded KM6LECGs monthly at home for 3 months. Families completed a questionnaire on their experience.

Setting Great Ormond Street Hospital Centre for Inherited Cardiovascular Diseases.

Participants 64 children: 22 with hypertrophic cardiomyopathy (HCM); 22 with long QT syndrome and 20 unaffected siblings (controls).

Main outcome measures Comparison of data extracted from the clinic 12-lead ECG and supervised KM6LECG, and the supervised and unsupervised KM6LECG recording.

Results Of 64 children (35% female, mean age 12 years), 58 had a baseline 12-lead ECG and appropriate baseline KM6LECG. In children with HCM, abnormalities in ventricular depolarisation/repolarisation in the limb leads of the 12-lead ECG were reliably reproduced. From the whole cohort, there was a strong positive correlation between the corrected QT interval from the 12-lead ECG and baseline KM6LECG (intraclass correlation coefficient=0.839) and baseline KM6LECG with an unsupervised KM6LECG (intraclass correlation coefficient=0.736). Suspected ‘lead’ misplacement impacted 18% of unsupervised recordings. Overall, the acceptability of the KM6LECG to families was good.

Conclusions The KM6LECG provides an accurate tool for assessing some ECG abnormalities associated with paediatric inherited cardiovascular disease and may provide a useful at-home adjunct to face-to-face clinical care of children requiring ECG assessment.

  • Paediatrics
  • Cardiology

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • The KardiaMobile 6-lead ECG (KM6LECG) can be used to document heart rhythm abnormalities outside the clinic setting.

  • In adults, the KM6LECG allows accurate measurement of the QT interval.

WHAT THIS STUDY ADDS

  • In a purely paediatric cohort of individuals with hypertrophic cardiomyopathy, long QT syndrome and controls, the KM6LECG can accurately and reproducibly identify some ECG abnormalities seen in the clinic.

  • Performed at home, unsupervised, these measurements can be reproduced by families, accepting some issues with suspected ‘lead’ misplacement.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • The KM6LECG may allow remote ECG assessment in children undergoing screening for inherited cardiac conditions and other cardiac conditions.

Introduction

The COVID-19 pandemic provided impetus to explore delivery of medical care outside the traditional face-to-face clinic model. Evaluation of the 12-lead ECG at routine clinic visits is a key component of serial follow-up for children with, or undergoing screening for, an inherited cardiac condition (ICC). Over 97% of children with hypertrophic cardiomyopathy (HCM) have an abnormal baseline ECG.1 In long QT syndrome (LQTS), the diagnosis usually requires the presence of QT prolongation, often associated with abnormal T-wave morphology, and the duration of the corrected QT interval (QTc) has prognostic value.2–4 Outside the ICC setting, accurately and reproducibly assessing cardiac repolarisation forms part of the monitoring for other paediatric conditions, such as during certain medical therapy and cancer treatment.5 6

KardiaMobile (AliveCor) is a mobile personal ECG device that uses a smartphone application to record a 30 s ECG, either single lead or six lead, analogous to the six limb leads (I, II, III, aVL, aVR, aVF) on a conventional 12-lead ECG.7 It has been validated in adults in the detection of atrial fibrillation and for assessing the QTc (receiving United States Food and Drug Administration approval for the measurement of the QTc in adults in 2021), other ECG intervals and in the remote assessment of conduction following transcatheter aortic valve implantation.8–13 In children, KardiaMobile has been shown in small studies to be a useful alternative to standard cardiac event monitoring for the investigation of palpitations14–16 and, in its single lead form, provides an accurate ECG in children with normal hearts and structural heart disease.17

The aim of this pilot study was to explore the utility and acceptability of the KardiaMobile 6-lead ECG (‘KM6LECG’) as a tool for remote monitoring in children with HCM and LQTS, as exemplars of ICC.

