- Email: [email protected]
Atrial Fibrillation After Bypass Surgery
Atrial Fibrillation After Bypass Surgery* Does the Arrhythmia or the Characteristics of the Patients Prolong Hospital Stay? Steven Borzak, MD; James E . Tisdale, PharmD; Neeta B. Amin, PharmD; A David Goldberg , MD; Diane Frank, RN; I. Desmond Padhi, PharmD; and Robert S. D. Higgins, MD
Study objectives: The goal of this study was to detennine whether prolonged hospital stay associated with atrial fibrillation or flutter (AF) after coronary artery bypass graft (CABG) surgery is attributable to the characteristics of patients who develop this arrhythmia or to the rhythm disturbance itself. Design: An investigation was conducted through a prospective case series. Setting: Patients were from a single urban teaching hospital. Participants: Consecutive patients undergoing isolated CABG surgery between December 1994 and May 1996 were included in the study. Interventions: No interventions were involved. Results: Of 436 patients undergoing isolated CABG surgery, 101 (23%) developed AF. AF patients were older and more likely to have obstructive lung disease than patients without AF, but both patients with and without AF had similar left ventricular function and extent of coronary disease. ICU and hospital stays were longer in patients with AF. Multivariate analysis, adjusted for age, gender, and race, demonstrated that postoperative hospital stay was 9.2±5.3 days in patients with AF and 6.4±5.3 days in patients without AF (p<0.001). Conclusions: Although AF is strongly associated with advanced age, most of the prolonged hospital stay appears to be attributable to the rhythm itself and not to patient characteristics. (CHEST 1998; 113:1489-91) Key words: atrial fibrill ation ; coronary artery bypass surgery; resource utilization Abbreviations: AF = atrial fibrill ation or flutter; CABG= coronary artery bypass graft
Atrial fibrillation or flutter (AF) after coronary
artery bypass graft (CABG) surgery occurs in up to one third of patients.I-3 It is rarely fatal but can cause symptoms, lead to clinical instability, and, despite its typically short duration, may be associated with stroke in the postoperative setting. 3 -5 Although overall length of stay after CABG surgery has declined steadily over the past decade, AF is strongly associated with a prolonged l ength of hospital stay and increased hospital costs.1 •2 Lazar and others 6 found that atrial arrhythmias were among the most common postoperative complications associated with
*From the Divisions of Cardiology (Drs. Borzak and Goldberg and Ms. Frank) and Cardiothoracic Surgery (Dr. Higgins), Henry Ford Heart and Vascular Institute, Henry Ford Hospital, and the College of Pharmacy (Drs. Tisdale, Amin , and Padhi ), Wayn e State University, Detroit. grant from Hoechst M arion Roussell. r Supported b y a esearch Manuscript received Ju ly 16, 1997; revision accepted November 7, 1997. Reprint requests: Steven Bor:::,ak, MD , Ca rdiovascular Division, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202
prolonged stay after CABG surgery. However, it is unclear whether prolonged hospitalization after CABG surgery in AF patients is due to the arrhythmia itself or to the advanced age and other clinical characteristics of patients in whom AF is likely to develop. The present analysis attempts to define the relative contributions of AF and age to hospitalization after CABG surgery.
MATERIALS AND METHODS rd Hospital between The study was conducted at H emy Fo December 1994 and May 1996 and was approved by the Human Rights Co mmittee (Institutional R eview Board). All patients undergoing isolated CABG surgery were prospectively e valuated and screened for entry into a study examining management of AF after CAB G surgery.' Demographic and outcome data were prospectively obtained b y anurse coordinator or re search f ellow from the medical record concurrent with the patient's hospitalization and retrospectively ve rified. Length of stay data were confirm ed b y rcoss-referencing with the hospital billing database. AF was considered to have occurred if th e arrhythmia persisted CHEST / 113 / 6 / JUNE, 1998
1489
at least 15 min and required treatment or if episodes fewer than 15 min in duration caused clinical instability and the need for treatment. Baseline clinical characteristics were compared in patients with and without AF with the use of Student's t test for continuous variables and x2 analysis for categorical variables. In addition to baseline variables shown in Table 1, other factors considered in analyses included the following: weight; smoking history; history of angina, heart failure, hypercholesterolemia, stroke, arrhythmia, and prior revascularization; baseline ECG characte1istics; and preoperative physical findings. Due to non-normality, lengths of stay were compared in patients with and without AF with the use of the Wilcoxon rank-sum test. Stepwise linear regression analysis was performed to determine factors associated \Vith postoperative length of stay. Analysis of covariance was used to calculate postoperative length of stay in the groups with and without AF after adjusting for other predictors of prolonged stay. Due to non-normality, the length of stay was logarithm-transformed b efore the linear regression and analysis of covariance were performed.
