Post-operative atrial fibrillation (POAF) is a common complication following cardiac surgery in adults, occurring about 17%–34% of the time with higher risk reported following valvular surgeries [1]. Risk factors that have been associated with POAF include increased age, left atrial enlargement, mitral valve disease, congestive heart failure, and past history of atrial fibrillation [2]. Importantly, the presence of POAF has been associated with increased length of stay, stroke, in-hospital mortality and 30-day mortality [1]. Given this burden, multiple national and international guidelines have been developed to provide evidence-based recommendations on the prevention and treatment of POAF [3–5]. While there is robust data regarding POAF in the general adult population following cardiac surgery, evidence and guidelines concerning POAF in adults with congenital heart disease (ACHD) is less robust.

ACHD have an increased incidence of atrial and ventricular arrhythmias when compared to the general population, with one study finding a 22-fold increased risk [6,7]. The previously described risk factors that apply to the general public such as left atrial enlargement, mitral valve disease, and congestive heart failure are also common among this population, especially those with previous interventions. Additionally, certain congenital lesions have an even higher lifetime incidence of atrial fibrillation, including secundum atrial septal defects, atrioventricular septal defects, tetralogy of Fallot, D-transposition of the great arteries with atrial switch, and hypoplastic left heart syndrome [6].

Importantly, ACHD patients who develop atrial tachyarrhythmias are at increased risk for adverse clinical events [8]. In one study, patients with atrial arrhythmias had a 50% increase in mortality, two-fold increase in stroke or onset of heart failure, and three-fold increased risk of requiring surgical or catheter-based intervention [9]. In a cohort of patients with a history of atrial switch procedure for transposition of the great arteries, atrial arrhythmia led to decreased systemic ejection fraction and worse right ventricular function [10]. Furthermore, a study of patients with tetralogy of Fallot identified atrial arrhythmia as a predictor of mortality and subsequent ventricular tachycardia [11].

A few prior studies have evaluated POAF specifically in the ACHD population. In one study of 239 patients, age greater than 60 years, pre-operative pulmonary hypertension, mitral valve intervention, and post-operative inotropic support independently increased the risk of POAF [12]. In a second study of 1598 patients, older age, hypertension, left atrial reserve strain, right atrial dysfunction, and non-systemic atrioventricular valve regurgitation were associated with POAF [13]. The incidence of POAF in both studies was 21%. Furthermore, one study showed POAF was associated with a longer total length of stay while both studies showed POAF was not associated with mortality [12,13].

Given these limited data, this study sought to understand the incidence, risk factors, and outcomes associated with POAF in the ACHD surgical population at our institution, which includes both an adult and a children’s hospital. Given the increasing numbers of ACHD surgeries being performed at children’s hospitals and the unique physiological and demographic differences when compared to the general adult cardiac population, these data offer the chance to improve understanding of surgical risks and outcomes in POAF ACHD patients.

A single-center, retrospective cohort study was performed of adult patients (age ≥ 18 years old) who underwent congenital cardiac surgery at either the University of Colorado Hospital or Children’s Hospital of Colorado at the University of Colorado-Anschutz Medical Campus between January 2013 and December 2021. Data was extracted from the Society of Thoracic Surgery (STS) congenital surgery registry and enriched with manual chart review of individual charts. Data collected included demographics (i.e., age, body mass index (BMI), sex, ACHD anatomical classification), echocardiography reports (i.e., atrial enlargement, ventricular enlargement or dysfunction, valvular stenosis or regurgitation), historical risk factors (i.e., pre-operative oxygen use, hypertension, pulmonary hypertension, supraventricular tachycardia, coronary artery disease, thyroid disease, obstructive sleep apnea, dyslipidemia, and pacemaker use), surgical interventions (i.e., STAT (The Society of Thoracic SurgeonsEuropean Association for Cardio-Thoracic Surgery) score, prior sternotomies, intra-operative arrhythmia, bypass and cross-clamp time), surgeries performed (i.e., intervention on atrioventricular valves or atria), post-operative risk factors (i.e., pericarditis, electrolytes disturbances, inotrope use), management (i.e., post-operative medication and anticoagulant use, discharged on anticoagulation or rate or rhythm control medication), and outcomes (i.e., mortality, stroke, bleeding events, length of stay, atrial fibrillation recurrence). Patients were included if they carried a formal congenital heart disease diagnosis and underwent congenital cardiac surgery. Patients were excluded who underwent heart transplant, catheter-based procedures, or cardiac surgery not requiring bypass. POAF was defined by progress note and/or discharge summary documentation for atrial fibrillation onset during hospitalization.

