Postoperative atrial fibrillation is the most common dysrhythmia to occur after coronary artery bypass graft surgery. It develops in 10% to 40% of patients and can lead to complications such as hemodynamic instability, heart failure, and stroke. Risk factors include hypertension, diabetes, chronic kidney disease, and obesity. Patients who experience postoperative atrial fibrillation often have longer hospital stays, are at higher risk for readmission, and have increased mortality. Protocols designed to reduce the incidence of the condition can decrease hospital costs, improve patient outcomes, and increase overall quality of care. This quality improvement project took place in a tertiary care center located in southeastern Michigan and focused on the development and implementation of an evidence-based postoperative atrial fibrillation prophylaxis protocol using amiodarone. The outcomes of this project suggest that amiodarone prophylaxis can reduce the incidence of postoperative atrial fibrillation in patients with no previous history of atrial fibrillation undergoing coronary artery bypass graft surgery.
Atrial fibrillation (AF) is a type of heart dysrhythmia characterized by a rapid, irregular beating, or "fibrillation," of the heart’s upper chambers.1 Atrial fibrillation disrupts normal blood flow through the heart by not allowing the atria time to completely empty their volume into the ventricles. The remaining blood swirls in the atria and can clot over time, putting the patient at risk for ischemic stroke. According to the Centers for Disease Control and Prevention, AF is responsible for 15% to 20% of ischemic strokes in the United States.1
Postoperative atrial fibrillation (POAF) is defined as an episode of AF that occurs after a surgical procedure. The condition can develop in patients without a previous history of AF and is considered the most common dysrhythmia to occur after coronary artery bypass graft (CABG) surgery, affecting 10% to 40% of patients.2–4 According to Thanavaro and colleagues,5 94% of POAF episodes occur within 7 days of surgery, with the highest incidence on postoperative day 2 or 3. Although POAF is often transient, it can persist for weeks after surgery, increasing the risk for further complications such as heart failure and stroke that could lead to increased length of stay (LOS), higher costs, hospital readmissions, and increased mortality.6,7
Factors Associated With Postoperative Atrial Fibrillation
Preoperative Risk Factors
Although the exact etiology of POAF is unknown, the incidence is highest among patients undergoing both CABG and valve surgery together. Male sex and advanced age have also been directly linked to increased incidence.6–8 The sex link to POAF is likely due to the higher number of men than women undergoing CABG surgery.9 However, the age link to POAF is not entirely understood.
Several other preoperative risk factors have been identified. For instance, patients with a CHA2DS2-VASc (Congestive heart failure [CHF], Hypertension, Age, Diabetes, Stroke/transient ischemic attack–Vascular disease) score of 2 or greater have a higher incidence of POAF.4,7 This higher risk is thought to be secondary to the comorbidities measured by the CHA2DS2-VASc score. These comorbidities are strongly associated with ventricular remodeling, diastolic dysfunction, and left atrial enlargement, which are known risk factors for POAF.7 Obesity, defined as a body mass index (BMI) greater than 30 kg/m2 , has also been linked to increased incidence of POAF.4,10 In addition, chronic obstructive pulmonary disease (COPD) and chronic kidney disease4 (CKD) are major risk factors.11,12 β-Blocker withdrawal, as when β-blockers are held in the immediate postoperative period because of hypotension, increases the risk of developing POAF by more than 50% after CABG surgery.2 This increased risk may be due to the existing comorbidities requiring treatment with antiplatelet therapy (ie, myocardial infarction, percutaneous coronary intervention) or to adverse reactions to anti-platelet therapy, such as thrombocytopenia and bleeding. Anemia associated with acute blood loss and blood transfusions are known risk factors for POAF.4 Furthermore, platelet activation itself is a prodysrhythmic process.13
