Preoperative Brain Natriuretic Peptide and Atrial Arrhythmias After Coronary Artery Bypass Graft Surgery Ethan Bernstein, MPH,*† Robert Block, MD, MPH,‡§ Peter Veazie, PhD,‡ Christine Tompkins, MD,§ C. Allen Bashour, MD,*║ and Alparslan Turan, MD* Objective: To assess the association of preoperative brain natriuretic peptide with atrial arrhythmias and length of stay after cardiac surgery. Design: A retrospective data analysis. Setting: All data were collected from patients who underwent cardiac surgery at a single institution, an academic hospital, between 2005 and 2010. Participants: Patient data were collected from the authors’ institution’s Perioperative Health Documentation System of cardiac surgeries. Interventions: None. Measurements and Main Results: The major findings were that individuals with a 10-pg/mL increase in brain natriuretic peptide were 1.005 (95% CI: 1.002, 1.009) times more likely to have an atrial arrhythmia than those with no increase in brain natriuretic peptide. A brain natriuretic peptide value Z306 pg/mL was calculated to best predict
an atrial arrhythmia. Those above the threshold were 1.455 (95% CI: 1.087, 1.947) times more likely to have an atrial arrhythmia than those below the threshold. Individuals above the threshold had a median of 3 days in the intensive care unit compared to 1 day for those below the threshold, as well as median hospital stays of 11 days for those below the threshold. Conclusions: The results indicated that elevated brain natriuretic peptide was associated with increased risk of atrial arrhythmias and prolonged length of hospital stay after cardiac surgery. Identifying at-risk patients is important to guide preventative strategies for postoperative atrial arrhythmias. & 2015 Elsevier Inc. All rights reserved.
A
Identifying predictors of atrial arrhythmias in patients undergoing cardiac surgery can help clinicians identify candidates for preventive treatments, to reduce the morbidity and mortality associated with postoperative AF. The goals of this study were to evaluate whether preoperative BNP predicted atrial fibrillation post-CABG and to investigate if BNP was associated with prolonged ICU and hospital length of stay.
FTER CORONARY ARTERY BYPASS GRAFT (CABG) surgery, patients are at risk for the development of atrial arrhythmias, including atrial fibrillation, atrial flutter, supraventricular tachycardia, and premature atrial contractions. Atrial fibrillation (AF) is the most common cardiac rhythm disturbance, occurring in 25% to 40% of patients undergoing cardiac surgery.1,2 Postoperative atrial arrhythmias are associated with increased risk of morbidity, mortality, readmission to the intensive care unit (ICU), length of hospital and ICU stay, perioperative myocardial infarction, congestive heart failure, and stroke.3–11 Because of the challenges of predicting and preventing AF, there has been substantial interest in the use of biomarkers to identify patients who are at risk for atrial arrhythmias. One of these potential biomarkers is brain natriuretic peptide (BNP).12–16 BNP commonly is used for detecting congestive heart failure, but recently has been examined as a possible biomarker for atrial arrhythmias.11 BNP is released predominantly from the ventricular myocytes in response to left ventricular hypertrophy, which has been associated with AF.17–21 Previous attempts to determine an association between elevated preoperative BNP levels and atrial arrhythmias following CABG surgery have provided conflicting results. In a prospective trial of 144 consecutive surgeries, Ata et al found an association between preoperative BNP and postoperative AF (226 pg/mL, in those with AF v 65.2 pg/mL, in those without AF; p o 0.001).22,23 In contrast, Sezai et al24 did not show a statistically significant association between preoperative BNP and AF in 234 consecutive patients (137.2 ⫾ 215.7 pg/dL, in those with AF v 149.7 ⫾ 262.0 pg/dL, in those without AF; p ¼ 0.751). Thus, the relationship between BNP concentrations and AF is controversial. Furthermore, these studies had a limited sample size, and they were reported from very different surgical settings. Although conducted retrospectively, the authors’ data were collected prospectively at the time of surgery and will be the largest study to assess the relationship between BNP and AF.
KEY WORDS: arrhythmias, cardiac, coronary artery bypass, natriuretic peptide, brain
METHODS
With approval from both institutions’ Institutional Review Board and Research Subjects Review Board, a retrospective analysis of prospectively collected data from the Perioperative Health Documentation System that were acquired between 2005 and 2010 was performed. Data were collected from both
From the *Department of Outcomes Research, Cleveland Clinic, Cleveland, Ohio, †Uniformed Services University of the Health Sciences, Bethesda, Maryland, ‡Department of Public Health Sciences, University of Rochester, Rochester, New York, §University of Colorado Anschutz Medical Campus Division of Cardiology, Aurora, Colorado; and ║Department of Cardiothoracic Anesthesiology, Cleveland Clinic, Cleveland, Ohio. Supported by the National Institutes of Health/National Center for Research Resources CTSI award (grant number KL2 RR024136) to (T. P.) at the University of Rochester. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health. Support was also provided solely from institutional and/or departmental sources. None of the authors has a personal financial interest in this research. Address reprint requests to Alparslan Turan, MD, Department of Outcomes Research, Cleveland Clinic, 9500 Euclid Avenue, P-77, Cleveland, OH, 44195. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2014.11.015
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 3 (June), 2015: pp 611–616
611
612
BERNSTEIN ET AL
the hospital and anesthesia records, which integrated preoperative, intraoperative, and postoperative outcomes. Statistical Methods The authors assessed the relationship between preoperative BNP and atrial arrhythmias using unadjusted logistic regression, followed by a backwards selection multivariate logistic regression model, adjusting for unevenly distributed covariates. Covariates were selected based upon previously determined relationships between the covariate and postoperative AF. The selection included all potential covariates, the squares of the continuous covariates, and interactions between BNP and the covariates. The authors analyzed 29 predetermined potential covariates between patients who developed atrial arrhythmias and patients who did not, using a student’s t-test (for continuous variables), chi-square test (for dichotomous variables), or Wilcoxon rank sum test (for non-normal distributions). Any variable with a p value o 0.10, between those who did and did not develop an atrial arrhythmia, were considered to be covariates and adjusted for in the analysis. Although originally included in the analysis, postoperative medications were removed because of their use, possibly resulting from arrhythmias, and not a predictor of arrhythmias. A sensitivity analysis including all patients with BNP concentrations and imputed data for those with missing covariate data was performed; imputed data were not used in the final analysis. The authors determined a threshold BNP concentration to best predict atrial arrhythmias following CABG surgery, using a receiver operating characteristic (ROC) analysis and the Youden index (the sum of the sensitivity and specificity). Using the calculated threshold value, an unadjusted and adjusted multivariate logistic regression model, with the same covariates as the adjusted model for a one-unit increase in BNP, was performed to assess the risk of atrial arrhythmias in individuals above and below the BNP threshold. Kaplan-Meier analyses, using the calculated threshold value, were conducted to assess the lengths of ICU and hospital stay. Wilcoxon and log-rank tests were used to compare the lengths of stay between those above and below the BNP threshold. Lengths of ICU and hospital stay were defined as the duration of stay from the date of surgery to the date of ICU and hospital discharge, respectively. SAS software version 9.2.2 (SAS Institute, Cary, NC) and PASS 11 software for Windows (NCSS, Kaysville, UT) were used for all statistical analyses.