ACHD anatomic classification was determined per the 2018 American Heart Association/American College of Cardiology guidelines for the Management of Adults with Congenital Heart Disease [14]. Atrial dilation was recorded as dilated vs. normal by echocardiography report. Valvular stenosis and regurgitation were considered positive if determined to be moderate or severe by echocardiography report. Ventricular dilation was determined to be significant if mild or greater by echocardiography report. Patients were determined to have a history of supraventricular tachycardia if they carried a history of any atrial arrhythmia including ectopic atrial tachycardia, atrial fibrillation, atrial flutter, atrioventricular re-entrant, and atrioventricular node re-entrant tachycardias. STAT score, which predict risk of mortality associated with surgical repair of congenital heart disease, was determined per the 2020 STS Surgical Congenital Heart Disease Mortality Categories [15]. Hypokalemia and hypomagnesemia were determined by values below the reference range of 3.5 mmol/L and 1.6 mg/dL, respectively, during the post-operative period. This study was approved by the University of Colorado Multiple Institutional Review Board (COMIRB 22-0503).

This study found POAF is a common complication following cardiac surgery in the ACHD population, occurring in 24% of our patients. Older age, history of supraventricular tachycardia, intra-operative arrhythmia, and post-operative hypokalemia were all independently associated with POAF, and patients who developed POAF had significantly longer hospital stays and increased risk of future recurrence of atrial fibrillation. Given the significant long-term outcomes of having atrial fibrillation, these data highlight the importance of identifying ways to predict, prevent, and optimally manage POAF [16].

To date, there is limited data available on the risk factors, management, and outcomes of POAF in the ACHD population. Two prior studies have sought to address this data gap and have identified an incidence of 21% for POAF, similar to our study’s incidence of 24%. These studies similarly identified older age as an independent predictor of POAF, but had not evaluated history of supraventricular tachycardia, intra-operative arrhythmia and post-operative hypokalemia which present novel risk factors for POAF. Prior identified risk factors, including hypertension, pulmonary hypertension, and certain postoperative inotropes were significant in univariate analysis but not multivariate analysis in this study. Conversely, mitral valve intervention and non-systemic atrioventricular valve regurgitation did not predict POAF in this population [12,13]. These data add important information and reiterate prior reported risk factors to help strengthen the understanding of predicting POAF in the ACHD population.

These results support prior studies findings regarding the higher incidence of POAF with older age. While the typical ACHD patient is younger than the general adult cardiac patient (average age of 30s vs. 60s), adults with congenital heart disease continue to live longer and thus it is important to recognize age as a significant risk factor [1,17,18]. Additionally, several novel risk factors were identified, including a history of supraventricular tachycardia, intra-operative arrhythmia, and post-operative hypokalemia. While this study did not distinguish between types of supraventricular tachycardia, it highlights the importance of prior history of supraventricular tachycardia when considering the odds of POAF were 7.89 times greater for those who had a history of supraventricular tachycardia in comparison to those who did not. This is consistent with prior adult literature which has shown the risk for history of atrial fibrillation causing POAF [19,20]. Specific to the ACHD population, history of paroxysmal atrial fibrillation was previously evaluated as significant in univariate analysis in one study but was not included in the subsequent multi-variate analysis [8]. Current guidelines for the management of atrial fibrillation include a 2a recommendation for the short-term prophylactic use of beta blockers or amiodarone in patients at high-risk for POAF [21,22]. Given these findings, further studies are needed to assess the indications for preventative medications in ACHD patients with a history of supraventricular tachycardia.

Regarding intra-operative arrhythmia, prior studies have not evaluated this variable and its significance may help highlight patients who need closer monitoring for POAF after leaving the operating room. Furthermore, these findings may highlight patients who may benefit most from prophylactic post-operative antiarrhythmic therapy. Finally, post-operative hypokalemia was found to be significantly associated with POAF in multivariate analysis. Many centers, including our own, have electrolyte repletion protocols to minimize electrolyte abnormalities; however, these protocols are typically responsive to lab values and do not completely eliminate electrolyte imbalances. While no prior study has assessed this factor in ACHD patients, given the known increased risk of atrial fibrillation due to hypokalemia, this identifies an important modifiable risk factor which may significantly impact rates of POAF and may warrant more aggressive or prophylactic treatment in the post-operative period [23].

