Electrocardiogram (ECG) is an essential tool for diagnoses and monitoring cardiac arrhythmias in veterinary medicine. However, a low-cost, user friendly, 6-leads, smartphone-based electrocardiograph is still lacking.

D-Heart® is a portable, smartphone-based device, which streams the tracing via Bluetooth to any smartphone, enabling the acquisition of an ECG on multiple leads, currently used in human cardiology.

The aim of this study was to determine the accuracy of D-Heart® tracings, compared with the gold standard non-portable 6-lead electrocardiograph in dogs.

Standard 6-lead ECGs and D-Heart® ECGs were subsequently acquired in conscious dogs in right lateral recumbency. Each tracing was saved with an anonymous ID. Two experienced observers, independent and impartial, performed data analysis. In case of disagreement, a third impartial observer adjudicated the tracing.

The sample was described, for continuous variables, by mean and standard deviation or median and interquartile range (IQR), when appropriate, and for categorical variable by proportion. The concordance between the ECG methods was assessed by weighted kw-Cohen index, with its relative significance, taking as the end point variable the standard 6-lead ECG group. The Bland-Altman method with a 95% confidence level was applied for P, PR, QRS, T, and QT interval measurements. Since differences between the two measurements did not follow a normal distribution, a non-parametric approach (median value and IQR) was used to determine the limits of agreement. P values were considered significant at the 0.05 level. Amplitude of the waves was not considered in this preliminary analysis because currently the software doesn't allow voltage variation.

One hundred fifteen dogs of different weights and breeds admitted to the cardiology service of DIMEVET were enrolled. Mean age of dogs was 7,5±4 years. Most were intact males (45%, n=51). The most represented breed were mongrels (27%, n=32).

Weighted Cohen's kappa (kw) test between ECGs demonstrated an excellent concordance of 0.989 (p<0.001) in the evaluation of the heart rhythm, of 0.991 (p<0,001) for ST segment morphology and of 0.838 (p=0.040) for T wave morphology. There was a 100% concordance among P morphology assignment. P, PR, QRS, T, and QT intervals comparison with Bland-Altman method showed an extremely good concordance for D-Heart® measurements (95% limit of agreement ±0.9 ms for P, ±10 ms for PR, ±35 ms for QRS, ±5 ms for T wave). Less concordance resulted for QT (±80 ms).

D-Heart® proved effective accurate recording of ECG comparable to standard 6-lead ECG, opening new perspectives to improve diagnostic tools in veterinary cardiology.

Disclosures

Disclosures to report.

This project was funded by D-Heart. NiccolÁ Maurizi and NicolÁ Briante are the co-founders of the social innovative start-up D-Heart.