Those who did not develop an atrial arrhythmia were found to have an average BNP concentration of 315.9 pg/mL, and those who did develop an atrial arrhythmia had an average concentration of 557.5 pg/mL. For a 10-unit increase in BNP, the unadjusted odds ratio for an atrial arrhythmia was 1.0005 (95% CI: 1.00047, 1.00054; p o 0.001). For the adjusted model, the odds ratio was (95% CI, p value) when adjusting for baseline BNP, age, body mass index, ejection fraction, race, sex, American Society of Anesthesiologists physical status classification, creatinine, smoking, preoperative AF, chronic obstructive pulmonary disease/asthma, preoperative statin, preoperative beta-blocker, preoperative anti-arrhythmic medication, length of surgery, valve surgery, creatinine squared, BNP squared, interaction of BNP and length of surgery, and an interaction of BNP and valve surgery. The adjusted odds ratio for a 10-unit increase in BNP is 1.005 (95% CI: 1.002, 1.009; standard error: 0.000291; p ¼ 0.004) for patients who did not have valve surgery and had the mean length of surgery. The calculated threshold value was 306 pg/mL, with a sensitivity of 61.7% and specificity of 70.5%, for a Youden index of 1.322. The area under the curve for the ROC analysis was 0.718 (Fig 2). The positive predictive value of the
RESULTS
The authors evaluated 18,064 patients who underwent cardiac surgery between 2005 and 2010. Of these patients, 9,378 were excluded because they did not have preoperative BNP concentrations, and an additional 3,887 were removed because of missing covariate data or being statistical outliers. The remaining 4,765 patients were included in the analysis (Fig 1). The baseline characteristics (Table 1) for individuals who developed atrial arrhythmias and those who did not were found to be different for all but 3 covariates: Smoking, preoperative atrial flutter, and preoperative nonsteroidal antiinflammatory drug use.
Fig 1. Flow chart of included and excluded patients from the authors’ institutions’ Perioperative Documentation System of cardiac surgeries 2005-2010.
613
BRAIN NATRIURETIC PEPTIDE AND CARDIAC SURGERY
Table 1. Baseline Characteristics of the Population of Patients with Preoperative BNP Values Variable
BNP, mean (median [Q1, Q3]), pg/mL Age, mean (median, Q1, Q3), yr BMI, mean (median, Q1, Q3), kg/m2 Male, n (%) Race, n (%) African-American, n (%) Asian/Pacific Islander, n (%) Caucasian, n (%) Hispanic/Latino, n (%) Other, n (%) ASA status,‡ n (%) ASA status 1, 2, or 3, n (%) ASA status 4, n (%) ASA status 5, n (%) Smoking, n (%) Current/former smoker, n (%) Never smoked, n (%) History of congestive heart failure, n (%) Hypertension, n (%) Preoperative atrial flutter, n (%) Preoperative atrial fibrillation, n (%) History of myocardial infarction, n (%) MI o1 week, n (%) MI 41 week, o3 months, n (%) MI 3-6 months, n (%) MI 46 months, n (%) No History of MI, n (%) Diabetes, n (%) Peripheral vascular disease, n (%) Alcohol consumption, n (%) Ejection fraction, mean (SD), % COPD/asthma, n (%) Creatinine, mean (median, Q1, Q3), mg/dL Troponin, mean (median, Q1, Q3), ng/mL Preoperative statin, n (%) Preoperative beta-blocker, n (%) Preoperative antiarrhythmic, n (%) Preoperative ACE inhibitor, n (%) Preoperative steroids, n (%) Preoperative NSAID, n (%) Preoperative nitrate, n (%) Length of surgery, mean (SD), H:MM Postoperative amiodarone, n (%) Postoperative dopamine, n (%) Postoperative epinephrine, n (%) Valve surgery, n (%) No Left Right Both
Atrial Arrhythmia (n ¼ 1,226)
No Atrial Arrhythmia (n ¼ 3,548)
557.5 70.0 28.8 735
(265 [111, 671]) (70 [62, 78]) (27.7 [24.3, 32.0]) (59.95)
315.9 63.5 28.1 2283
(122.5 [45, 322]) (65 [74, 65]) (27.2 2 [4.2, 31.2]) (64.35)
63 9 1118 22 14
(5.1) (0.73) (91.2) (1.8) (1.1)
219 34 3164 54 77
(6.2) (0.96) (89.2) (1.5) (2.2)
p Value*
o0.001† o0.001† o0.001† 0.006 0.0898
o0.001 99 (8.1) 1088 (88.7) 39 (3.2)
577 (16.3) 2892 (81.5) 79 (1.6) 0.005
677 549 594 929 25 457
(55.2) (44.8) (48.4) (75.8) (2.0) (37.3)
1793 1755 1193 2634 56 863
(50.5) (49.5) (33.6) (66.6) (1.58) (24.3)
213 65 119 8 821 397 209 131 46.3 245 1.37 0.26 554 614 234 385 129 16 155 7:15 255 17 785
(17.4) (5.3) (9.7) (0.65) (67.0) (32.4) (17.1) (10.7) (15.48) (20.0) (1.1, 0.9, 1.4) (0.00 [0.00, 0.03]) (45.2) (50.08) (19.1) (31.4) (10.5) (1.31) (12.6) (1:53) (20.8) (1.39) (64.0)
452 148 200 24 2724 850 405 468 50.8 451 1.17 0.15 1507 1465 554 1050 265 46 323 6:29 167 17 1487
(12.7) (5.3) (5.6) (0.68) (76.8) (24.0) (11.4) (13.2) (13.53) (12.7) (1.0, 0.8, 1.2) (0.00 0 [.00, 0.01]) (42.5) (41.3) (15.61) (29.6) (7.5) (1.30) (9.1) (1:41) (4.7) (0.48) (41.9)
336 679 36 175
(27.4) (55.4) (1.72) (14.3)
1047 2141 61 299
(29.5) (60.3) (1.72) (8.4)
o0.001 o0.001 0.2815 o0.001 o0.001
o0.001 o0.001 0.022 o0.001 o0.001 o0.001† o0.001† 0.0983 o0.001 0.005 0.2337 o0.001 0.9818 o0.001 o0.001 o0.001 0.001 o0.001 o0.001
Abbreviations: ACE, angiotensin-converting enzyme; BMI, body mass index; BNP, brain natriuretic peptide; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; NSAID, nonsteroidal anti-inflammatory drug. *p values calculated using t-test unless otherwise noted. †p value calculated using the Wilcoxon signed-rank test. ‡ASA status refers to the American Society of Anesthesiologists (ASA) physical status classification.