Table 2-Postoperative Outcomes Group Va1iable Total length of hospital stay, mean:t:SD Postoperative length of stay, mean:t:SD Postoperative ICU stay, mean:t:SD
With AF (n = 101)
Without AF (n=335)
p Value
11.6:':9.1
8.7:':4.3
< 0.001
9.4:':8.8
6.3 :':3.5
< 0.001
2.7:':6.3
l.7:t:2.4
0.004
age, gender, and race, AF remained a significant predictor of prolonged hospital stay (Table 3).
DISCUSSION RESULTS
A total of 436 consecutive patients undelWent isolated CABG surgery during the study period and were included in the analysis . AF developed in 101 patients (23%). Characteristics of the groups with and without AF are compared in Table 1, showing that the members of the AF group were older. Postoperative outcomes are compared in Table 2. Hospital stay was prolonged in AF patients. To determine the relative contributions of AF and other clinical features of patients to length of stay, the final model of stepwise linear regression is shown in Table 3. The occurrence of AF was the strongest predictor of prolonged postoperative length of stay, followed by advanced age, female gender, and African-American race (Table 3). When analysis of covariance was performed to adjust length of stay for
Twenty-three percent of patients undergoing isolated CABG surgery experienced clinically relevant AF; this percentage is in accord with that of other contemporary series. 1 -3 ·6 ·8 Characteristics of AF patients differed significantly from those of patients remaining in sinus rhythm in that AF patients were older, which is consistent with findings in previous reports.I- 3 As previously described by others,2 this study found an association with pulmonary disease; however, it is the first investigation to describe an association between prolonged hospital stay and African-American race, a finding which would appear not to have a biologic hypothesis but which merits further investigation.
Table 3-Multivariate Logistic Regression Analysis of Length of Stay Vmi ables Associated With Prolonged Hospital Stay
Table !-Comparison of Baseline Characteristics in Groups With and Without AF Group Vmiable
With AF Without AF (n= 101 ) (n=335 ) p Value
66.4:':9.4 61.2:':10.1 < 0.001 Mean age, years; :t:SD 72 72 Male gender, % 0.95 16 23 0.13 African-Ameri can, % History of hypertension, % 69 61 0.13 31 30 0.92 Histmy of diabetes, % Prior myocardial infarction, % 31 35 0.46 COPD , % 11 3 0.001 56:t:14 58:t:14 Left ventricular ejection fraction , 0.22 mean:t:SD 2.7:':0.8 0.22 No. of di seased vessels, mean:t:SD 2.6:':0.8 Presence of valve disease, % 5.9 3.6 0.30 77 84 0.12 Postoperative beta blocker use, %
1490
Variable Intercept AF Age g reater than average Male gender African-American race
Parameter Standard F Estimate Error Statistic 1.408 0.2.56 0.007 -0.139 0.160
0.116 0.042 0.002 0.039 0.043
147.1 36.6 15.8 12.7 13.8
p Value < 0.001 < 0.001 < 0.001 < 0.001 < 0.001
Unadjusted and Adjusted Length of Stay by AF Status Unadjusted With AF Postoperative length of stay, d
Without AF
Adjusted* With AF
Without AF
9.4 :t:8.8 6.3 :t:3.5 1 9.2 :t:5.3 6.4 :t:5.3 1
*Adjusted for age, race, and gender. 1 p<0.001 with AF vs without AF. Clinical Investigations
As others have found, 1-3 this study showed that AF was associated with significantly longer hospital stay following CABG surgery. Unlike previous studies, however, the analysis in this one was specifically directed at understanding whether the longer hospitalization was due to the arrhythmia itself or due to the clinical characteristics of AF patients. Analysis in this study suggests that after accounting for advanced age, gender, and race, the key variables associated with prolonged stay, AF was still associated with a nearly 3-day difference in length of stay. One limitation of the present study is that reasons for the prolonged hospital stay attributable to AF were not examined. One factor may be that achievement of ventricular rate control with digoxin, the standard first approach to treating AF after CABG surgery, 3 ·9 may take up to 12 to 24 h. 7 Aranki and colleagues 1 found patients with AF had an increased incidence of pneumonia, prolonged mechanical ventilation, and other respiratory complications, as well as renal failure and more serious ventricular arrhythmias. This study also found an association between pulmonary disease and AF (Table 1), but this association did not account for prolonged hospital stay in the multivariate analysis. AF after CABG surgery also may be associated rarely with stroke.4 However, these more serious complications, despite their increased tendency to occur in AF patients, were still relatively rare, suggesting that in the majority of patients, treatment of the AF itself may explain a large portion of the longer stay. Therefore, AF is common after CABG surgery and occurs more commonly in older patients. AF is associated with an additional 3 days of hospitalization after accounting for other clinical features which prolong stay. This study suggests the hypothesis that
strategies to restore sinus rhythm aggressively should be prospectively evaluated against standard rate control to reduce hospital stay. ACKNOWLEDGMENT: We thank Drs. Gaetano Paone, Norman Silverman, and Charles Webb for their invaluable contributions and Gordon Jacobsen for statistical analyses.