Intraoperative and Postoperative Risk Factors
In addition to preoperative risk factors, both intraoperative and postoperative risk factors can lead to POAF. Intraoperative risk factors are often multifactorial. Inadequate atrial cooling, atrial ischemia, and local inflammation of cardiac structures can create uneven changes in the atrial refractory period, resulting in POAF.11,14,15 After surgery, systemic inflammatory response or sepsis can trigger numerous pathways to POAF that include elevations in C-reactive protein, electrolyte abnormalities, and lactic acidosis.4,11 Airway complications from bronchospasm, pneumonia, pleural effusion, pulmonary embolus, and pneumothorax often result in poor ventilation and perfusion, right ventricular strain, hypoxia, and hypoxemia, leading to secondary POAF.8,16 Increased myocardial stress from catecholamine and inotropic infusions, physiological response to pain, hypervolemia or hypovolemia, pericardial inflammation, and pericardial effusion can also contribute to POAF.8,11,14–16
Complications and Associated Costs
Complications of POAF include hemodynamic instability, discomfort, anxiety, cognitive impairment, thromboembolic events such as stroke, and heart failure.12,17 Increased hospital LOS due to complications of POAF may result in costs of more than $10 000 per patient.12,17 The average increase in hospital LOS due to POAF is 2 to 4 days beyond the CABG diagnosis-related group reimbursable LOS of 5 to 7 days.8,12 Furthermore, POAF after CABG surgery puts patients at higher risk of readmission within 30 days of discharge, frequently due to dysrhythmias or recurrent AF. In general, hospital readmissions within 30 days of discharge account for more than $41.3 billion of uncompensated care nationwide.18
In addition to increased costs, patients with POAF have increased mortality. Following surgery, all categories of mortality (in-hospital, 6-month, 2-year, and long-term) are higher in patients with POAF than in those without POAF.12,19,20 Mortality is generally higher in women than in men.21
There is no question that the development of POAF can be a serious and expensive complication with long-term consequences. According to LaPar and colleagues,22 "Protocols designed to reduce the incidence of POAF have the potential to significantly impact patient outcomes and the delivery of high-quality, cost-effective patient care." Therefore, a quality improvement (QI) project was initiated at a tertiary care center in southeastern Michigan with the goal of developing and implementing an evidence-based POAF prophylaxis protocol to reduce the incidence of new-onset atrial fibrillation after CABG surgery. As a preliminary step, we performed a literature review to search for specific protocols, research studies, and guidelines on the prevention of POAF in CABG surgery patients.
The benefits of amiodarone for the preven tion of POAF have been widely studied for more than 20 years. Amiodarone has β-adrenergic effects as well as sodium, potassium, and calcium channel blocking abilities and is used to manage atrial and ventricular dysrhythmias. It is the only antidysrhythmic agent that slows cardiac conduction by working on all 4 phases of the cardiac cycle.8,15
Our literature search revealed several meta-analyses and 2 guidelines on the use of amiodarone for POAF prophylaxis. Meta-analyses conducted by Aasbo et al23 and Bagshaw et al6 documented evidence that amiodarone significantly reduced the incidence of POAF. Both meta-analyses suggested that amiodarone significantly reduced the incidence of stroke, hospital costs, and hospital LOS.
