Feasibility of Patient-Managed ECG Recordings to Detect the Time of Atrial Fibrillation Recurrence after Electrical Cardioversion: Results from the PRE-ELECTRIC Study

Background: Electrical cardioversion (ECV) is a common procedure to terminate persistent atrial fibrillation (AF). The recurrence rate is high, and the patients often fail to recognize AF recurrence. Objectives: The aim of the study was to evaluate the feasibility of patient-managed electrocardiography (ECG) to detect the time to AF recurrence after ECV. Methods: PRE-ELECTRIC (predictors for recurrence of atrial fibrillation after electrical cardioversion) is a prospective, observational study. Patients ≥18 years of age scheduled for ECV of persistent AF at Bærum Hospital were eligible for inclusion in the study. Time to recurrence of AF was detected by thumb ECG, recorded twice daily and whenever experiencing symptoms. The observation period was 28 days. We defined adherence as the observed number of days with ECG recordings divided by the expected number of days with ECG recordings. Study personnel contacted the participants by phone to assess their awareness of AF recurrence after a recurrence was detected in the thumb ECG. Results: The study enrolled 200 patients scheduled for ECV of persistent AF at Bærum Hospital between 2018 and 2022. The mean age was 66.2 ± 9.3 years, and 21.0% (42/200) were women. The most frequent comorbidities were hypertension (n = 94, 47.0%) and heart failure (n = 51, 25.5%). A total of 164 participants underwent ECV of AF. The procedure was initially successful in 90.9%, of which 50.3% had a recurrence of AF within 4 weeks. The median time to recurrence was 5 days. Among the cardioverted participants, 123 (75.0%) had no missing days of thumb ECG recording during the observation period, and 97.0% had ≤3 missing days. More than a third (37.3%) of the participants with AF recurrence were unaware of the recurrence at the time of contact. Women were older and more symptomatic than men but had similar outcomes after ECV. Conclusions: Recurrence of AF after ECV was common. Using patient-managed thumb ECG was a feasible method to detect AF recurrence following ECV. Further studies are needed to investigate whether patient-managed ECG after ECV can optimize AF treatment.

of days with ECG recordings divided by the expected number of days with ECG recordings. Study personnel contacted the participants by phone to assess their awareness of AF recurrence after a recurrence was detected in the thumb ECG. Results: The study enrolled 200 patients scheduled for ECV of persistent AF at Baerum Hospital between 2018 and 2022. The mean age was 66.2 ± 9.3 years, and 21.0% (42/200) were women. The most frequent comorbidities were hypertension (n = 94, 47.0%) and heart failure (n = 51, 25.5%). A total of 164 participants underwent ECV of AF. The procedure was initially successful in 90.9%, of which 50.3% had a recurrence of AF within 4 weeks. The median time to recurrence was 5 days. Among the cardioverted participants, 123 (75.0%) had no missing days of thumb ECG recording during the observation period, and 97.0% had ≤3 missing days. More than a third (37.3%) of the participants

Introduction
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia. Electrical cardioversion (ECV) is a method to restore sinus rhythm to reduce AF-related symptoms or evaluate if symptoms are AFrelated [1]. Recent research supports early rhythm control to improve cardiovascular outcomes [2,3]. More than 90% of patients achieve sinus rhythm immediately after an ECV, but 30-64% experience a recurrence of AF within a few weeks [4][5][6][7], leaving the treatment of patients with persistent AF a challenge. Assessment of symptoms, quality of life, and consideration of rate versus rhythm control is part of the recommended ABC pathway of AF treatment [1]. However, AF-related symptoms are variable, and symptom rhythm correlation is generally poor [6,7,8,9]. We aimed to investigate the feasibility of patientmanaged out-of-hospital electrocardiography (ECG) to detect AF recurrence after ECV.