REFERENCES 1 Aranki SF, Shaw DP, Adams DH, et a!. Predictors of atrial fibrillation after coronmy artery surgery: current trends and impact on hospital resources. Circulation 1996; 94:390-97 2 Mathew JP, Parks R, Savino JS, et a!. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA 1996; 276:300-06 3 Lauer MS, Eagle KA, Buckley MJ, et a!. Atrial fibrillation following coronary artery bypass surgery. Prog Carcliovasc Dis 1989; 31:367-78 4 Taylor GJ, Malik SA, Colliver JA, et a!. Usefulness of atrial fibrillation as a predictor of stroke after isolated coronary artery bypass grafting. Am J Cardiol 1987; 60:905-07 5 Creswell LL, Schuessler RB, Rosenbloom M, eta!. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993; 56: 539-49 6 Lazar HL, Fitzgerald C, Gross S, et a!. Determinants of length of stay after coronary artery bypass graft surgery. Circulation 1995; 92(suppl II):II-20-II-24 7 Tisdale JE, Padhi ID, Goldberg AD, eta!. Diltiazem controls ventricular rate more rapidly but not more effectively than digoxin in patients with atrial fibrillation following coronary arte1y bypass surgery: a randomized double-blind study. Am Heart J 1998 (in press) 8 Hashimoto K, Ilstrup DM, SchaffHV. Influence of clinical and hemodynamic variables on risk of supraventricular tachycardia after coronary artery bypass. J Thorac Cardiovasc Surg 1991; 101:56-65 9 Pires LA, Wagshal AB, Lancey R, et a!. Arrhythmias and conduction disturbances after coronary artery bypass graft surgery: epidemiology, management and prognosis. Am Heart J 1995; 129:799-808
CHEST /113 I 61 JUNE, 1998
1491
Study objectives: The goal of this study was to detennine whether prolonged hospital stay associated with atrial fibrillation or flutter (AF) after coronary artery bypass graft (CABG) surgery is attributable to the characteristics of patients who develop this arrhythmia or to the rhythm disturbance itself. Design: An investigation was conducted through a prospective case series. Setting: Patients were from a single urban teaching hospital. Participants: Consecutive patients undergoing isolated CABG surgery between December 1994 and May 1996 were included in the study. Interventions: No interventions were involved. Results: Of 436 patients undergoing isolated CABG surgery, 101 (23%) developed AF. AF patients were older and more likely to have obstructive lung disease than patients without AF, but both patients with and without AF had similar left ventricular function and extent of coronary disease. ICU and hospital stays were longer in patients with AF. Multivariate analysis, adjusted for age, gender, and race, demonstrated that postoperative hospital stay was 9.2±5.3 days in patients with AF and 6.4±5.3 days in patients without AF (p<0.001). Conclusions: Although AF is strongly associated with advanced age, most of the prolonged hospital stay appears to be attributable to the rhythm itself and not to patient characteristics. (CHEST 1998; 113:1489-91) Key words: atrial fibrill ation ; coronary artery bypass surgery; resource utilization Abbreviations: AF = atrial fibrill ation or flutter; CABG= coronary artery bypass graft
Atrial fibrillation or flutter (AF) after coronary
artery bypass graft (CABG) surgery occurs in up to one third of patients.I-3 It is rarely fatal but can cause symptoms, lead to clinical instability, and, despite its typically short duration, may be associated with stroke in the postoperative setting. 3 -5 Although overall length of stay after CABG surgery has declined steadily over the past decade, AF is strongly associated with a prolonged l ength of hospital stay and increased hospital costs.1 •2 Lazar and others 6 found that atrial arrhythmias were among the most common postoperative complications associated with
*From the Divisions of Cardiology (Drs. Borzak and Goldberg and Ms. Frank) and Cardiothoracic Surgery (Dr. Higgins), Henry Ford Heart and Vascular Institute, Henry Ford Hospital, and the College of Pharmacy (Drs. Tisdale, Amin , and Padhi ), Wayn e State University, Detroit. grant from Hoechst M arion Roussell. r Supported b y a esearch Manuscript received Ju ly 16, 1997; revision accepted November 7, 1997. Reprint requests: Steven Bor:::,ak, MD , Ca rdiovascular Division, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202
prolonged stay after CABG surgery. However, it is unclear whether prolonged hospitalization after CABG surgery in AF patients is due to the arrhythmia itself or to the advanced age and other clinical characteristics of patients in whom AF is likely to develop. The present analysis attempts to define the relative contributions of AF and age to hospitalization after CABG surgery.