Method

This was a single-centre prospective cohort study. 64 individuals followed up at the Great Ormond Street Hospital Centre for Inherited Cardiovascular Diseases were recruited: 22 children with HCM; 22 children with LQTS; and 20 unaffected siblings attending clinic for clinical screening (controls). All participants underwent standard systematic clinical evaluation as part of their outpatient clinic visit, including 12-lead ECG and cardiac imaging when indicated.

Evaluation of KM6LECG

Following the 12-lead ECG, participants immediately undertook a supervised KM6LECG, using the positioning recommended by the device company. Participants were then loaned the device and asked to record further KM6LECGs once a month for 3 months. Participants sent a portable document format (PDF) copy of the unique anonymised ECGs securely to the research team.

Baseline 12-lead ECG, supervised KM6LECG and one of the monthly unsupervised KM6LECGs were evaluated by an investigator blinded to clinical details. ECG sweep speed was 25 mm/s and voltage 10 mm/mV. 12-Lead ECG machine (GE MAC5500HD) was set up using the standard 150 Hz filter. The KM6LECG filter was left as ‘automatic’ (device assessment as to whether a 50 or 60 Hz filter is applied). Measurements were taken from lead II or V5 for the 12-lead ECG and lead II from the KM6LECG, averaged over three beats unless otherwise specified. The following data were collected: heart rhythm; presence of ectopic beats; heart rate; P-wave morphology (classified as ‘left atrial dilatation’ if notched with a peak-to-peak interval of >0.4 ms or ‘right atrial dilatation’ if >2.5 mm amplitude in lead II18 19); PR interval; presence of abnormal Q-waves; QRS axis, duration (>120 ms classified as abnormal) and morphology; ST-segment abnormality; T-wave morphology (‘giant’ >10 mm, asymmetric or flat <2 mm in multiple leads classified as abnormal)1 20; rate-corrected QT (calculated using the maximal slope technique and corrected for heart rate using Bazett’s formula). ‘Lead’ misplacement on the KM6LECG was suspected if the P-wave axis was abnormal with additional abnormalities of the QRS and T-wave axes, discrepant from the baseline 12-lead ECG. If lead misplacement was suspected, QTc was still calculated, but other data were not collected. KM6LECG quality was labelled as ‘poor’ if there was significant baseline wander or ‘noise’. If an unsupervised trace was poor quality or lead misplacement was suspected it was not chosen for analysis, unless there was no alternative. The additional two unsupervised KM6LECGs were assessed for suspected ‘lead’ misplacement and quality. A second investigator assessed the baseline supervised 12-lead ECG and KM6LECG in a random subset of 12 children to determine interobserver variability.

ECG parameters known to be most commonly abnormal in HCM and LQTS were the focus of analysis,1 21 as outlined in table 1.

Table 1

Electrocardiographic features assessed in children with LQTS and HCM

In keeping with previous work, ‘excellent agreement’ between QTc on the baseline 12-lead ECG and KM6LECG was defined as <20 ms absolute difference, and ‘clinically acceptable agreement’ was defined as <40 ms absolute difference.11

For children with HCM, two additional combined categorical variables were created as surrogates for the possible presence of disease: ‘any repolarisation abnormality’ (including any of abnormal T-wave morphology, ST depression or elevation, abnormal T-wave inversion or QTc prolongation as defined above) and ‘HCM-suggestive features’ (including any of abnormal T-wave inversion, pathological Q-waves, left or extreme axis deviation or, for the 12-lead ECG only, voltage criteria for left ventricular hypertrophy). Similarly, for children with LQTS, a combined variable of ‘prolonged QTc and/or abnormal T-wave morphology’ was created.