Materials and Methods
PRE-ELECTRIC (predictors for recurrence of atrial fibrillation after electrical cardioversion) is a prospective, observational study. The study enrolled patients ≥18 years of age scheduled for ECV of persistent AF between 2018 and 2022 at Baerum Hospital, Gjettum, Norway. There were no exclusion criteria. The study was approved by the Regional Committees for Medical and Health Research Ethics in Norway (ref. 2017/1604), and all participants signed informed consent following the Helsinki Declaration. ECV was performed according to clinical guidelines [1]. The ECV was defined as successful if the patient had sinus rhythm in an ECG 2-4 h after the procedure. The participants underwent clinical examination and 12lead ECG recording the week before ECV, 3-6 h after ECV, and 4-6 weeks after ECV. During the first visit, a physician classified the participants' AF symptoms according to the modified European Heart Rhythm Association (mEHRA) score. The questionnaire "The Symptom Checklist: Frequency and Severity Scale" (SCL) measures the patient's perception of the AF symptoms [10], and the participants completed the questionnaire at the visit before ECV and the end of the observation period. All participants were instructed to record at-home ECGs using a Zenicor ® thumb ECG twice a day and when experiencing symptoms during 28 days after the ECV. The participants continued with ECG recordings after a documented recurrence of AF. We classified AF as persistent if the ECGs showed AF in all recordings ≥7 consecutive days. Participants with a history of persistent AF needing new cardioversion (electrical or pharmacological) before 7 days of recurrent AF duration were also classified as having persistent AF. We defined adherence as the number of observed days with ECG recordings divided by the number of expected days with ECG recordings. The time to AF recurrence was documented by the thumb ECG. The study physicians routinely evaluated the ECGs 2-3 times per week and whenever contacted by the participants. Study personnel normally contacted the participants within 3 days after a detected AF recurrence. If there was uncertainty about the rhythm status, we invited the participant to the hospital for a 12-lead ECG.

Statistical Analysis
Continuous variables are presented as mean ± standard deviation (SD), and differences between groups are examined for statistical significance using a t test for two independent samples. Categorical variables are presented as frequencies (percent), and differences between groups are compared using χ 2 or Fisher's exact test when appropriate. Missing SCL data were replaced by person mean. Kaplan-Meier survival curves were compared using the logrank test. We used STATA 17 (StataCorp) for statistical analyses.

Results
A total of 200 patients were included in the study. Baseline characteristics are shown in Table 1. The mean age was 66.2 ± 9.3 years, and 42 (21%) participants were women. The most frequent comorbidities were hypertension (n = 94, 47.0%), heart failure (n = 51, 25.5%), and ischemic heart disease (n = 21, 10.5%). The average body mass index was 27.8 ± 4.5 kg/m 2 , and 146 (73%) participants were overweight or obese. Women were older and reported more frequent and severe symptoms than men. Twenty-five participants spontaneously converted to sinus rhythm before ECV. Eight participants presented with an atrial flutter at the first visit and were excluded from further analyses (Fig. 1).
We scheduled the baseline visit for the week before ECV. However, 13 participants had a postponed ECV due to heart failure (n = 3), high digoxin level (n = 1), missed dose(s) of oral anticoagulants (n = 5), electrolyte imbalance (n = 1), patient request (n = 1), and unknown (n = 2). The most extended ECV postponement was 148 days due to treatment of heart failure.
Among the 164 participants with an ECV of AF, 123 participants (75.0%) recorded thumb ECGs every day during the observation period, and 97.0% had ≤3 missed days. All the participants were capable of using the thumb ECG device.
Fifteen participants were not successfully cardioverted; five never achieved sinus rhythm, and ten presented with a recurrent AF at the ECG 2-4 h after the procedure. ECV was successful for 149 participants; however, 75 (50.3%) had a recurrence of AF during the observation period. Most AF recurrences were persistent (n = 68, 90.7%). The median time to recurrence of AF was 5 days, and 97.3% of the recurrences of AF occurred within 17 days (Fig. 2b). Day 3 had the highest number of AF recurrences, and 37.3% of the recurrences had manifested by this time. Men and women had similar outcomes after ECV (Fig. 2a). Twenty-eight of 75 participants (37.3%) were unaware of the recurrence of AF when the study personnel contacted them.