MATERIALS AND METHODS rd Hospital between The study was conducted at H emy Fo December 1994 and May 1996 and was approved by the Human Rights Co mmittee (Institutional R eview Board). All patients undergoing isolated CABG surgery were prospectively e valuated and screened for entry into a study examining management of AF after CAB G surgery.' Demographic and outcome data were prospectively obtained b y anurse coordinator or re search f ellow from the medical record concurrent with the patient's hospitalization and retrospectively ve rified. Length of stay data were confirm ed b y rcoss-referencing with the hospital billing database. AF was considered to have occurred if th e arrhythmia persisted CHEST / 113 / 6 / JUNE, 1998
1489
at least 15 min and required treatment or if episodes fewer than 15 min in duration caused clinical instability and the need for treatment. Baseline clinical characteristics were compared in patients with and without AF with the use of Student's t test for continuous variables and x2 analysis for categorical variables. In addition to baseline variables shown in Table 1, other factors considered in analyses included the following: weight; smoking history; history of angina, heart failure, hypercholesterolemia, stroke, arrhythmia, and prior revascularization; baseline ECG characte1istics; and preoperative physical findings. Due to non-normality, lengths of stay were compared in patients with and without AF with the use of the Wilcoxon rank-sum test. Stepwise linear regression analysis was performed to determine factors associated \Vith postoperative length of stay. Analysis of covariance was used to calculate postoperative length of stay in the groups with and without AF after adjusting for other predictors of prolonged stay. Due to non-normality, the length of stay was logarithm-transformed b efore the linear regression and analysis of covariance were performed.
Table 2-Postoperative Outcomes Group Va1iable Total length of hospital stay, mean:t:SD Postoperative length of stay, mean:t:SD Postoperative ICU stay, mean:t:SD
With AF (n = 101)
Without AF (n=335)
p Value
11.6:':9.1
8.7:':4.3
< 0.001
9.4:':8.8
6.3 :':3.5
< 0.001
2.7:':6.3
l.7:t:2.4
0.004
age, gender, and race, AF remained a significant predictor of prolonged hospital stay (Table 3).
DISCUSSION RESULTS
A total of 436 consecutive patients undelWent isolated CABG surgery during the study period and were included in the analysis . AF developed in 101 patients (23%). Characteristics of the groups with and without AF are compared in Table 1, showing that the members of the AF group were older. Postoperative outcomes are compared in Table 2. Hospital stay was prolonged in AF patients. To determine the relative contributions of AF and other clinical features of patients to length of stay, the final model of stepwise linear regression is shown in Table 3. The occurrence of AF was the strongest predictor of prolonged postoperative length of stay, followed by advanced age, female gender, and African-American race (Table 3). When analysis of covariance was performed to adjust length of stay for
Twenty-three percent of patients undergoing isolated CABG surgery experienced clinically relevant AF; this percentage is in accord with that of other contemporary series. 1 -3 ·6 ·8 Characteristics of AF patients differed significantly from those of patients remaining in sinus rhythm in that AF patients were older, which is consistent with findings in previous reports.I- 3 As previously described by others,2 this study found an association with pulmonary disease; however, it is the first investigation to describe an association between prolonged hospital stay and African-American race, a finding which would appear not to have a biologic hypothesis but which merits further investigation.
