Accuracy of Cardiac Arrhythmias and Conduction Disorders Diagnosis Using a Smartwatch

Overview

Implantable devices increase the detection of silent atrial fibrillation in high-risk populations, being a useful tool in the early diagnosis of these patients, allowing monitoring for a longer time. Currently, these devices can be used for monitoring and possible early diagnosis of abnormalities in order to prevent health complications and bring about better outcomes, as well as monitor the user's progress and thus prevent more serious developments. Objective: to analyze, in patients with cardiovascular disease, the electrocardiographic aspects detected from the device in comparison to the conventional electrocardiogram, identifying aspects of similarity between the tracings obtained and automatic diagnostic accuracy by the device and by the visual assessment of the tracing. Method: a cross-sectional observational study of diagnostic accuracy will be carried out at the University Hospital of the Federal University of Piaui. The following parameters were considered: significance level of 95%, test power of 80%, value of the disease diagnosis rate of 84%, area under the receiver operating characteristic curve of 94% and an equal proportion of the number of positive and negative cases. The minimum sample size consisted of 100 patients, and an additional 10% will be added to compensate for possible losses. A total of 110 patients will be invited to participate in the study. Patients aged ≥ 18 years will be included, with the presence of one or more diagnoses below: systemic arterial hypertension, heart failure, valvular disease, coronary atherosclerotic disease, myocardial infarction, endocarditis, myocarditis, diabetes, cardiac arrhythmias and patients with cardiac pace or implantable automatic defibrillator. The variables analyzed will be: clinical-demographic data of the patients, including age, sex, race, socioeconomic profile; comorbidities and electrocardiographic findings as measured by the clock and conventional ECG. The data will be analyzed comparatively between the groups: 1 = EKG on the clock; 2 = 12-lead conventional EKG and DI (single lead). The primary outcome will be the positive diagnostic correlation of the data obtained in group 1 in relation to group 2. Secondary outcomes: successful measurement of the electrocardiographic record by the clock; diagnostic accuracy of atrial fibrillation and other heart rhythm disorders.