Discussion
We found that using a thumb ECG was feasible to detect the recurrence of AF after ECV. AF recurrence occurred in 50.3% of the successfully cardioverted participants, and the Data are presented as mean (±SD) or n (%). AF, atrial fibrillation; mEHRA, modified European Heart Rhythm Association symptom classification for AF; CHA 2 DS 2 -VASc, congestive heart failure, hypertension, age ≥75 (doubled), diabetes mellitus, stroke (doubled), vascular disease, age 65-74, and sex category (female); AAD, antiarrhythmic drug. 349 median time to recurrence was 5 days. More than a third of the patients were unaware of their early recurrence of AF at the time of contact. Women were older and more symptomatic than men but had similar outcomes after ECV.
The reported AF recurrence rate after ECV of~50% is in line with previous studies [6,7,11]. Hellman et al. [4] and Pluymaekers et al. [5] reported the median time to AF recurrence to be 8 and 12 days, respectively, whereas we found the median time to be 5 days. The accuracy of determining the time to recurrence varies according to the method used, and ECG recorded after the presentation of symptoms may underdiagnose or delay the diagnosis of AF recurrence. All patients in our study participated actively by using the thumb ECG, which may have increased the accuracy of AF recurrence. Other studyspecific variations in the underlying cause, comorbidity, duration, and treatment of AF can make it difficult to compare the recurrence of AF across different studies [12]. In addition, differences in antiarrhythmic drug regimens for ECV are likely to impact both AF recurrence rate and time to recurrence [11].
Symptom assessment is a part of the ESC guidelines to improve quality of life and is essential to consider when choosing a rhythm or a rate control strategy [1]. However, identifying AF-related symptoms may be difficult as symptoms vary between and within the patient, and factors like the patient's emotion or mood [13] or the number or burden of the underlying comorbidities may play a role [7]. As described by Hermans et al. [6,7], we found that many of our study participants were unaware of their AF recurrence. The high proportion of patients unaware of an AF recurrence may have important clinical implications. First, in patients with strong indications for rhythm control, early detection of AF recurrence may allow for a prompt second ECV and effective antiarrhythmic medication, which may improve rhythm outcomes. Second, early detection of recurrent AF in unaware patients with non-specific symptoms suggests that symptoms are unrelated to AF and may support an alternative rate control strategy. Third, patients with low stroke risk may be recommended to continue oral anticoagulation if they are eligible for a further rhythm control strategy.
The few women in our study may show a tendency toward earlier findings of possible sex-specific differences in AF treatment [14]. However, the female group in our study is small, and the generalizability is limited. More studies are needed to assess if and how the use of patient-managed ECG can improve personalized AF management.

Limitations
The study was performed in a single-center study, and the sample size was relatively small. Most of the study population were men of European ancestry, which limits the generalizability of the results. The symptoms in the "Symptom Checklist: Frequency and Severity Scale" were reported by the participants and may have been affected by recall bias. Study participation may have affected the high adherence rate, and the adherence may be lower in a real-world setting. Finally, ECG recordings by thumb ECG twice a day may not detect all episodes with recurrent paroxysmal AF, leading to misclassification.

Conclusion
Recurrence of AF after ECV is common. Using a digital thumb ECG was feasible to detect the time of AF recurrence. More than one-third of the participants with AF recurrence in our study were not aware of their AF recurrence. Further studies are needed to investigate if patient-managed ECG after ECV may improve personalized AF management.

Data Availability Statement
The data supporting this study's findings are not publicly available due to the original patient consent and national privacy regulations (including GDPR). Any requests for data sharing will be handled according to the regulation by the Data Protection Officer at Vestre Viken Hospital Trust. Further inquiries can be directed to the corresponding author.