In meta-analyses completed by Buckley et al17 and Chatterjee et al,3 randomized controlled trials related to route, timing, and dose-response relationships were evaluated. The bioavailability of oral amiodarone ranges from 30% to 70%, with a mean weighted average of 48%, which was used to determine amiodarone prophylaxis dosing.17 Buck-ley et al17 found that prophylactic doses of amiodarone between 3000 and 5000 mg led to a significant decrease in the incidence of POAF, doses less than 3000 mg were associated with a somewhat higher incidence of POAF, and doses greater than 5000 mg did not demonstrate a significant benefit but were associated with a higher incidence of adverse effects. No significant differences have been found in outcomes according to the timing of amiodarone prophylaxis (preoperative, intraoperative, or postoperative)3,17,23 or according to whether amiodarone was administered orally or intravenously (IV).3,17
Patel and colleagues24 reviewed 18 clinical trials that specifically evaluated adverse outcomes. This meta-analysis revealed that amiodarone prophylaxis increased the risk of developing symptomatic bradycardia by 70% and symptomatic hypotension by 62%. The incidence of adverse effects (eg, bradycardia, hypotension) was greatest among patients receiving IV amiodarone and among those receiving oral or IV amiodarone doses greater than 1000 mg in a 24-hour period.3,24 Currently, there are no clinical practice guidelines specific to the prophylactic use of amiodarone for prevention of POAF in the cardiac surgical patient. The American College of Cardiology, American Heart Association, and Heart Rhythm Society 2014 guidelines on AF endorse the preoperative administration of amiodarone for POAF prophylaxis (class IIa, level of evidence [LOE] A) but are not specific regarding the dose, frequency, or route of administration.25
The American Association for Thoracic Surgery 2014 guidelines are more specific regarding the dose, frequency, and route of administration of amiodarone for POAF prevention but apply to high-risk pulmonary resection (class IIa, LOE A) and esophagectomy patients (class IIb, LOE B), not the cardiac surgical patient population. The association recommends a 300-mg IV bolus of amiodarone over 60 minutes, followed by 600 mg orally twice daily for 3 to 5 days, a guideline based in part on the cardiac surgical literature.8
This QI project was implemented in a 632-bed, level 2 tertiary care center in southeastern Michigan that serves a multicultural, diverse patient population. In 2016, 283 CABG surgeries were performed at this facility. The overall incidence of POAF among these patients was 24%. At the initiation of this project, there was no amiodarone POAF prophylaxis protocol in place. The existing practice was to administer amiodarone reactively to treat POAF rather than proactively to prevent it.
Preimplementation Data Collection
In 1989 the Society of Thoracic Surgeons (STS) developed a national database as a QI and patient safety initiative for cardiothoracic surgeons.26 The database contains multiple data points that can be used to assess performance and opportunities for improvement and to provide benchmarks for comparison across facilities. One of the data points is the incidence of POAF.
In August 2017 the STS database for this facility was queried to obtain current data to serve as a baseline. The query requested 6 months of data, from January to June 2017. Data from CABG-only surgery patients included preexisting AF; history of COPD, diabetes, liver disease, thyroid disease, and POAF; readmission occurrence and reason; age; sex; and LOS.
The query revealed that a total of 111 CABG-only surgeries had been performed during the 6-month period. Patients with preexisting AF, thyroid disease, and liver disease were excluded from this group, because they would not have been candidates for amiodarone prophylaxis. All patients were receiving a β-blocker and a statin unless the treatments were contraindicated. The incidence of POAF in this baseline sample was 25% (n = 28), similar to the 2016 proportion. Among the 28 patients with new-onset POAF, 82% were male and 18% were female. The sex ratio was roughly similar in the non-POAF group. Most of the POAF patients fell into the age group of 66 to 75 years (n = 13), followed by 56 to 65 years (n = 6). The inverse was true in the non-POAF group, with the largest proportion of patients falling into the age range of 56 to 65 years (n = 31), followed by 66 to 75 years (n = 26). Most of the patients in both groups were white, and the top 3 comorbidities in both groups were hypertension, diabetes, and obesity. The average LOS for patients with POAF was 8 days, compared with 7 days in the non-POAF group. Six (21%) of the POAF patients were readmitted to the hospital within 30 days of discharge, compared with 9 (11%) in the non-POAF group. The reasons for readmission in the POAF group included dysrhythmia and stroke. The STS does not specify the type of dysrhythmia but rather groups all types under a single heading. There were no inpatient deaths in the POAF group, and there was 1 inpatient death in the non-POAF group that was due to postoperative cardiac arrest with lack of response to resuscitative efforts.
The average cost for a surgical intensive care unit (SICU) bed at this facility is $5078 per day. The maximum reimbursable amount for the CABG surgery diagnosis-related group is based on a 7-day LOS. The data query revealed that the average LOS of patients in the group with new-onset POAF (n = 28) was 8 days, resulting in a possible combined loss of more than $142 000. This cost estimate does not include additional diagnostic testing or procedures that may have been required. Likewise, readmissions within 30 days of discharge after CABG surgery are not reimbursable. Depending on level of care, diagnostic testing, procedures, and LOS required, the cost of a hospital readmission could be considerably higher.