Statistical analysis

Statistical analysis was performed using SPSS V.28 and SAS V.9.4. Baseline demographics were reported as mean±SD if normally distributed or median (IQR) if not. Analysis was undertaken between data extracted from the clinic 12-lead ECG and clinic supervised KM6LECG, and supervised and an unsupervised KM6LECG recording. For continuous measures, the differences between measurements assessed by two methods, adjusted for patient groups, were compared using a linear mixed model with a random intercept to account for the clustering by individuals. To estimate the correlation between two measurements assessed by two methods made on the same individual, the intraclass correlation coefficient was used. Categorical data were examined for sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) using the 12-lead ECG as the ‘gold standard’. For the analysis of categorical variables, the kappa statistic was used. Further detailed statistical analysis including the sensitivity analysis and interobserver variability assessment are outlined in online supplemental web appendix 1.

Results

The mean age was 12 years; demographic data are outlined in table 2. All children successfully underwent a 12-lead ECG at a face-to-face clinic visit. Five children were missing baseline KM6LECG data and one had suspected ‘lead’ misplacement on their baseline KM6LECG. These six children were excluded from subsequent comparative analysis, leaving 58 children with a baseline 12-lead ECG and appropriate baseline KM6LECG for analysis (18 control, 20 HCM, 20 LQTS) (table 2). Genetic variants in the children with LQTS are summarised in online supplemental table 1. An example of a baseline 12-lead ECG from a child in the control group and their supervised KM6LECG and one of their unsupervised KM6LECGs is shown in figure 1(1A)–(1C).

Table 2

Baseline characteristics of children and electrocardiographic data availability and quality

Figure 1

Examples of the standard 12-lead ECG and the corresponding supervised and unsupervised KardiaMobile 6-lead ECGs (KM6LECGs). Examples of limb leads from standard 12-lead ECG (25 mm/s, 10 mm/mV) (A) and KM6LECG traces supervised on the same day as baseline ECG (B) and unsupervised KM6LECG traces at a later date (C), KM6LECG traces display formatted to 12-lead ECG layout, lead I–II and aVR, aVL, aVF readings taken in parallel (1) child from control group with normal resting ECG, (2) child with hypertrophic cardiomyopathy baseline ECG demonstrating abnormal repolarisation with giant T-waves, (3) child with long QT syndrome with baseline ECG demonstrating QT prolongation and asymmetric T-wave morphology.

The median time between supervised and unsupervised KM6LECGs used for data analysis was 37 days (IQR 37). All but three readings were more than 3 weeks apart, up to a maximum of 136 days apart. Six children had their paired clinic supervised 12-lead and KM6LECGs after the three remote KM6LECGs.

Heart rate, depolarisation and repolarisation interval measurement on the KM6LECG

There was a strong positive statistically significant correlation between the 12-lead ECG and baseline KM6LECG for QRS axis (intraclass correlation coefficient=0.850), QRS duration (intraclass correlation coefficient=0.839) and QTc (intraclass correlation coefficient=0.839). Sensitivity analysis for QTc led to similar results, included in online supplemental web appendix 1. There was moderate agreement for heart rate (intraclass correlation coefficient=0.624). The mean heart rate increased by 3.25 beats/min (p=0.025) between the supine 12-lead ECG and seated KM6LECG.

There was substantial agreement for the QTc being abnormal or normal between the baseline 12-lead and baseline KM6LECGs (kappa 0.714, p<0.001). 39 of 56 children (69%) had an excellent agreement between the readings, and an additional 12 had clinically acceptable readings, resulting in 51 of 56 (91%) having a clinically acceptable agreement between readings. The overall sensitivity of the KM6LECG in detecting a prolonged QTc interval (17 of 56 children) was 76%, with a specificity of 92%.

The KM6LECG in paediatric HCM

Contingency analysis is shown in table 3, using the baseline 12-lead ECG as reference. For the two combined variables ‘any repolarisation abnormality’ and ‘HCM suggestive features’, when only features in the limb leads of the 12-lead ECG were included, the sensitivity improved. A visual example of the limb leads from a 12-lead ECG in a child with HCM, the corresponding supervised KM6LECG and one of the child’s unsupervised KM6LECGs is shown in figure 1(2A)–(2C). Three children with HCM or LQTS were mislabelled as having abnormal repolarisation on the basis of the KM6LECG when this was not felt to be the case on the 12-lead ECG, in each instance the T-wave appeared asymmetric on the KM6LECG.