threshold was 41.9%, and the negative predictive value was 84.2%. Using the threshold resulted in 1,490 individuals with an elevated BNP and 3,275 individuals with a nonelevated BNP.
The unadjusted odds ratio was 2.367 (95% CI: 2.068, 2.709; p o 0.001), and the adjusted odds ratio was 1.455 (95% CI: 1.087, 1.947; p ¼ 0.031) for individuals above the threshold of 306 pg/mL who had no valve surgery and the mean length of
614
BERNSTEIN ET AL
DISCUSSION
Fig 2. Receiver operating characteristic (ROC) curve for the calculated threshold value of 306 pg/mL, Area under the curve ¼ 0.72.
surgery compared to individuals below the threshold who had no valve surgery and the mean length of surgery. BNP concentrations above the threshold were significantly associated with increased lengths of ICU (Fig 3) and hospital stay (Fig 4). Using both the log-rank and Wilcoxon tests to compare the curves, a significant difference (p o 0.001) was found in the time to ICU (Table 2) and hospital discharge (Table 3) between those with a BNP value above the threshold compared to those with a BNP value below the threshold. Table 3
Fig 3.
The results of this study indicated that increased levels of preoperative BNP were associated with an increased risk of atrial arrhythmias following CABG surgery. The authors found an odds ratio of 1.005 (95% CI: 1.002, 1.009; p ¼ 0.003) for a 10-unit change in BNP, and for individuals above the calculated threshold of 306 pg/mL, an odds ratio of 1.455 (95% CI: 1.087, 1.947; p ¼ 0.031). Those above the threshold also were found to have increased lengths of ICU and hospital stay. Elevated BNP can be used as an identifier of those at risk for the development of an atrial arrhythmia and may benefit from prophylactic therapy with antiarrhythmic drugs.25 The threshold value determined by this study was higher than the 135-pg/mL threshold previously reported by Ata et al.23 Furthermore, their threshold value had a sensitivity of 73.7%, specificity of 71.2%, and area under the curve of 0.75, which were similar to the diagnostic characteristics of the threshold, which had a sensitivity of 70.5%, specificity of 61.7% and area under the curve of 0.72.23 It should be noted that if the threshold that Ata et al determined was used with the authors’ patient population, 70.45% of patients who developed an atrial arrhythmia and 48.20% of patients who did not develop an atrial arrhythmia would be above the threshold. Additionally, in the authors’ population, the threshold would have an area under the curve of 0.61, limiting the predictive ability of the threshold. This is because of the higher average level of BNP found in both patients who did and did not develop atrial arrhythmias, when compared to previous studies. The authors’ population of patients who did not develop an atrial arrhythmia were found to have an average BNP concentration of 315.9 pg/mL and 557.5 pg/mL for those who did develop an atrial arrhythmia, which can be compared to the
Kaplan-Meier curves of time to intensive care unit (ICU) discharge with brain natriuretic peptide (BNP) threshold of 306 pg/mL.
615
BRAIN NATRIURETIC PEPTIDE AND CARDIAC SURGERY
Fig 4.
Kaplan-Meier curve of time to hospital discharge with brain natriuretic peptide (BNP) threshold of 306 pg/mL.
study population of Ata et al, who had an average BNP concentration of 65 pg/mL for those who did not develop AF and 226 pg/mL for those who did develop AF. Sezai et al had an average BNP concentration of 148.7 pg/mL for those who did not develop AF and 137.2 pg/mL for those who did develop AF.24 Although Sezai et al did not find a correlation between BNP and AF, they found a correlation between atrial natriuretic peptide, a peptide related to BNP that is also released from cardiac myocytes in response to stretch and AF.24 Additionally, the mean BNP concentration for the authors’ patients was much higher than what was found in the outpatient setting by Silvet et al, who reported concentrations of 131 pg/mL for those with AF and 49 pg/mL for those without AF.22 It was possible that the authors’ higher levels of BNP were because of the sicker patient population who visit the authors’ institution compared to other hospitals (although not published, this has been determined using the risk stratification system developed by 1 of the departments in the authors’ institution), and could help explain the difference in results found between the current study and that of Sezai et al.24,26 Although each increment of 10 increases in BNP was associated with only a small increase in risk of developing an atrial arrhythmia, the use of a threshold to predict those at risk does provide an estimated threshold for the clinician to anticipate which patient may develop an atrial arrhythmia.
Table 2. Mean and Median Lengths of ICU Stay for Above and Below BNP Threshold
Mean Median
Above BNP Threshold
Below BNP Threshold
6.615 Days (SE ¼ 0.197) 3 Days
3.0364 Days (SE ¼ 0.075) 1 Days
Abbreviations: BNP, brain natriuretic peptide; ICU, intensive care unit; SE, standard error in days.