On the basis of an extensive literature review, the authors of this article collaborated to develop an amiodarone POAF prophylaxis protocol (Figure 2). An order set for the protocol was then created and embedded into the electronic medical record (EMR) within the cardiac surgery postoperative order set.
Patients who met inclusion criteria (Figure 2) received IV amiodarone after surgery according to the protocol following adequate volume resuscitation and transfusion of emergent blood products. Adequate volume resuscitation was determined by the cardiothoracic nurse practitioner or surgeon, in collaboration with the registered nurse, on the basis of the patient’s hemodynamic trends and physical assessment. Patients were generally considered stable for amiodarone prophylaxis if they were not dependent on epicardial pacing, had a cardiac index greater than 2 L/min/m2 without inotropic agents, and had a systemic vascular resistance between 800 and 1200 dynes/s/cm for 3 consecutive hours.
The rationale for IV administration during the first 24 hours was 2-fold: first, to maximize the bioavailability of amiodarone; and second, to avoid potential gastrointestinal upset with oral amiodarone before the patient could tolerate a diet. The dosing strategy was developed by synthesizing the findings of meta-analyses by Chatterjee et al,3 Buckley et al,17 and Patel et al.24 Patients on the protocol received a total IV loading dose of 990 mg within the first 24 hours after surgery, followed by 400 mg of oral amiodarone twice daily with meals for 5 days. The bioavailability of oral amiodarone is increased to approximately 70% when the drug is given with food.17,28–30 Taking this into consideration, the total prophylactic dose of amiodarone was estimated to be approximately 3790 mg by the end of postoperative day 5; this dose falls within the therapeutic range identified by Buckley and colleagues17 of 3000 to 5000 mg and keeps the dose less than 1000 mg in 24 hours, which helps avoid the potential adverse effects identified by Chatterjee et al3 and Patel et al.24
Patients who underwent emergent and nonemergent CABG surgery with cardiopulmonary bypass were considered for the amiodarone POAF prophylaxis protocol. Before initiation of the protocol, a cardiothoracic nurse practitioner determined whether the patient met the inclusion criteria by evaluating his or her medical history, diagnostic test results, and current medication list. Patients who met the inclusion criteria received the protocol and were monitored for development of POAF throughout their postoperative hospitalization.
Permission to implement this project was granted by the institutional review board. We maintained strict adherence to human rights protections, ethical standards of practice, and codes of conduct. No identifiable patient data were collected.
Data Collection and Analysis
Baseline data were obtained from the EMR by a cardiothoracic nurse practitioner before surgery to determine risk factors for POAF and whether the patient was a candidate for the protocol. Patients who did not meet inclusion criteria, as well as those who met inclusion criteria but were never started on the protocol because of staff concerns or misunderstanding of the protocol, were excluded from postimplementation analysis. Also excluded were patients who had CABG surgery from January 1 to 14, 2018, when there was a nationwide shortage of amiodarone; the protocol was placed on hold during this period to conserve resources for emergencies. The protocol was resumed on January 15, 2018, after the shortage had resolved. Patients who were not able to complete the protocol because of adverse effects were also excluded.
The cardiothoracic nurse practitioner ordered the protocol for all patients who were found to be eligible candidates upon arrival to the SICU after surgery. Data were collected throughout the patient’s postoperative course and were updated daily. Only cardiothoracic nurse practitioners had access to the data, which was kept in an unmarked binder that was locked in the cardiothoracic nurse practitioner’s office. No identifiable patient information was collected. Patients were assigned a chronological number based on the timing of their enrollment in the protocol. Data were collected from November 1, 2017, through February 15, 2018. The incidence of POAF and frequency distribution of demographic information were calculated using descriptive statistics (Table).