Table 3

Findings on the KardiaMobile 6-lead KM6LECG compared with the baseline 12-lead ECG in children with HCM and LQTS

The KM6LECG in paediatric LQTS

The KM6LECG was sensitive and specific in identifying a prolonged QTc in children with LQTS. Of the five children with a QTc >500 ms, the KM6LECG QTc was >500 ms in three (60%), 499 ms in one and not possible to assign in one on the KM6LECG due to a combination of abnormal T-wave morphology and a poor-quality trace. Contingency analysis is shown in table 3. A visual example of the abnormalities described on the limb leads of a 12-lead ECG in LQTS and the corresponding supervised KM6LECG and one of the child’s unsupervised KM6LECGs is shown in figure 1(3A)–(3C).

Reproducibility

Of the supervised KM6LECGs, 4/59 (6.8%) were labelled as ‘poor quality’ versus 10/168 (6%) of unsupervised KM6LECGs (table 2). ‘Lead’ misplacement was suspected in 31/168 (18.5%) of KM6LECGs performed at home, versus 1/59 (1.7%) performed supervised. This precluded the ability to assess the axis and some repolarisation abnormalities, although QTc was still calculated.

For those with sufficient data, there was a strong positive correlation between QRS parameters expected to remain static over a short duration of follow-up, with near perfect agreement between the supervised QRS axis and duration as measured on the KM6LECG in clinic and at home (QRS axis intraclass correlation coefficient=0.962, QRS duration intraclass correlation coefficient=0.911). There was a strong positive correlation between the QTc as calculated on the supervised KM6LECG compared with an unsupervised recording (intraclass correlation coefficient=0.736). There was substantial agreement between the QTc being abnormal or normal on the baseline KM6LECG compared with an unsupervised recording (kappa 0.788, p<0.001).

There was a high interobserver correlation for QTc measurement in the subset where this was assessed. Full interobserver analysis results are included in online supplemental web appendix 1.

Child and parent satisfaction

Overall, from the 56 survey responses returned, the acceptability of the KM6LECG was good. Three of the 18 children aged over 12 years reported some technical difficulties. The questions and responses are shown in online supplemental figure 1.

Discussion

This study demonstrates that the KM6LECG can be used in children in a ‘real world’ setting to accurately replicate the abnormalities identified in the limb leads of the ECG, including in children with both HCM and LQTS. Furthermore, these recordings can be accurately reproduced by unsupervised families at home and were felt by families to be an acceptable and ‘potentially helpful’ tool for remote ECG monitoring.

The KM6LECG reliably replicated the abnormal ECG findings from the limb leads of children with HCM, including presence of Q-waves and abnormalities in QRS axis, duration and ventricular repolarisation. Notably, our study has shown that these parameters, expected to remain static in short-term follow-up, can also be reliably measured in children from KM6LECGs taken by families unsupervised at home. The absence of the precordial leads in the KM6LECG did mean that in six children who had abnormal findings only in the precordial leads, the KM6LECG was reported as normal.

For children with LQTS, the KM6LECG was a sensitive (91%) and fairly specific (88%) tool at identifying a prolonged QTc and replicating some abnormalities in T-wave morphology. This is similar to previous paediatric and adult work using a supervised single lead recording.11 22 The finding that the five children with a QTc >500 ms on their baseline 12-lead ECG, arguably at highest risk, had similarly abnormal KM6LECGs, with QTc >499 ms in four and one with abnormal T-wave morphology (making accurate calculation impossible), was reassuring. Using a binary outcome of normal or abnormal QTc based on prespecified values, with the 12-lead ECG as reference gold standard, the sensitivity (76%) of the KM6LECG for detecting a prolonged QTc was higher than the 54% reported in a previous work using the KM6LECG in children, with similar specificity.23 From the sensitivity analysis, the overall reliability of the unsupervised KM6LECG was improved by removing potentially overestimated measures.