Preoperative elevated BNP concentrations were associated with increased ICU and hospital stays. Using the calculated threshold value, the increase in the number of ICU and hospital days predicted by levels of BNP above the threshold were 3.6 and 7.9, respectively. It was possible that those patients with increased levels of BNP developed an atrial arrhythmia and then initiated treatment before discharge. Another important reason for delay in discharge can be complications like stroke or thrombotic events resulting from atrial arrhythmias, as well.5 Patients who do have longer lengths of ICU and hospital stays have greater hospital costs.7 Although not specifically related to AF, previous studies have determined the mean cost per day in the ICU for a nonventilated patient to be approximately $13,000.25 Thus, the care of patients with a BNP value above the threshold, who have a predicted 3.6 more days spent in the ICU, would cost an additional $46,800. There were several limitations to this study. It was an observational study, which may have resulted in selection and information bias. Every precaution was taken by including all cases with complete data over more than 5 years, and the results also were compared with imputed data for those individuals who had missing data and a difference between these results was not found, but no imputed data were used in the results presented in this study. This limited the possibility of selection bias. Patients with missing BNP values were not included in the study; it was possible that this population of
Table 3. Mean and Median Lengths of Hospital Stay for Above and Below BNP Threshold
Mean Median
Above BNP Threshold
Below BNP Threshold
17.366 (SE ¼ 0.473) 11
9.436 (SE ¼ 0.127) 7
Abbreviations: BNP, brain natriuretic peptide; SE, standard error in days.
616
BERNSTEIN ET AL
patients was different from those included in this study. Interestingly, the authors found an increased risk of atrial arrhythmias for those who had right-sided valve surgery compared to those who had left or no valve surgery. This increase could have been because of the limited number of individuals who had right-sided valve surgery. Additionally, this study was carried out at a single institution, and, thus, the patient population may not accurately represent the patient population of other hospitals performing CABG surgery. This study did, however, have strengths. This was by far the largest study that has evaluated BNP and atrial arrhythmias in cardiac surgical patients and demonstrated a relationship between the two. The database was quite rich and allowed
for the inclusion of many covariates into the analyses. This allowed the authors to explore the contribution of particular patient characteristics to atrial arrhythmias. In conclusion, elevated preoperative BNP concentrations predicted atrial arrhythmias following CABG surgery and increased lengths of ICU and hospital stays. This information potentially could be applied clinically to risk-stratify patients and to guide preventative strategies aimed at minimizing perioperative atrial arrhythmias. Further research is needed to determine the efficacy of clinical interventions aimed at either lowering BNP values or administering antiarrhythmic medications in the preoperative phase for at-risk individuals.
REFERENCES 1. DeFrances CJ, Lucas CA, Buie VC, et al: 2006 National hospital discharge survey. Natl Health Stat Report:1-20, 2008 2. Eagle KA, Guyton RA, Davidoff R, et al: ACC/AHA guidelines for coronary artery bypass graft surgery: executive summary and recommendations: A report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee to revise the 1991 guidelines for coronary artery bypass graft surgery). Circulation 100:1464-1480, 1999 3. Aranki SF, Shaw DP, Adams DH, et al: Predictors of atrial fibrillation after coronary artery surgery. Current trends and impact on hospital resources. Circulation 94:390-397, 1996 4. Cox JL: A perspective of postoperative atrial fibrillation in cardiac operations. Ann Thorac Surg 56:405-409, 1993 5. Creswell LL, Schuessler RB, Rosenbloom M, et al: Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 56:539-549, 1993 6. Leitch JW, Thomson D, Baird DK, et al: The importance of age as a predictor of atrial fibrillation and flutter after coronary artery bypass grafting. J Thorac Cardiovasc Surg 100:338-342, 1990 7. Mathew JP, Parks R, Savino JS, et al: Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. Multicenter Study of Perioperative Ischemia Research Group. JAMA 276:300-306, 1996 8. Mariscalco G, Engstrom KG: Atrial fibrillation after cardiac surgery: Risk factors and their temporal relationship in prophylactic drug strategy decision. Int J Cardiol 129:354-362, 2008 9. Mariscalco G, Engstrom KG: Postoperative atrial fibrillation is associated with late mortality after coronary surgery, but not after valvular surgery. Ann Thorac Surg 88:1871-1876, 2009 10. Mathew JP, Fontes ML, Tudor IC, et al: A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA 291:1720-1729, 2004 11. Mahoney EM, Thompson TD, Veledar E, et al: Costeffectiveness of targeting patients undergoing cardiac surgery for therapy with intravenous amiodarone to prevent atrial fibrillation. J Am Coll Cardiol 40:737-745, 2002 12. Chang SL, Chen YC, Chen YJ, et al: Mechanoelectrical feedback regulates the arrhythmogenic activity of pulmonary veins. Heart 93: 82-88, 2007 13. Kalifa J, Jalife J, Zaitsev AV, et al: Intra-atrial pressure increases rate and organization of waves emanating from the superior pulmonary veins during atrial fibrillation. Circulation 108:668-671, 2003
14. Davies MJ, Pomerance A: Pathology of atrial fibrillation in man. Br Heart J 34:520-525, 1972 15. Goette A, Juenemann G, Peters B, et al: Determinants and consequences of atrial fibrosis in patients undergoing open heart surgery. Cardiovasc Res 54:390-396, 2002 16. Hayashi H, Wang C, Miyauchi Y, et al: Aging-related increase to inducible atrial fibrillation in the rat model. J Cardiovasc Electrophysiol 13:801-808, 2002 17. McCullough PA, Nowak RM, McCord J, et al: B-type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: Analysis from Breathing Not Properly (BNP) Multinational Study. Circulation 106:416-422, 2002 18. Morita E, Yasue H, Yoshimura M, et al: Increased plasma levels of brain natriuretic peptide in patients with acute myocardial infarction. Circulation 88:82-91, 1993 19. Dao Q, Krishnaswamy P, Kazanegra R, et al: Utility of B-type natriuretic peptide in the diagnosis of congestive heart failure in an urgent-care setting. J Am Coll Cardiol 37:379-385, 2001 20. Omland T, Persson A, Ng L, et al: N-terminal pro-B-type natriuretic peptide and long-term mortality in acute coronary syndromes. Circulation 106:2913-2918, 2002 21. Goetze JP, Friis-Hansen L, Rehfeld JF, et al: Atrial secretion of B-type natriuretic peptide. Eur Heart J 27:1648-1650, 2006 22. Silvet H, Young-Xu Y, Walleigh D, et al: Brain natriuretic peptide is elevated in outpatients with atrial fibrillation. Am J Cardiol 92:1124-1127, 2003 23. Ata Y, Turk T, Ay D, et al: Ability of B-type natriuretic peptide in predicting postoperative atrial fibrillation in patients undergoing coronary artery bypass grafting. The Heart Surgery Forum 12: E211-E216, 2009 24. Sezai A, Hata M, Niino T, et al: Study of the factors related to atrial fibrillation after coronary artery bypass grafting: A search for a marker to predict the occurrence of atrial fibrillation before surgical intervention. J Thorac Cardiovasc Surg 137:895-900, 2009 25. Dasta JF, McLaughlin TP, Mody SH, et al: Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 33:1266-1271, 2005 26. Sessler DI, Sigl JC, Manberg PJ, et al: Broadly applicable risk stratification system for predicting duration of hospitalization and mortality. Anesthesiology 113:1026-1037, 2010
an atrial arrhythmia. Those above the threshold were 1.455 (95% CI: 1.087, 1.947) times more likely to have an atrial arrhythmia than those below the threshold. Individuals above the threshold had a median of 3 days in the intensive care unit compared to 1 day for those below the threshold, as well as median hospital stays of 11 days for those below the threshold. Conclusions: The results indicated that elevated brain natriuretic peptide was associated with increased risk of atrial arrhythmias and prolonged length of hospital stay after cardiac surgery. Identifying at-risk patients is important to guide preventative strategies for postoperative atrial arrhythmias. & 2015 Elsevier Inc. All rights reserved.