While not significant in multivariate analysis, the significance of pre-operative oxygen use in univariate analysis may indicate the potential role of pre-operative hypoxemia in causing POAF. In the general adult surgery population, hypoxia has been described as contributing to POAF through ischemia of myocardial atria cells increasing ectopy and reducing atrial refractoriness [24–26]. It is plausible that pre-operative hypoxemia may increase the risk of POAF in the ACHD population through a similar mechanism. Given the significant altitude of our center, it is possible that the elevation may have led to a higher incidence of pre-operative hypoxemia than similar centers at sea level.

Conversely, there were several variables that were associated with POAF in prior studies that were not reproduced in this study. Hypertension and pulmonary hypertension were significant in univariate analysis but not multivariate analysis, though this may be limited by the low power of the study. As hypertension is a known risk factor for both atrial fibrillation and POAF in the general population, it is plausible that this variable is associated with POAF in this population and should be studied further [27,28]. A similar connection between pulmonary hypertension and POAF has been described in the general population [29]. Additionally, non-systemic valve regurgitation was positive in prior studies but was not positive in univariate analysis in our study. As non-systemic valve regurgitation frequently accompanied pulmonary hypertension, is it possible that these variables are inter-related [30].

Of note, left atrial enlargement was predictive of atrial fibrillation in univariate analysis in this study but not multi-variate analysis. However, left atrial enlargement is a prior described risk factor for atrial fibrillation in the general public [31–33]. Proposed mechanisms include disturbed impulse propagation caused by alterations in the functional and structure of the atria, including increased atrial stretch creating heterogeneous refractory periods within the atria which predisposes to atrial fibrillation [31,33,34]. Furthermore, studies have shown that residual left atrial enlargement is a risk factor for inability to maintain sinus rhythm after successful cardioversion as well as the need for repeat ablation of atrial fibrillation [33,35]. Therefore, it is reasonable that left atrial enlargement would precipitate atrial fibrillation in the ACHD population and may be significant in larger sample sizes.

Overall, as a class post-operative inotropes were not predictive of developing POAF; however, specific inotropes, including norepinephrine, epinephrine, and vasopressin, were associated in developing POAF. While Brock et al. suggested a significant impact of post-operative inotropes on POAF, it was unclear which type of inotropes were used [12]. Therefore, these data suggest further investigation is needed to determine if specific classes of inotropes are more likely to contribute to the development of POAF.

Our data shows a similar finding to other studies in which POAF is not associated with increased mortality, stroke, or bleeding events. While in the general population, POAF is associated with increased risk of stroke and mortality, the much lower prevalence of adults with congenital heart disease undergoing cardiac surgery limits our ability to study this potential complication [1]. Similar to Brock et al., we found an increased length of stay for patients with POAF [12]. As increased length of stay can be associated with increased complications, financial burden, and morbidity, these data emphasize the importance of close surveillance throughout the hospitalization [36]. A prior meta-analysis in the general adult cardiac population showed the impact of prophylactic pharmacological and non-pharmacological intervention on preventing post-operative atrial fibrillation and reduced hospital length of stay [37]. As the incidence of POAF is quite high in the ACHD population, these data suggest the importance of future studies evaluating methods to prevent POAF and the potential of inpatient POAF protocols in institutions treating ACHD patients [12,13].

There were several limitations of our study. First, the small sample size may preclude the ability to determine the significance of several variables. Additionally, this study was conducted at a single center. As each study conducted to date has been a single-center study, there may be factors that are significant only to specific centers and regional variability which may confound the identification of nationwide risk factors. While each patient chart was rigorously assessed, relying on documentation alone may introduce error into the identification of atrial fibrillation.

Atrial fibrillation is a common complication after cardiac surgery in the ACHD population and is associated with prolonged length of hospital stay and increased risk of atrial fibrillation recurrence. Several risk factors, including older age, history of supraventricular tachycardia, intra-operative tachycardia, and post-operative hypokalemia were predictive of POAF in multivariable analysis. These results reinforce the importance of close surveillance in the postoperative period, and given the high incidence of POAF in this population, argue for further pre-operative risk assessment tools as well as inpatient POAF management protocols.