The incidence of atrial fibrillation and its impact on the length of stay following valvular heart surgery

Atrial fibrillation (AF) is one of the most commonly reported post-operative complications following cardiac surgery.¹ Despite advances in surgical techniques and post-operative care, the incidence of post-operative atrial fibrillation (POAF) remains around 31.5% following cardiac surgery.²

POAF has been associated with prolonged length of intensive care unit (ICU) and hospital stay and adverse outcomes including stroke, myocardial infarction and death.³ While advances have been made in the prediction, prevention and treatment of POAF to minimise the associated complications and adverse outcomes,^1,4 the incidence of POAF following cardiac surgery remains relatively unchanged in past decades.¹ The majority of research has focussed on patients undergoing coronary artery bypass grafting (CABG) surgery, with limited relevance to other cardiac surgery patient groups, such as those undergoing valvular heart surgery. There are no national guidelines in Aotearoa New Zealand to guide prevention and treatment of POAF following cardiac surgery. Further understanding of the incidence and effects of POAF could identify areas for future research to adapt preventative and curative interventions for patients following valvular heart surgery. The primary aim of this study was to determine the incidence of new onset POAF following valvular heart surgery and its impact on length of stay (LOS) in hospital.

Methods

This single-centre, observational, retrospective study enrolled patients admitted to the Cardiothoracic and Vascular Intensive Care Unit (CVICU) following valvular heart surgery using cardiopulmonary bypass (CPB) at Auckland City Hospital, Auckland, Aotearoa New Zealand between 1 January 2021 and 30 June 2021. Auckland City Hospital is a large tertiary metropolitan teaching hospital. The study was approved by the Auckland Health Research Ethics Committee (AH23353), with the need for informed consent waived due to the retrospective, non-interventional design of the study.

Study eligibility

Patients who were 16 years or older undergoing valvular heart surgery with or without concomitant CABG surgery were eligible for inclusion.

The collective term of valvular heart surgery referred to either repair or replacement using both mechanical and tissue valves of the aortic, mitral, tricuspid and pulmonary valves, or a combination of these procedures.

Patients were excluded based on the following criteria: history of congenital heart conditions, undergoing isolated CABG or valve-sparing aortic root replacement surgery, and any previously documented history of AF, atrial flutter or paroxysmal AF.

Study end points

The primary end points of the study were the incidence of new onset POAF and the LOS in both the ICU and overall hospital admission.

Data collection

Patients fulfilling eligibility criteria were identified through the CVICU database and patient-coded data of Auckland City Hospital and were included in the study. The occurrence of POAF was determined by the researcher through manual inspection of patient medical records for any documented arrhythmia reported by the medical team or recorded on the observation chart, or an electrocardiogram (ECG) recording a heart rhythm with undiscernible P waves and irregular RR intervals longer than or equal to 30 seconds.

Demographic data were collected, including age, gender, ethnicity and the European System for Cardiac Operative Risk Evaluation II (EuroSCORE II), which is a risk model predicting the risk of mortality after major cardiac surgery. Surgical characteristics of patients and clinical outcomes including mortality and ICU readmission were also collected. Post-operative interventions, including cardiac pacing use, and physiologic variables, including potassium and magnesium levels measured closest to the development of the arrhythmia, were extracted if routinely recorded.

The ICU LOS was determined as the time from admission into CVICU to the time that the patient was documented by the medical team as appropriate for discharge from CVICU. Where this was unavailable, the discharge time recorded on the transfer letter from ICU or co-located high dependency unit to the surgical ward was used. The overall hospital LOS was calculated from the time of admission into CVICU until the patient was discharged from the surgical ward to their home, identified from the discharge letter.

Statistical analysis

Data were collated into an Excel spreadsheet with missing data omitted. A statistician from The University of Auckland was consulted to assist with data analysis using the software R (version 4.2.0). Data were tested for normality using Shapiro–Wilk test. Mann–Whitney–Wilcoxon test was applied to all continuous variables when comparing between medians. Comparison between categorical data was performed using Pearson’s Chi-squared test or Fisher’s exact test. Bivariate analysis using simple logistic regression was conducted to examine the significance of the type of valvular heart surgery on the ICU and hospital LOS. A two-sided p-value <0.05 was determined to be of statistical significance. Descriptive data are presented as mean ± standard deviation, median (interquartile range [IQR]) or number (percentage), where appropriate.