A
Identifying predictors of atrial arrhythmias in patients undergoing cardiac surgery can help clinicians identify candidates for preventive treatments, to reduce the morbidity and mortality associated with postoperative AF. The goals of this study were to evaluate whether preoperative BNP predicted atrial fibrillation post-CABG and to investigate if BNP was associated with prolonged ICU and hospital length of stay.
FTER CORONARY ARTERY BYPASS GRAFT (CABG) surgery, patients are at risk for the development of atrial arrhythmias, including atrial fibrillation, atrial flutter, supraventricular tachycardia, and premature atrial contractions. Atrial fibrillation (AF) is the most common cardiac rhythm disturbance, occurring in 25% to 40% of patients undergoing cardiac surgery.1,2 Postoperative atrial arrhythmias are associated with increased risk of morbidity, mortality, readmission to the intensive care unit (ICU), length of hospital and ICU stay, perioperative myocardial infarction, congestive heart failure, and stroke.3–11 Because of the challenges of predicting and preventing AF, there has been substantial interest in the use of biomarkers to identify patients who are at risk for atrial arrhythmias. One of these potential biomarkers is brain natriuretic peptide (BNP).12–16 BNP commonly is used for detecting congestive heart failure, but recently has been examined as a possible biomarker for atrial arrhythmias.11 BNP is released predominantly from the ventricular myocytes in response to left ventricular hypertrophy, which has been associated with AF.17–21 Previous attempts to determine an association between elevated preoperative BNP levels and atrial arrhythmias following CABG surgery have provided conflicting results. In a prospective trial of 144 consecutive surgeries, Ata et al found an association between preoperative BNP and postoperative AF (226 pg/mL, in those with AF v 65.2 pg/mL, in those without AF; p o 0.001).22,23 In contrast, Sezai et al24 did not show a statistically significant association between preoperative BNP and AF in 234 consecutive patients (137.2 ⫾ 215.7 pg/dL, in those with AF v 149.7 ⫾ 262.0 pg/dL, in those without AF; p ¼ 0.751). Thus, the relationship between BNP concentrations and AF is controversial. Furthermore, these studies had a limited sample size, and they were reported from very different surgical settings. Although conducted retrospectively, the authors’ data were collected prospectively at the time of surgery and will be the largest study to assess the relationship between BNP and AF.
KEY WORDS: arrhythmias, cardiac, coronary artery bypass, natriuretic peptide, brain
METHODS
With approval from both institutions’ Institutional Review Board and Research Subjects Review Board, a retrospective analysis of prospectively collected data from the Perioperative Health Documentation System that were acquired between 2005 and 2010 was performed. Data were collected from both
From the *Department of Outcomes Research, Cleveland Clinic, Cleveland, Ohio, †Uniformed Services University of the Health Sciences, Bethesda, Maryland, ‡Department of Public Health Sciences, University of Rochester, Rochester, New York, §University of Colorado Anschutz Medical Campus Division of Cardiology, Aurora, Colorado; and ║Department of Cardiothoracic Anesthesiology, Cleveland Clinic, Cleveland, Ohio. Supported by the National Institutes of Health/National Center for Research Resources CTSI award (grant number KL2 RR024136) to (T. P.) at the University of Rochester. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health. Support was also provided solely from institutional and/or departmental sources. None of the authors has a personal financial interest in this research. Address reprint requests to Alparslan Turan, MD, Department of Outcomes Research, Cleveland Clinic, 9500 Euclid Avenue, P-77, Cleveland, OH, 44195. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2014.11.015
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 3 (June), 2015: pp 611–616
611
612
BERNSTEIN ET AL
the hospital and anesthesia records, which integrated preoperative, intraoperative, and postoperative outcomes. Statistical Methods The authors assessed the relationship between preoperative BNP and atrial arrhythmias using unadjusted logistic regression, followed by a backwards selection multivariate logistic regression model, adjusting for unevenly distributed covariates. Covariates were selected based upon previously determined relationships between the covariate and postoperative AF. The selection included all potential covariates, the squares of the continuous covariates, and interactions between BNP and the covariates. The authors analyzed 29 predetermined potential covariates between patients who developed atrial arrhythmias and patients who did not, using a student’s t-test (for continuous variables), chi-square test (for dichotomous variables), or Wilcoxon rank sum test (for non-normal distributions). Any variable with a p value o 0.10, between those who did and did not develop an atrial arrhythmia, were considered to be covariates and adjusted for in the analysis. Although originally included in the analysis, postoperative medications were removed because of their use, possibly resulting from arrhythmias, and not a predictor of arrhythmias. A sensitivity analysis including all patients with BNP concentrations and imputed data for those with missing covariate data was performed; imputed data were not used in the final analysis. The authors determined a threshold BNP concentration to best predict atrial arrhythmias following CABG surgery, using a receiver operating characteristic (ROC) analysis and the Youden index (the sum of the sensitivity and specificity). Using the calculated threshold value, an unadjusted and adjusted multivariate logistic regression model, with the same covariates as the adjusted model for a one-unit increase in BNP, was performed to assess the risk of atrial arrhythmias in individuals above and below the BNP threshold. Kaplan-Meier analyses, using the calculated threshold value, were conducted to assess the lengths of ICU and hospital stay. Wilcoxon and log-rank tests were used to compare the lengths of stay between those above and below the BNP threshold. Lengths of ICU and hospital stay were defined as the duration of stay from the date of surgery to the date of ICU and hospital discharge, respectively. SAS software version 9.2.2 (SAS Institute, Cary, NC) and PASS 11 software for Windows (NCSS, Kaysville, UT) were used for all statistical analyses.