A total of 68 CABG surgeries were performed from November 1, 2017, to February 15, 2018. Seventeen patients were excluded from the protocol because of previous history of AF, thyroid dysfunction, elevated liver function test results, bradycardia, or hypotension. Four patients had the protocol discontinued within 48 hours because of adverse effects: suspected pulmonary toxicity (2), symptomatic bradycardia (1), and symptomatic hypotension (1). Forty-seven patients completed the protocol and were included in the statistical analysis. Amiodarone prophylaxis was started an average of 3 hours after arrival at the SICU. All patients in the amiodarone protocol were also started on a statin and a low-dose β-blocker as standard practice unless the treatments were contraindicated (eg, because of allergy, bradycardia, hypotension).
Eight of 47 (17%) patients receiving amiodarone prophylaxis developed POAF, which occurred on postoperative day 2 (n = 4) or 3 (n = 4). Patients who developed POAF were primarily male (n = 7), with an age range of 64 to 75 years. Four of these patients were white, and all had comorbidities. Four (50%) patients developed sustained or paroxysmal POAF that persisted for more than 24 hours, requiring anticoagulation. Fifty percent of patients with POAF had a previous history of CHF, with an ejection fraction of 40% or lower at the time of surgery. Fifty percent also had a history of CKD, and 50% had a BMI of greater than 30 kg/m2 . Sixty-three percent of patients with POAF had diabetes, and 100% of patients with POAF had a history of hypertension and a CHA2DS2-VASc score of 2 or greater. One patient died on postoperative day 5 as a result of complications associated with heparin-induced thrombocytopenia (HIT), end-stage renal disease, and cardiogenic shock. Six (15%) patients from the non-POAF group were readmitted within 30 days of discharge. No POAF patients were readmitted with dysrhythmia. The mean (SD) LOS for the entire study group (N = 47) was 6 (1) days, which falls within the CABG reimbursement time frame of 7 days.
The outcomes of this QI initiative reflect the findings from the literature. As in the meta-analyses by Bagshaw et al6 and Aasbo et al,23 amiodarone prophylaxis reduced the incidence of POAF in patients after CABG surgery. Although the amiodarone prophylaxis group (N = 47) was smaller than the baseline group (N = 111), a decrease in POAF (17% vs 25%) was noted. In addition, the mean (SD) LOS in the prophylaxis group was shorter than that in the baseline group (6  vs 8  days in those with POAF and 6  vs 7  days in those without POAF). These results indicate that the amiodarone POAF prophylaxis protocol could significantly reduce costs, improve patient outcomes, and increase the overall quality of care.6,22,23
The apparent risk factors of POAF were also similar to those identified in the literature. In examining the comorbidities of patients who experienced POAF despite amiodarone prophylaxis, 100% of them had a history of hypertension, and most were between the ages of 64 and 75. Findings were similar in the baseline group. Chronic hypertension, whether primary or secondary, often results in left ventricular hypertrophy, which can lead to left atrial dilation.11,31 Taking into account the age-related changes in the heart, such as increased collagen and loss of ventricular compliance, the likelihood of POAF was high.32
In addition to hypertension, 63% of the patients who experienced POAF had diabetes. The pathophysiology of diabetes can lead to CHF and CKD, which are also risk factors for POAF.4,7 This, too, was reflected in the sample; 50% of patients who experienced POAF had a history of CKD and CHF. Additionally, 50% of the POAF patients were obese (BMI > 30 kg/m2 ) and had systolic dysfunction (ejection fraction ≤ 40%) at the time of surgery. All patients in the POAF group had a CHA2DS2-VASc score of 2 or greater, which supports the finding by Chua et al7 that a CHA2DS2-VASc score of 2 or greater is a significant predictor of POAF. Similar to findings by Thanavaro et al,5 all patients in the POAF group experienced POAF on day 2 or 3. Although the correlation between risk factors and POAF appears stronger in the intervention group than in the baseline group (likely because of difference in sample size), both sets of results seem to suggest that patients with hypertension, CHF, CKD, and obesity are at higher risk for developing POAF with or without amiodarone prophylaxis.
Surprisingly, only 1 of 6 (17%) patients with COPD developed POAF while on the protocol, whereas 21% of patients with COPD in the baseline group developed POAF. These results could suggest that amiodarone prophylaxis reduces the incidence of POAF in patients with COPD. However, the severity of COPD was not evaluated in this project, and further research with a larger sample of patients would be necessary to determine the full impact of amiodarone prophylaxis on the development of POAF in patients with COPD.
