Atrial fibrillation: real-life experience of a rhythm control with electrical cardioversion in a community hospital

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia with significant associated morbidity and mortality [1]. Its prevalence is growing along with the ageing of the population. AF is mutually connected with congestive heart failure, their co-existence being referred to as heart failure and AF dual epidemic [2]. Treatment of AF may be comprehended via the ABC pathway delineated in ESC Guidelines: anticoagulation/avoid stroke, better symptom control and detection and management of cardiovascular risk factors and concomitant disease. Better symptom control can be achieved through antiarrhythmic therapies, rate control medications, acute treatment of paroxysms with electrical and pharmacological cardioversion, and radiofrequency ablation of arrhythmia triggers and substrate [3]. Two primary strategies for the treatment of AF are rate and rhythm control. In the rate control strategy, the arrhythmic substrate and triggers per se are left unaddressed, while a safe reduction in ventricular response rate becomes the primary goal of treatment. Within the rhythm control strategy, treatment is directed at restoring and maintaining sinus rhythm. Restoration of sinus rhythm can be achieved by pharmacologic means or by electrical cardioversion - a procedure involving the application of a short burst of direct electric current to the atrial myocardium under deep sedation or general anaesthesia.

The management paradigm of AF has significantly changed over time. Introducing direct current electrical cardioversion and antiarrhythmic drugs enabled physicians to restore and maintain sinus rhythm, albeit with moderate efficacy [4]. Still, earlier randomised clinical trials, RACE, and AFFIRM, have not shown mortality benefit, which curbed enthusiasm about rhythm control for nearly two decades [5, 6]. Despite criticism of these studies with regards to the group crossover rates, cessation of anticoagulation in a substantial proportion of rhythm-control subjects, and the relatively older age of the trial population, these landmark trials have shifted priorities from arrhythmia treatment to thromboembolism prevention with anticoagulation. Randomised trial of rhythm control versus rate control of AF in patients with congestive heart failure also has not shown improvement in mortality [7]. While rate control had become the primary treatment strategy, rhythm control was still widely used for symptomatic patients [8]. Given no impact on prognosis, decisions about rate control or rhythm restoration are often driven by individual preferences and considerations of quality of life. Still, recent interest in early rhythm control strategy, particularly in younger patients with first or rare episodes of AF, reemerged with new observational and randomised data. In the GARFIELD registry, earlier cardioversion was associated with lower mortality compared to those who did not receive early cardioversion [9]. In the CASTLE-AF trial, catheter ablation for AF was shown to reduce the risk of death or hospitalisation in patients with heart failure [10]. EAST-AFNET trial had shown improved cardiovascular outcomes with rhythm control in early treatment of AF in patients with concomitant cardiovascular conditions [11]. Moreover, the advent of radiofrequency ablation and the results of new trials made rhythm control strategy more appealing again [12]. Most data on treating AF in real-world settings comes from academic centres [13], while most patients are still treated for AF within the primary care community settings. Reports on managing AF in primary care and non-academic settings are limited [14, 15] . Thus, our study aims to present data on rhythm control with electrical cardioversion in a community hospital in Russia.

A total of 1,380 records corresponding to the search criteria were extracted (see Supplementary material, Fig. S1). Of them, 130 were excluded due to no cardioversion described in the record, cardioversion or defibrillation for non-AF arrhythmia (regular supraventricular tachycardia, ventricular tachycardia, and ventricular fibrillation), and duplicate records describing the same procedure. One record was excluded due to missing a valid date of birth. Twenty two additional records describing cardioversions but not corresponding to search criteria were found during the analysis of the found records and TEE studies. All the records were analysed, and prespecified data were extracted.