Those who did not develop an atrial arrhythmia were found to have an average BNP concentration of 315.9 pg/mL, and those who did develop an atrial arrhythmia had an average concentration of 557.5 pg/mL. For a 10-unit increase in BNP, the unadjusted odds ratio for an atrial arrhythmia was 1.0005 (95% CI: 1.00047, 1.00054; p o 0.001). For the adjusted model, the odds ratio was (95% CI, p value) when adjusting for baseline BNP, age, body mass index, ejection fraction, race, sex, American Society of Anesthesiologists physical status classification, creatinine, smoking, preoperative AF, chronic obstructive pulmonary disease/asthma, preoperative statin, preoperative beta-blocker, preoperative anti-arrhythmic medication, length of surgery, valve surgery, creatinine squared, BNP squared, interaction of BNP and length of surgery, and an interaction of BNP and valve surgery. The adjusted odds ratio for a 10-unit increase in BNP is 1.005 (95% CI: 1.002, 1.009; standard error: 0.000291; p ¼ 0.004) for patients who did not have valve surgery and had the mean length of surgery. The calculated threshold value was 306 pg/mL, with a sensitivity of 61.7% and specificity of 70.5%, for a Youden index of 1.322. The area under the curve for the ROC analysis was 0.718 (Fig 2). The positive predictive value of the
RESULTS
The authors evaluated 18,064 patients who underwent cardiac surgery between 2005 and 2010. Of these patients, 9,378 were excluded because they did not have preoperative BNP concentrations, and an additional 3,887 were removed because of missing covariate data or being statistical outliers. The remaining 4,765 patients were included in the analysis (Fig 1). The baseline characteristics (Table 1) for individuals who developed atrial arrhythmias and those who did not were found to be different for all but 3 covariates: Smoking, preoperative atrial flutter, and preoperative nonsteroidal antiinflammatory drug use.
Fig 1. Flow chart of included and excluded patients from the authors’ institutions’ Perioperative Documentation System of cardiac surgeries 2005-2010.
613
BRAIN NATRIURETIC PEPTIDE AND CARDIAC SURGERY
Table 1. Baseline Characteristics of the Population of Patients with Preoperative BNP Values Variable
BNP, mean (median [Q1, Q3]), pg/mL Age, mean (median, Q1, Q3), yr BMI, mean (median, Q1, Q3), kg/m2 Male, n (%) Race, n (%) African-American, n (%) Asian/Pacific Islander, n (%) Caucasian, n (%) Hispanic/Latino, n (%) Other, n (%) ASA status,‡ n (%) ASA status 1, 2, or 3, n (%) ASA status 4, n (%) ASA status 5, n (%) Smoking, n (%) Current/former smoker, n (%) Never smoked, n (%) History of congestive heart failure, n (%) Hypertension, n (%) Preoperative atrial flutter, n (%) Preoperative atrial fibrillation, n (%) History of myocardial infarction, n (%) MI o1 week, n (%) MI 41 week, o3 months, n (%) MI 3-6 months, n (%) MI 46 months, n (%) No History of MI, n (%) Diabetes, n (%) Peripheral vascular disease, n (%) Alcohol consumption, n (%) Ejection fraction, mean (SD), % COPD/asthma, n (%) Creatinine, mean (median, Q1, Q3), mg/dL Troponin, mean (median, Q1, Q3), ng/mL Preoperative statin, n (%) Preoperative beta-blocker, n (%) Preoperative antiarrhythmic, n (%) Preoperative ACE inhibitor, n (%) Preoperative steroids, n (%) Preoperative NSAID, n (%) Preoperative nitrate, n (%) Length of surgery, mean (SD), H:MM Postoperative amiodarone, n (%) Postoperative dopamine, n (%) Postoperative epinephrine, n (%) Valve surgery, n (%) No Left Right Both
Atrial Arrhythmia (n ¼ 1,226)
No Atrial Arrhythmia (n ¼ 3,548)
557.5 70.0 28.8 735
(265 [111, 671]) (70 [62, 78]) (27.7 [24.3, 32.0]) (59.95)
315.9 63.5 28.1 2283
(122.5 [45, 322]) (65 [74, 65]) (27.2 2 [4.2, 31.2]) (64.35)
63 9 1118 22 14
(5.1) (0.73) (91.2) (1.8) (1.1)
219 34 3164 54 77
(6.2) (0.96) (89.2) (1.5) (2.2)
p Value*
o0.001† o0.001† o0.001† 0.006 0.0898
o0.001 99 (8.1) 1088 (88.7) 39 (3.2)
577 (16.3) 2892 (81.5) 79 (1.6) 0.005
677 549 594 929 25 457
(55.2) (44.8) (48.4) (75.8) (2.0) (37.3)
1793 1755 1193 2634 56 863
(50.5) (49.5) (33.6) (66.6) (1.58) (24.3)
213 65 119 8 821 397 209 131 46.3 245 1.37 0.26 554 614 234 385 129 16 155 7:15 255 17 785
(17.4) (5.3) (9.7) (0.65) (67.0) (32.4) (17.1) (10.7) (15.48) (20.0) (1.1, 0.9, 1.4) (0.00 [0.00, 0.03]) (45.2) (50.08) (19.1) (31.4) (10.5) (1.31) (12.6) (1:53) (20.8) (1.39) (64.0)
452 148 200 24 2724 850 405 468 50.8 451 1.17 0.15 1507 1465 554 1050 265 46 323 6:29 167 17 1487
(12.7) (5.3) (5.6) (0.68) (76.8) (24.0) (11.4) (13.2) (13.53) (12.7) (1.0, 0.8, 1.2) (0.00 0 [.00, 0.01]) (42.5) (41.3) (15.61) (29.6) (7.5) (1.30) (9.1) (1:41) (4.7) (0.48) (41.9)
336 679 36 175
(27.4) (55.4) (1.72) (14.3)
1047 2141 61 299
(29.5) (60.3) (1.72) (8.4)
o0.001 o0.001 0.2815 o0.001 o0.001
o0.001 o0.001 0.022 o0.001 o0.001 o0.001† o0.001† 0.0983 o0.001 0.005 0.2337 o0.001 0.9818 o0.001 o0.001 o0.001 0.001 o0.001 o0.001
Abbreviations: ACE, angiotensin-converting enzyme; BMI, body mass index; BNP, brain natriuretic peptide; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; NSAID, nonsteroidal anti-inflammatory drug. *p values calculated using t-test unless otherwise noted. †p value calculated using the Wilcoxon signed-rank test. ‡ASA status refers to the American Society of Anesthesiologists (ASA) physical status classification.
threshold was 41.9%, and the negative predictive value was 84.2%. Using the threshold resulted in 1,490 individuals with an elevated BNP and 3,275 individuals with a nonelevated BNP.