Another unexpected finding was related to readmissions within 30 days of discharge. In the intervention group, 1 patient (13%) with POAF was readmitted for an unrelated issue, whereas 6 patients (15%) without POAF were readmitted. Additionally, none of the patients were readmitted with dysrhythmia. These results are contrary to the findings of Espinoza et al33 and Iribarne et al34 and differ from the outcomes of the 2017 baseline group, in which 21% of patients with POAF were readmitted. The difference could be a result of the small sample size of the amiodarone prophylaxis group, or it may suggest that amiodarone prophylaxis reduces readmissions after CABG surgery.
The review of the literature suggested that patients with POAF may have an increased mortality rate.19,20 Because of the small sample size, it is difficult to determine the impact of amiodarone on mortality. Continuation of this protocol with a larger sample size may provide a better predictive value related to mortality. Although the sample was small, 1 patient out of 8 with POAF died. The patient’s death was likely due to the presence of multiple comorbidities, which were risk factors for POAF. The patient developed HIT secondary to heparin anticoagulation for new-onset paroxysmal POAF, which resulted in coagulopathy and, ultimately, death. De Jong and Dekker13 found that platelet activation is a prodysrhythmic process. The pathophysiology of HIT may have perpetuated the paroxysmal POAF. More research would be necessary to determine whether a direct relationship exists between HIT and POAF.
This project had several identified limitations. The order set was not initially built into the EMR, and for the first week of the project the protocol had to be ordered "à la carte," with single orders linked together. Second, there was a steep learning curve during the first month of the protocol. Consequently, the protocol was initiated late or was not initiated at all in a few patients who were eligible for it, thus excluding them from the data analysis. In addition, a few patients had the protocol discontinued unnecessarily or had doses held owing to misunderstanding, excluding them from the data analysis. This problem was resolved by implementing one-on-one education with the nursing staff to remedy any misconceptions about the protocol. Third, a nationwide shortage of amiodarone occurred during the project, resulting in a 2-week pause in data collection because resources had to be reserved for treatment rather than prophylaxis. During this period, 10 patients would have had the protocol initiated had the amiodarone been available. This episode leads to the fourth limitation, which is the small sample size. A larger sample would have strengthened the value of the outcomes of this project. Finally, as the bioavailability of oral amiodarone varies widely from 30% to 70%, whether patients received a therapeutic dose is unknown. Although the oral doses were given with food in an attempt to increase the drug’s absorption, whether a therapeutic level (3000-5000 mg) was consistently attained remains unclear. If 70% of oral amiodarone was absorbed, the total therapeutic dose would have been 3790 mg. If only 50% or less was absorbed, the dose would fall below therapeutic levels. This issue has prompted consideration of whether to increase the oral dose of amiodarone to 600 mg twice daily with meals. However, such an increase could also increase the risk of adverse effects; further study is necessary.
Implications for Future Research
Data collection using the current protocol will continue. Depending on the outcomes, the oral dose of amiodarone may be increased to 600 mg twice daily with food to determine whether a further reduction of POAF occurs and whether the benefits of the increased dose outweigh the risk of adverse effects. Also under consideration is adding the potent anti-inflammatory agent colchicine to the protocol to mitigate the inflammatory etiology of POAF. In a study by Imazio and colleagues,35 the prophylactic administration of colchicine after CABG surgery reduced the incidence of POAF from 22% to 12% (P = .02). Analysis and revision of the current protocol by the authors will continue in order to generate additional data for a possible clinical practice guideline specific to the prevention of POAF in the cardiac surgical patient population.
The outcomes of this QI project provide additional evidence that the use of a prophylactic amiodarone protocol may reduce the incidence of POAF in patients undergoing CABG surgery. Confounding variables may contribute to the development of POAF despite antidysrhythmic therapy. Ongoing data collection and analysis will eventually provide additional insight into the occurrence of POAF in CABG patients.
The authors declare no conflicts of interest.