In the study period, 1,272 procedures were performed on 435 patients, with a median (IQR) 1 (1–3) procedures per patient. Among 435 patients, 51% were female, with a median age of 66. Thirty seven percent had heart failure, and 48% had arterial hypertension. Data on weight and height extracted from the echocardiography database were available only in 239 patients, while echocardiographic parameters of interest were available for 56 to 134 patients due to incomplete echocardiography reports. Patient characteristics are provided in Table 1.

Most procedures (87%) were performed in outpatient settings. The median (IQR) dose of propofol was 80 (60–100) mg. In 88% of cases, a single shock was delivered, two in 6.9%, and three or more shocks in 4.9%. The median shock energy was 150 (IQR 150–150) Joules. TEE to exclude left atrial thrombus was performed before 25% of procedures. There were 1,095 (86%) cardioversions for atrial fibrillation and 177 (14%) for atrial flutter. More than half of the procedures were performed for AF of less than one-week duration. The length of the hospital stay for inpatient procedures was not reported because hospitalisations were mostly due to causes unrelated to the procedure. Inpatient settings were associated with older age, female sex, coronary heart disease and heart failure. Characteristics of cardioversions are provided in Table 2.

Rhythm control drugs, namely amiodarone, sotalol, and propafenone, were prescribed after the cardioversion procedure in 8.6%, 17%, and 1.5% of cases. Beta-blockers were prescribed after 49% of procedures, and in 22%, neither rhythm control nor rate control drugs were prescribed. Rhythm control drugs were prescribed at least once after cardioversion to 29% of patients.

Periprocedural administration of antiarrhythmic drugs was performed in 5.9% of procedures, and in most cases, it was amiodarone. Sixty-nine percent of patients were chronically anticoagulated before cardioversion, including 17% on warfarin and 52% on direct oral anticoagulants. Details are provided in Table 2.

A total of 379 TEE studies were performed before planned electrical cardioversions, with 56 of them resulting in declined cardioversions: 30 due to left atrial thrombus, 12 due to left atrial sludge or suspicion of thrombus, six due to failure to intubate, four due to spontaneous restoration of sinus rhythm and four due to other reasons.

Our data show high immediate effectiveness of direct current cardioversion similar to other studies [20, 21] with low risk of serious complications. This procedure was safe, with no serious complications in over a thousand cases. In most cases, it could be performed in outpatient settings, with patients discharged home in one or two hours. Its effectiveness is more than 90%, meaning sinus rhythm will be restored in 9 out of 10 patients. Interestingly, in all registries, we have found men were predominant in patients undergoing electrical cardioversion (73.2% in [20], 64.4% in [9], more than 70% in [22], 68% in [21]) while in our study there were almost equal numbers of men and women. Most of the procedures were performed in outpatient settings, and no procedure-related factors were associated with inpatient settings except for the duration of arrhythmia and heart failure. Hospitalisations were not related to cardioversion per se but were related to other indications, mostly heart failure and social factors. Overall, TEE was performed before nearly every fifth procedure in outpatient settings and every fourth in inpatient settings. The availability of TEE allows procedures to be performed on the day of presentation, which is more convenient for patients.

While the effectiveness of the cardioversion differed in inpatient and outpatient settings, we do not consider settings as a predictor of success not only because of confounding of age and comorbidities, which can be adjusted for, but also because failure to restore sinus rhythm can cause hospitalisation.

The association between longer duration of atrial fibrillation and cardioversion failure, although not universally observed, has been shown in some previous studies [23]. It can be explained by atrial remodelling, the presumed mechanism of atrial fibrillation self-perpetuation [3]. The greater effectiveness of cardioversion in repeat procedures compared to the first procedure has not been reported in other studies. In fact, in the RHYTHM-AF Registry data from Poland, the first cardioversion was associated with greater success [20]. A possible explanation is a selection based on some unobserved patient factors: repeated cardioversions were not performed if the first cardioversion was unsuccessful due to these factors, and a rate control strategy was pursued, whereas in patients prone to successful cardioversion procedures were repeated for subsequent arrhythmia recurrences.