Results

Between 1 January 2021 and 30 June 2021, a total of 428 patients were identified as having received valvular heart surgery. Of these, 311 patients were assessed for eligibility for the study (see Figure 1). A total of 120 patients were included in the data analysis.

View Figure 1–2, Table 1–3.

Overall, the median patient age was 67 years (IQR 58–74) (Table 1). Patients who were of Māori (60% vs 40%) or Asian (75% vs 25%) ethnicity were more likely to develop AF than not, but this was not statistically significant. Other baseline characteristics were similar between groups (Table 1).

Surgical characteristics were similar between those who developed AF and those who did not (Table 2). The most common type of surgery was valvular surgery in combination with CABG (n=40, 33.3%) followed by aortic valve replacement (AVR) (n=33, 27.5%) and mitral valve replacement or repair (n=24, 20%). There were no significant differences in the incidence of new onset AF following the different types of valve surgery (p=0.6), and whether the patient received amiodarone intraoperatively or not (p=1). Of those who were administered amiodarone intraoperatively, 42% (n=21) developed new onset AF post-operatively.

Primary end points

The overall incidence of new onset AF following valvular heart surgery was 42.5% (n=51), developing mostly on the second post-operative day (Figure 2).

LOS in the ICU was significantly longer in those who developed new onset AF compared with those who did not (2.8 days vs 1.2 days; p=0.002) (Table 3), as was overall hospital LOS (9 days vs 7.6 days, p=0.006). No differences were observed in mortality or readmission rate to the ICU between the two groups. The ICU LOS was 2.5 times longer in patients who had a double valvular heart surgery compared to patients who had a single AVR (p=0.033), and 3.1 times longer in patients who had a mitral valve repair surgery (p=0.026). However, there was no significant difference in the overall hospital LOS observed between all other types of valvular heart surgery (p=0.56).

Of the 51 patients that developed new onset AF post-operatively, the majority were not receiving epicardial pacing at the time, with only 3.9% (n=2) receiving atrial pacing and 7.8% (n=4) receiving dual chamber pacing. The potassium level recorded closest to the development of the patients’ first AF episode following their valvular heart surgery was median (IQR) 4.3mmol/L (0.4), and the magnesium level was median (IQR) 0.93mmol/L (0.17).

Discussion

This study found that the incidence of new onset POAF was 42.5% and was significantly associated with prolonged ICU and hospital LOS.

The overall incidence of new onset POAF identified in this study is consistent with other studies.^5–8 This is not surprising, as valvular heart surgery has been shown to substantially increase the risk of developing POAF.⁹ Furthermore, a recent study found a significantly higher incidence of POAF following valvular heart surgery in combination with CABG than a singular valvular heart surgery (p<0.001), suggesting that patients are at higher risk of developing POAF when undergoing combined procedures.⁸ This may be due to prolonged CPB and cross-clamp durations, as patients who experience longer CPB and cross-clamp times were also reported to be more likely to develop POAF.⁸ This may be explained by the systemic inflammation CPB induces, which is a contributing factor for POAF, increasing the likelihood of AF following a combined surgery compared to isolated valvular heart surgery.¹⁰ In this study, we saw no association between the type of valve operated on, CPB and cross-clamp durations and the development of new onset POAF. The inconsistent findings may be due to previous studies being conducted in different countries with different patient baseline characteristics. Turkkolu et al. conducted a study in Turkey with baseline characteristics such as reduced ejection fraction, diabetes and renal impairment, found to be independent predictors of POAF,⁸ whereas these were not observed in this study. The number of patients who underwent a combined surgery observed in the previous study (12.6%) compared to the current study (33.3%), as well as its larger sample size of 1,191 participants, may contribute to the varying results observed between studies.⁸

None of the other variables analysed in this study, such as age, lower ejection fraction, male gender and post-operative serum electrolyte levels, were associated with the development of new onset POAF to be suggestive of being possible risk factors. Post-operative hypokalaemia and hypomagnesaemia are often monitored following cardiac surgery.^4,11 A recent study determined an increased risk of developing POAF associated with a post-operative potassium level below 4.5mmol/L, whereas a magnesium level below 1mmol/L post-operatively was not associated with the increased risk of AF.¹² In our study, a median potassium level of 4.3mmol/L and median magnesium level of 0.93mmol/L measured closest to the development of AF was found, suggesting that targeting a higher level of electrolytes may play a role in reducing risks of POAF.