Table 3-Multivariate Logistic Regression Analysis of Length of Stay Vmi ables Associated With Prolonged Hospital Stay
Table !-Comparison of Baseline Characteristics in Groups With and Without AF Group Vmiable
With AF Without AF (n= 101 ) (n=335 ) p Value
66.4:':9.4 61.2:':10.1 < 0.001 Mean age, years; :t:SD 72 72 Male gender, % 0.95 16 23 0.13 African-Ameri can, % History of hypertension, % 69 61 0.13 31 30 0.92 Histmy of diabetes, % Prior myocardial infarction, % 31 35 0.46 COPD , % 11 3 0.001 56:t:14 58:t:14 Left ventricular ejection fraction , 0.22 mean:t:SD 2.7:':0.8 0.22 No. of di seased vessels, mean:t:SD 2.6:':0.8 Presence of valve disease, % 5.9 3.6 0.30 77 84 0.12 Postoperative beta blocker use, %
1490
Variable Intercept AF Age g reater than average Male gender African-American race
Parameter Standard F Estimate Error Statistic 1.408 0.2.56 0.007 -0.139 0.160
0.116 0.042 0.002 0.039 0.043
147.1 36.6 15.8 12.7 13.8
p Value < 0.001 < 0.001 < 0.001 < 0.001 < 0.001
Unadjusted and Adjusted Length of Stay by AF Status Unadjusted With AF Postoperative length of stay, d
Without AF
Adjusted* With AF
Without AF
9.4 :t:8.8 6.3 :t:3.5 1 9.2 :t:5.3 6.4 :t:5.3 1
*Adjusted for age, race, and gender. 1 p<0.001 with AF vs without AF. Clinical Investigations
As others have found, 1-3 this study showed that AF was associated with significantly longer hospital stay following CABG surgery. Unlike previous studies, however, the analysis in this one was specifically directed at understanding whether the longer hospitalization was due to the arrhythmia itself or due to the clinical characteristics of AF patients. Analysis in this study suggests that after accounting for advanced age, gender, and race, the key variables associated with prolonged stay, AF was still associated with a nearly 3-day difference in length of stay. One limitation of the present study is that reasons for the prolonged hospital stay attributable to AF were not examined. One factor may be that achievement of ventricular rate control with digoxin, the standard first approach to treating AF after CABG surgery, 3 ·9 may take up to 12 to 24 h. 7 Aranki and colleagues 1 found patients with AF had an increased incidence of pneumonia, prolonged mechanical ventilation, and other respiratory complications, as well as renal failure and more serious ventricular arrhythmias. This study also found an association between pulmonary disease and AF (Table 1), but this association did not account for prolonged hospital stay in the multivariate analysis. AF after CABG surgery also may be associated rarely with stroke.4 However, these more serious complications, despite their increased tendency to occur in AF patients, were still relatively rare, suggesting that in the majority of patients, treatment of the AF itself may explain a large portion of the longer stay. Therefore, AF is common after CABG surgery and occurs more commonly in older patients. AF is associated with an additional 3 days of hospitalization after accounting for other clinical features which prolong stay. This study suggests the hypothesis that
strategies to restore sinus rhythm aggressively should be prospectively evaluated against standard rate control to reduce hospital stay. ACKNOWLEDGMENT: We thank Drs. Gaetano Paone, Norman Silverman, and Charles Webb for their invaluable contributions and Gordon Jacobsen for statistical analyses.
REFERENCES 1 Aranki SF, Shaw DP, Adams DH, et a!. Predictors of atrial fibrillation after coronmy artery surgery: current trends and impact on hospital resources. Circulation 1996; 94:390-97 2 Mathew JP, Parks R, Savino JS, et a!. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA 1996; 276:300-06 3 Lauer MS, Eagle KA, Buckley MJ, et a!. Atrial fibrillation following coronary artery bypass surgery. Prog Carcliovasc Dis 1989; 31:367-78 4 Taylor GJ, Malik SA, Colliver JA, et a!. Usefulness of atrial fibrillation as a predictor of stroke after isolated coronary artery bypass grafting. Am J Cardiol 1987; 60:905-07 5 Creswell LL, Schuessler RB, Rosenbloom M, eta!. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993; 56: 539-49 6 Lazar HL, Fitzgerald C, Gross S, et a!. Determinants of length of stay after coronary artery bypass graft surgery. Circulation 1995; 92(suppl II):II-20-II-24 7 Tisdale JE, Padhi ID, Goldberg AD, eta!. Diltiazem controls ventricular rate more rapidly but not more effectively than digoxin in patients with atrial fibrillation following coronary arte1y bypass surgery: a randomized double-blind study. Am Heart J 1998 (in press) 8 Hashimoto K, Ilstrup DM, SchaffHV. Influence of clinical and hemodynamic variables on risk of supraventricular tachycardia after coronary artery bypass. J Thorac Cardiovasc Surg 1991; 101:56-65 9 Pires LA, Wagshal AB, Lancey R, et a!. Arrhythmias and conduction disturbances after coronary artery bypass graft surgery: epidemiology, management and prognosis. Am Heart J 1995; 129:799-808
CHEST /113 I 61 JUNE, 1998
1491