A cardiologist or internist performed the procedure. No anesthesiologist was present for sedation. In Russia, this procedure is traditionally performed in inpatient settings by cardiologists and anesthesiologists, hospitalisation being a barrier to the wider use of this procedure, especially in rural areas and community hospitals, where this procedure is underused. While cardioversion service is led by nurses in some countries, [24‐26] in Russia, nurses perform only a few minor procedures independently.

Compared to other studies, fewer antiarrhythmic drugs were prescribed to patients in our study. Both in AFFIRM and EAST-AFNET 4 trials comparing rate and rhythm control strategies, significantly more than half of the patients in the rhythm-control arm were receiving antiarrhythmic drugs despite higher availability of radiofrequency ablation in the latter trial [6, 11]. In the Canadian registry, 41% of patients were on rhythm control medication and 12.3% on beta-blockers after diagnosis of AF [27]. In our study, rhythm control drugs were prescribed only after 27% of procedures, and only 29% of patients were prescribed any rhythm control drug at least once after cardioversion. The majority of the patients received no rhythm control drugs. Beta-blockers were prescribed after almost half of the procedures to prevent fast ventricular response during further paroxysms or for other indications. This seems reasonable considering readily available access to cardioversion service, controversies about antiarrhythmics conferring mortality risk, and difficulties in assessing their effectiveness when paroxysms of arrhythmia are infrequent.

Our study has several limitations. While our hospital is the only hospital in the area, some patients from neighbouring districts receive care here, and some patients from our district can be treated in other places, like tertiary hospitals in the regional capital or Moscow. Still, we are the only facility in the broader region where electrical cardioversion is available in the outpatient settings, which increases the chances that patients will be referred and followed up in our hospital. Also, our study was performed in a single center and did not represent common practice in other community hospitals in Russia. As far as we know, electrical cardioversion for AF is scarcely used in most other hospitals in our region and is rarely done in outpatient settings in this country because of reimbursement issues, logistics, tradition, or legal concerns.

An important limitation is related to the retrospective nature of our study and the use of EHR as the main data source. While EHR greatly facilitates the collection and analysis of medical data, the reliability of this data fully depends on the accuracy of physicians and nurses entering them. Templates with required fields ensure the completeness of the data, but the use of copy-paste makes it much less reliable [28]. In patients with repeated procedures, records tend to be too short, like “cardioversion performed as usual under propofol, patient discharged in sinus rhythm on same medications”.

Another limitation of our study is the absence of data on CHA(2)DS(2)-VASc and HAS-BLED scores, so we could not assess whether anticoagulation was administered according to guidelines. Data on cardioversion’s long-term effects were also not analysed in this study. While presented analyses are useful for practising physicians in assessing the immediate effects of the procedure, longer-term outcomes, like time in sinus rhythm, are also important. This study will be extended to capture the long-term effects of cardioversion.

Our study also has some strengths. It is performed in community settings, which enables it to inform practice in other community hospitals. All consecutive electrical cardioversions were included in the analysis, which makes selection bias less probable. Unlike in other AF and electrical cardioversion registries, there was an equal number of men and women in our study. It informs physicians and policymakers and directly answers the questions that can arise in patients with AF about the procedure.

While EHR provides the opportunity to perform retrospective studies and assess clinical practice in primary care and community settings, many interesting and possibly important data are missing. This can be avoided in prospective studies and dedicated registries, but these are often performed in academic centres and tertiary clinics, so they are not always generalisable. Possibly, some research framework should be established to make participation in clinical research more readily available to physicians and nurses in community settings, as it was proposed more than a century ago [29].

In conclusion, electrical cardioversion is part of AF rhythm control strategy treatment. It is an immediately effective procedure that can be safely performed in community hospitals in outpatient settings. Collecting and analysing data on performed treatments is important to inform patients during shared decision-making. This study will be supplemented by the analysis of long-term outcomes, which would add additional value to this project.