The unadjusted odds ratio was 2.367 (95% CI: 2.068, 2.709; p o 0.001), and the adjusted odds ratio was 1.455 (95% CI: 1.087, 1.947; p ¼ 0.031) for individuals above the threshold of 306 pg/mL who had no valve surgery and the mean length of
614
BERNSTEIN ET AL
DISCUSSION
Fig 2. Receiver operating characteristic (ROC) curve for the calculated threshold value of 306 pg/mL, Area under the curve ¼ 0.72.
surgery compared to individuals below the threshold who had no valve surgery and the mean length of surgery. BNP concentrations above the threshold were significantly associated with increased lengths of ICU (Fig 3) and hospital stay (Fig 4). Using both the log-rank and Wilcoxon tests to compare the curves, a significant difference (p o 0.001) was found in the time to ICU (Table 2) and hospital discharge (Table 3) between those with a BNP value above the threshold compared to those with a BNP value below the threshold. Table 3
Fig 3.
The results of this study indicated that increased levels of preoperative BNP were associated with an increased risk of atrial arrhythmias following CABG surgery. The authors found an odds ratio of 1.005 (95% CI: 1.002, 1.009; p ¼ 0.003) for a 10-unit change in BNP, and for individuals above the calculated threshold of 306 pg/mL, an odds ratio of 1.455 (95% CI: 1.087, 1.947; p ¼ 0.031). Those above the threshold also were found to have increased lengths of ICU and hospital stay. Elevated BNP can be used as an identifier of those at risk for the development of an atrial arrhythmia and may benefit from prophylactic therapy with antiarrhythmic drugs.25 The threshold value determined by this study was higher than the 135-pg/mL threshold previously reported by Ata et al.23 Furthermore, their threshold value had a sensitivity of 73.7%, specificity of 71.2%, and area under the curve of 0.75, which were similar to the diagnostic characteristics of the threshold, which had a sensitivity of 70.5%, specificity of 61.7% and area under the curve of 0.72.23 It should be noted that if the threshold that Ata et al determined was used with the authors’ patient population, 70.45% of patients who developed an atrial arrhythmia and 48.20% of patients who did not develop an atrial arrhythmia would be above the threshold. Additionally, in the authors’ population, the threshold would have an area under the curve of 0.61, limiting the predictive ability of the threshold. This is because of the higher average level of BNP found in both patients who did and did not develop atrial arrhythmias, when compared to previous studies. The authors’ population of patients who did not develop an atrial arrhythmia were found to have an average BNP concentration of 315.9 pg/mL and 557.5 pg/mL for those who did develop an atrial arrhythmia, which can be compared to the
Kaplan-Meier curves of time to intensive care unit (ICU) discharge with brain natriuretic peptide (BNP) threshold of 306 pg/mL.
615
BRAIN NATRIURETIC PEPTIDE AND CARDIAC SURGERY
Fig 4.
Kaplan-Meier curve of time to hospital discharge with brain natriuretic peptide (BNP) threshold of 306 pg/mL.
study population of Ata et al, who had an average BNP concentration of 65 pg/mL for those who did not develop AF and 226 pg/mL for those who did develop AF. Sezai et al had an average BNP concentration of 148.7 pg/mL for those who did not develop AF and 137.2 pg/mL for those who did develop AF.24 Although Sezai et al did not find a correlation between BNP and AF, they found a correlation between atrial natriuretic peptide, a peptide related to BNP that is also released from cardiac myocytes in response to stretch and AF.24 Additionally, the mean BNP concentration for the authors’ patients was much higher than what was found in the outpatient setting by Silvet et al, who reported concentrations of 131 pg/mL for those with AF and 49 pg/mL for those without AF.22 It was possible that the authors’ higher levels of BNP were because of the sicker patient population who visit the authors’ institution compared to other hospitals (although not published, this has been determined using the risk stratification system developed by 1 of the departments in the authors’ institution), and could help explain the difference in results found between the current study and that of Sezai et al.24,26 Although each increment of 10 increases in BNP was associated with only a small increase in risk of developing an atrial arrhythmia, the use of a threshold to predict those at risk does provide an estimated threshold for the clinician to anticipate which patient may develop an atrial arrhythmia.
Table 2. Mean and Median Lengths of ICU Stay for Above and Below BNP Threshold
Mean Median
Above BNP Threshold
Below BNP Threshold
6.615 Days (SE ¼ 0.197) 3 Days
3.0364 Days (SE ¼ 0.075) 1 Days
Abbreviations: BNP, brain natriuretic peptide; ICU, intensive care unit; SE, standard error in days.
Preoperative elevated BNP concentrations were associated with increased ICU and hospital stays. Using the calculated threshold value, the increase in the number of ICU and hospital days predicted by levels of BNP above the threshold were 3.6 and 7.9, respectively. It was possible that those patients with increased levels of BNP developed an atrial arrhythmia and then initiated treatment before discharge. Another important reason for delay in discharge can be complications like stroke or thrombotic events resulting from atrial arrhythmias, as well.5 Patients who do have longer lengths of ICU and hospital stays have greater hospital costs.7 Although not specifically related to AF, previous studies have determined the mean cost per day in the ICU for a nonventilated patient to be approximately $13,000.25 Thus, the care of patients with a BNP value above the threshold, who have a predicted 3.6 more days spent in the ICU, would cost an additional $46,800. There were several limitations to this study. It was an observational study, which may have resulted in selection and information bias. Every precaution was taken by including all cases with complete data over more than 5 years, and the results also were compared with imputed data for those individuals who had missing data and a difference between these results was not found, but no imputed data were used in the results presented in this study. This limited the possibility of selection bias. Patients with missing BNP values were not included in the study; it was possible that this population of
Table 3. Mean and Median Lengths of Hospital Stay for Above and Below BNP Threshold
Mean Median
Above BNP Threshold
Below BNP Threshold
17.366 (SE ¼ 0.473) 11
9.436 (SE ¼ 0.127) 7
Abbreviations: BNP, brain natriuretic peptide; SE, standard error in days.