The KM6LECG is acceptable to, and able to be used by, families at home, although the number of unsupervised KM6LECGs suffering from suspected ‘lead’ placement issues was not insignificant. We are not aware of suspected ‘lead’ misplacement being historically being assessed. Notably, QTc was still assessable even when ‘lead’ misplacement was suspected.

Here, we have shown that the KM6LECG provides a remote option as an adjunct to in-clinic visits for measurement of parameters that may affect risk stratification, reflect disease progression or medication side effects, such as changes to depolarisation and repolarisation, without requiring attendance face-to-face. The accuracy, reproducibility and acceptability of the KM6LECG may also provide an opportunity for optimisation of clinical care for children commencing new medication or undergoing oncological treatment, outside the ICC setting, with potentially important resource implications, although further studies are required to confirm this. In the ICC setting, it may provide a timely modality for remote preliminary assessment of paediatric first-degree relatives referred for clinical screening, or those undergoing predictive genetic testing outside of a paediatric cardiac setting for the familial gene variant after appropriate counselling but prior to ‘in clinic’ review. Using general principles,24 the high PPV of the KM6LECG is of use in identifying those who may be at higher risk, noting that the pretest probability for an underlying disease susceptibility may be as high as 50% in some first-degree relatives. The low NPV of the KM6LECG in the subset of children with HCM where the abnormal ECG features were evident only in the precordial leads (roughly a third of this cohort), means that a normal KM6LECG cannot be used to exclude an abnormal 12-lead ECG in this group.

Limitations of this study include the small numbers for some of the ECG abnormalities being examined, and thus the choice to use ‘combined variables’ in some instances. Due to repositioning requirements, the baseline 12-lead ECG and KM6LECG were done in succession rather than concurrently; minor minute-to-minute or position-related fluctuation in ECG parameters (aside from heart rate) are not accounted for. Similarly, in follow-up, there was theoretically the chance for ECG parameters to have changed; however, this was not found to be the case for the parameters assessed. This study focused on HCM and LQTS as exemplars of ICC, so the findings may not be applicable to other populations. Further studies to evaluate the impact on clinical workflow and possible longitudinal use of this technology in those being reviewed repeatedly for screening during childhood are warranted.

Conclusion

The KM6LECG provides an accurate and reproducible method of assessing the paediatric limb lead ECG and replication of ECG repolarisation abnormalities such as those associated with paediatric HCM and LQTS, both in clinic and unsupervised at home. It is acceptable to families and may provide a useful at-home adjunct to face-to-face clinical care for some children and families, including those affected by an ICC.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by the National Health Service Research Ethics Committee, the Health Research Authority and the Great Ormond Street Hospital Research and Development team (IRAS ID 284257; Local Reference Number 20HL21). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We thank the statistical reviewer for their helpful suggestions.

References

Supplementary materials

Footnotes

  • CML and KL-W are joint first authors.

  • X @cmlawley

  • Contributors JPK, LS and EC conceptualised the study and formed the study design. EF and AB collected data. KL-W and CML undertook data extraction. CML and S-CC undertook data analysis, with interpretation supported by JPK. CML, S-CC and JPK prepared the final manuscript. All authors were involved in critically reviewing the manuscript and approved the final version to be published. JPK acts as the guarantor.

  • Funding This work was funded by a project grant from Great Ormond Street Hospital Children’s Charity to JPK (VC0721). EF and JPK are supported by Great Ormond Street Hospital Children’s Charity. EF and JPK are supported by Max’s Foundation. JPK is supported by a Medical Research Council Clinical (MRC)—National Institute for Health Research (NIHR) Clinical Academic Research Partnership (CARP) award (MR/T024062/1). This work is (partly) funded by the NIHR GOSH BRC.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.