Previous studies have identified risk factors significantly associated with the development of new onset POAF, with advanced age a consistently identified variable, although different median ages were identified.^8,11,13,14 Conversely, the current study did not find an association between older age and development of POAF, which agrees with an earlier study that found no difference between age and the likelihood of developing new onset POAF.⁷ The differing findings may be due to studies being conducted in different countries with patients undergoing other cardiac procedures such as isolated CABG or isolated AVR surgery, whereas this study included all types of valvular heart surgery.

In this study, patients who underwent valvular heart surgery and developed new onset POAF had an ICU LOS of approximately 2 days more than those who did not, which was evident in another study.¹⁴ Furthermore, this study also demonstrated that patients who developed POAF stayed longer in hospital compared to those who remained in sinus rhythm. A significantly longer hospital LOS of approximately 3.5 days for patients who developed AF post-operatively was also indicated in a previous study.⁵ The longer LOS observed may be due to the management required to restore haemodynamic stability and to optimise the patient’s condition to revert back to sinus rhythm.¹⁴ However, the longer ICU LOS may have been confounded by bed availability on the post-operative ward, causing delays in patients being discharged from the ICU. Prolonged hospitalisation is associated with increased total cost and resource utilisation, as well as compromising patient quality of recovery with increased risks of hospital-acquired infections.¹⁵ Delayed discharges place increased demands on hospitals, causing cancellations in elective surgery, delayed treatment and repercussions for following services.¹⁶

Interestingly, despite there being no difference seen in the incidence of POAF and the type of valvular heart surgery, there was a significantly longer ICU LOS observed in patients who underwent double valve surgery in comparison to patients who had a single valve surgery. The longer ICU LOS may be due to patients having poorer outcomes with developing AF in conjunction with post-operative complications associated with valvular heart surgery—as identified by the Society of Thoracic Surgeons Adult Cardiac Surgery Database—which can include prolonged ventilation, renal failure, mediastinitis, reoperation and stroke.¹⁷ Other studies reporting the outcomes of patients undergoing double valve surgery have also shown greater hospital and long-term mortality risks when compared to single valvular heart surgery, with no mention of the impact on the LOS in hospital.^18,19

The contrasting results may be attributed to studies investigating cardiac surgeries that have included a larger proportion of patients undergoing isolated CABG procedures, varying amiodarone dosages and inconsistent timing of amiodarone administration peri-operatively.^8,13,14 With no guidelines in place at the study centre, intraoperative amiodarone was administered in this study based on the preference of the surgeon or anaesthetist, in conjunction with the condition and heart rhythm the patient was in during surgery.

As POAF remains an identified issue with an associated prolonged LOS following cardiac surgery, consideration of future studies that are more inclusive and specify the type of heart valve being operated on may advance knowledge regarding the applicability and efficacy of the potential prevention and management strategies against POAF. The incidence and burden of new onset AF following cardiac surgery may be reduced through establishing standardised guidelines that could be adopted locally and internationally. Addressing the issue of not having a national guideline in Aotearoa New Zealand to prevent and manage AF following cardiac surgery that is inclusive of all types of common cardiac procedures, especially for valvular heart surgery, may lead to a reduction in the incidence of new onset POAF, shorter LOS in hospital and improved patient outcomes.

Limitations

The retrospective nature of this study design may have introduced bias and with the observational design causal relationships cannot be established. The small sample size, as a result of the time constraints of this project, may have led to selection bias.²⁰ Due to the variables chosen for this study, other risk factors may be present that were not measured in the dataset. This exposes the study to confounding bias and may produce inconsistent findings compared to other studies that may choose different risk factors to measure. The internal validity of this study, therefore, may be compromised, as unaccounted confounding variables may potentially mislead the association the study is attempting to identify.²⁰ Furthermore, findings may not be generalisable to other settings that care for post-operative cardiac patients in Aotearoa New Zealand or other countries.

Conclusion

This single-centre study found that valvular heart surgery was associated with a high incidence of new onset POAF, and prolonged ICU and hospital LOS, particularly following double valve surgery. An introduction of a national guideline for the management of POAF could lead to a reduction in POAF and associated adverse outcomes following valvular heart surgery in Aotearoa New Zealand.