The main objective of this study is to analyze, in patients with cardiovascular disease, the electrocardiographic aspects detected from the smartwatch device in comparison to the conventional electrocardiogram, identifying aspects of similarity between the tracings obtained and automatic diagnostic accuracy by the device and by visual assessment. of the layout. The secondary objectives will be to evaluate in the same population: aspects of practicality for the use of smart watches in clinical practice in cardiology; the success rate in obtaining the electrocardiographic tracing with adequate quality by the smart watch and describe the types of failure and assess the clinical characteristics related to the failure in obtaining the electrocardiographic tracing by the smart watch. Method: An observational, transversal, study of diagnostic accuracy will be carried out at the University Hospital of the Federal University of Piaui. The sample size was calculated using MedCalc Statistical Software version 19.2.3 (MedCalc Software Ltd, Ostend, Belgium). For the sample calculation, the following parameters were considered: significance level of 95%, test power of 80%, value of the disease diagnosis rate of 84%, area under the receiver operating characteristic curve of 94% and equal proportion of number of positive and negative cases. The minimum sample size consisted of 100 patients, and an additional 10% will be added to compensate for possible losses. A total of 110 patients will be invited to participate in the study. Inclusion criteria: • Age ≥ 18 years; • Patients followed up in the cardiology sector, hospitalized or outpatient, with the presence of one or more of the following diagnoses: systemic arterial hypertension, heart failure, valve disease, coronary atherosclerotic disease, myocardial infarction, endocarditis, myocarditis, diabetes, cardiac arrhythmias and patients with cardiac pacemakers or implantable automatic defibrillators. Exclusion criteria: • Patients who have electrocardiographic tracings, whether obtained from the watch or the conventional method, of poor quality that make it impossible or that makes the diagnosis difficult due to the technical aspect of the performance. • Withdrawal of the consent form at the request of the patient or guardian. Primary outcome: • Positive diagnostic correlation of the data obtained in group 1 (clock) in relation to group 2 (conventional ECG 12 leads and DI - single lead). Secondary outcome: • Successful verification of the electrocardiographic record by the watch; • Diagnostic accuracy of atrial fibrillation; • Diagnostic accuracy of heart rhythm disorders, excluding atrial fibrillation. Ethical aspects: All patients will be informed about the principles and objectives of the protocol before inclusion in the study and will be given explanations of potential advantages or disadvantages. All procedures involved in this study will be in accordance with the Helsinki Declaration of 1975, updated in 2013. The study will follow the principles of good clinical practice and the terms of Resolution No. 466/12 of December 12, 2012 of the National Health Council Ministry of Health. All patients or legal guardians will sign the informed consent form in writing before selection. Patients will be guaranteed that early withdrawal from the study will always be possible, without any consequence or impairment in their subsequent medical care. Data will be collected only after approval by the ethics and research committee. Those who do not sign the Informed Consent will not be included. Data collect The variables analyzed will be: clinical-demographic data of the patients, including age, sex, race, socioeconomic profile; comorbidities and electrocardiographic findings as measured by the clock and conventional ECG. The data will be analyzed comparatively between the groups: 1 = ECG on the clock; 2 = 12-lead conventional ECG and DI (single lead). Electrocardiographic recording devices The smart watch used will be the Apple Watch series 4, from Apple Inc. (Specifications: Sensor I, 511 Hz, mobile operating system 13.1.3, operating system 6.0.1, Watch 4.2). For the recording of the conventional electrocardiogram, a mobile device will be used that allows the recording of 12 simultaneous leads and the rhythm tracing of the DI lead. Criteria for electrocardiographic diagnosis The criteria for electrocardiographic diagnosis of morphological aspects, intervals and conclusive report will be based on the guideline of the Brazilian Society of Cardiology on the analysis and issuance of electrocardiographic reports. For the analysis of similarity between the traces obtained by the clock and the conventional ECG, the parameters of regularity of the RR cycle beat by beat, heart rate in beats per minute, amplitude and duration of the p wave, QRS and T wave will be evaluated, in addition to the PR and QT corrected intervals. The unit in milliseconds will be used for the duration and millimeters for the amplitude. To obtain the automatic electrocardiographic diagnosis of the clock, two ECG tracings of 30 seconds will be obtained, in a calm environment, with the patient at rest, in the horizontal supine position. The device will be attached to the wrist on the left side with the use of the fingers of the right hand on the sensor button of the watch to complete the electrocardiographic DI derivation following the Einthoven triangle derivation criteria. The automatic diagnosis obtained must be the same in both plots to be validated. A third trace will be performed if the automatic diagnosis of the second trace is different from the first analysis. If there is a new divergence in the automatic diagnosis of the clock's third ECG trace in relation to the two previously obtained traces, the diagnosis will be considered inconclusive. The specific results of the automatic diagnosis obtained by the watch will be: low beats with FC≤40 beats per minute (bpm); high beats with HR≥120bpm; sinus rhythm when interpreted as normal by the clock, atrial fibrillation, inconclusive and does not allow analysis due to poor technical quality and did not allow tracing. The conventional ECG tracing will be obtained in the 12 leads of the frontal plane (DI, DII, DIII, aVR, AVL, AVF) and the horizontal plane (V1 to V6) and a 30-second rhythm trace in the DI lead. The diagnostic results of conventional ECG tracings will be: normal or abnormal; sinus rhythm, AF, atrial flutter or other non-sinus rhythm; intraventricular conduction disorder: left bundle branch block (BRE), right bundle branch block (BRD), right posteroinferior lower block (BDPI or left anterior superior (BDAS), isolated or associated; normal electrical axis, shifted to the right or shifted to the right) left, inconclusive and does not allow analysis due to poor technical quality. The electrocardiographic tracings will be analyzed blindly by two independent cardiologists with a cardiology degree from Brazilian Medical Association and Brazilian Society of Cardiology and not participating in the research data collection stage. Statistical analysis: The two independent cardiologists will be evaluated for calibration using the Kappa method of agreement. If there is disagreement regarding the diagnosis of the two cardiologists, the divergent path in question will be evaluated by the main researcher of the research. The normal distribution of variables will be verified using the Shapiro-Wilk test, where continuous variables with normal distribution will be expressed as means and standard deviation (SD) using the chi-square test for comparison between groups. Variables with non-Gaussian distribution will be expressed as medians and interquartile ranges using Mann-Whitney tests for comparison between groups. For statistical significance, p \<0.05 and 95% confidence interval will be considered. The analyzes will be performed using the Statistical Package for the Social Sciences (SPSS®, version 22.0 for Windows®).