616
BERNSTEIN ET AL
patients was different from those included in this study. Interestingly, the authors found an increased risk of atrial arrhythmias for those who had right-sided valve surgery compared to those who had left or no valve surgery. This increase could have been because of the limited number of individuals who had right-sided valve surgery. Additionally, this study was carried out at a single institution, and, thus, the patient population may not accurately represent the patient population of other hospitals performing CABG surgery. This study did, however, have strengths. This was by far the largest study that has evaluated BNP and atrial arrhythmias in cardiac surgical patients and demonstrated a relationship between the two. The database was quite rich and allowed
for the inclusion of many covariates into the analyses. This allowed the authors to explore the contribution of particular patient characteristics to atrial arrhythmias. In conclusion, elevated preoperative BNP concentrations predicted atrial arrhythmias following CABG surgery and increased lengths of ICU and hospital stays. This information potentially could be applied clinically to risk-stratify patients and to guide preventative strategies aimed at minimizing perioperative atrial arrhythmias. Further research is needed to determine the efficacy of clinical interventions aimed at either lowering BNP values or administering antiarrhythmic medications in the preoperative phase for at-risk individuals.
REFERENCES 1. DeFrances CJ, Lucas CA, Buie VC, et al: 2006 National hospital discharge survey. Natl Health Stat Report:1-20, 2008 2. Eagle KA, Guyton RA, Davidoff R, et al: ACC/AHA guidelines for coronary artery bypass graft surgery: executive summary and recommendations: A report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee to revise the 1991 guidelines for coronary artery bypass graft surgery). Circulation 100:1464-1480, 1999 3. Aranki SF, Shaw DP, Adams DH, et al: Predictors of atrial fibrillation after coronary artery surgery. Current trends and impact on hospital resources. Circulation 94:390-397, 1996 4. Cox JL: A perspective of postoperative atrial fibrillation in cardiac operations. Ann Thorac Surg 56:405-409, 1993 5. Creswell LL, Schuessler RB, Rosenbloom M, et al: Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 56:539-549, 1993 6. Leitch JW, Thomson D, Baird DK, et al: The importance of age as a predictor of atrial fibrillation and flutter after coronary artery bypass grafting. J Thorac Cardiovasc Surg 100:338-342, 1990 7. Mathew JP, Parks R, Savino JS, et al: Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. Multicenter Study of Perioperative Ischemia Research Group. JAMA 276:300-306, 1996 8. Mariscalco G, Engstrom KG: Atrial fibrillation after cardiac surgery: Risk factors and their temporal relationship in prophylactic drug strategy decision. Int J Cardiol 129:354-362, 2008 9. Mariscalco G, Engstrom KG: Postoperative atrial fibrillation is associated with late mortality after coronary surgery, but not after valvular surgery. Ann Thorac Surg 88:1871-1876, 2009 10. Mathew JP, Fontes ML, Tudor IC, et al: A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA 291:1720-1729, 2004 11. Mahoney EM, Thompson TD, Veledar E, et al: Costeffectiveness of targeting patients undergoing cardiac surgery for therapy with intravenous amiodarone to prevent atrial fibrillation. J Am Coll Cardiol 40:737-745, 2002 12. Chang SL, Chen YC, Chen YJ, et al: Mechanoelectrical feedback regulates the arrhythmogenic activity of pulmonary veins. Heart 93: 82-88, 2007 13. Kalifa J, Jalife J, Zaitsev AV, et al: Intra-atrial pressure increases rate and organization of waves emanating from the superior pulmonary veins during atrial fibrillation. Circulation 108:668-671, 2003
14. Davies MJ, Pomerance A: Pathology of atrial fibrillation in man. Br Heart J 34:520-525, 1972 15. Goette A, Juenemann G, Peters B, et al: Determinants and consequences of atrial fibrosis in patients undergoing open heart surgery. Cardiovasc Res 54:390-396, 2002 16. Hayashi H, Wang C, Miyauchi Y, et al: Aging-related increase to inducible atrial fibrillation in the rat model. J Cardiovasc Electrophysiol 13:801-808, 2002 17. McCullough PA, Nowak RM, McCord J, et al: B-type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: Analysis from Breathing Not Properly (BNP) Multinational Study. Circulation 106:416-422, 2002 18. Morita E, Yasue H, Yoshimura M, et al: Increased plasma levels of brain natriuretic peptide in patients with acute myocardial infarction. Circulation 88:82-91, 1993 19. Dao Q, Krishnaswamy P, Kazanegra R, et al: Utility of B-type natriuretic peptide in the diagnosis of congestive heart failure in an urgent-care setting. J Am Coll Cardiol 37:379-385, 2001 20. Omland T, Persson A, Ng L, et al: N-terminal pro-B-type natriuretic peptide and long-term mortality in acute coronary syndromes. Circulation 106:2913-2918, 2002 21. Goetze JP, Friis-Hansen L, Rehfeld JF, et al: Atrial secretion of B-type natriuretic peptide. Eur Heart J 27:1648-1650, 2006 22. Silvet H, Young-Xu Y, Walleigh D, et al: Brain natriuretic peptide is elevated in outpatients with atrial fibrillation. Am J Cardiol 92:1124-1127, 2003 23. Ata Y, Turk T, Ay D, et al: Ability of B-type natriuretic peptide in predicting postoperative atrial fibrillation in patients undergoing coronary artery bypass grafting. The Heart Surgery Forum 12: E211-E216, 2009 24. Sezai A, Hata M, Niino T, et al: Study of the factors related to atrial fibrillation after coronary artery bypass grafting: A search for a marker to predict the occurrence of atrial fibrillation before surgical intervention. J Thorac Cardiovasc Surg 137:895-900, 2009 25. Dasta JF, McLaughlin TP, Mody SH, et al: Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 33:1266-1271, 2005 26. Sessler DI, Sigl JC, Manberg PJ, et al: Broadly applicable risk stratification system for predicting duration of hospitalization and mortality. Anesthesiology 113:1026-1037, 2010
