- Email: [email protected]
Predictors of Low Cardiac Output Syndrome After Isolated Coronary Artery Bypass Surgery: Trends Over 20 Years
ADULT CARDIAC
Predictors of Low Cardiac Output Syndrome After Isolated Coronary Artery Bypass Surgery: Trends Over 20 Years Khaled D. Algarni, MD, MHS, Manjula Maganti, MS, and Terrence M. Yau, MD, MS Division of Cardiovascular Surgery, Peter Munk Cardiac Center, Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada
Background. Postoperative low cardiac output syndrome (LCOS) is associated with high morbidity and mortality after coronary artery bypass grafting (CABG). We sought to examine trends in predictors of LCOS after isolated CABG. Methods. A total of 25,176 consecutive patients who underwent isolated CABG between 1990 and 2009 were included. To examine trends over time, we divided patients into four equivalent eras (1990 to -1994, n ⴝ 6,489; 1995 to 1999, n ⴝ 8,175; 2000 to 2004, n ⴝ 6,741; 2005 to 2009, n ⴝ 3,797). We used multivariable analysis to identify predictors of LCOS. Results. The prevalence of LCOS declined from 9.1% (1990 to 1994) to 2.4% (2005 to 2009, p < 0.001). The following were the major independent predictors of LCOS for the entire cohort (odds ratios in parentheses): reoperative CABG (4.1); earlier year of operation (4.1, 2.6, 1.7 for the first, second, and third eras, respectively); left
ventricular ejection fraction (LVEF) less than 0.20 (3.5), emergency surgery (2.7), cardiogenic shock (2.3), female gender (2), and LVEF 0.20 to 0.39 (2). Unlike other risk factors, the impact of LVEF less than 0.20 on development of postoperative LCOS increased substantially in the latest era (odds ratio, 7.8) compared with (odds ratios, 3.1, 4.3, and 3.2) the first, second, and third eras, respectively. Conclusions. The impact of LVEF less than 0.20 on development of postoperative LCOS has increased markedly in the latest era of our study. Prudent preoperative evaluation in patients with severe left ventricular dysfunction is critical. Further innovative research in myocardial protection and circulatory support is warranted in patients with severe left ventricular dysfunction.
T
Patients and Methods
he development of low cardiac output syndrome (LCOS) after coronary artery bypass grafting (CABG) surgery is associated with a 10-fold to 17-fold increase in mortality and markedly increased morbidity [1].The reported prevalence of LCOS after isolated CABG is 3% to 14% [1–3]. The prevalence and the independent predictors of LCOS were described previously by Rao and associates [1]. However, no studies have examined temporal changes in predictors of LCOS in patients undergoing isolated CABG surgery. The purpose of this study was to examine trends over time in the prevalence of LCOS in patients undergoing CABG surgery over a 20-year period, as well as to identify changes in the independent predictors of LCOS with time. These findings may assist in identification of patients at high risk of postoperative LCOS and therefore facilitate preoperative planning and patient counselling, and guide potential preoperative interventions for optimization of high-risk subgroups.
Accepted for publication June 8, 2011. Address correspondence to Dr Yau, Division of Cardiovascular Surgery, Toronto General Hospital, 4N-470, 200 Elizabeth St, Toronto, Ontario, Canada, M5G 2C4; e-mail: [email protected].
© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2011;92:1678 – 85) © 2011 by The Society of Thoracic Surgeons
Data Collection and Study Outcomes Clinical, operative, and outcome data were collected prospectively in a computerized database for 25,176 consecutive patients undergoing isolated CABG at our institution between January 1, 1990 and December 31, 2009. Patients undergoing CABG with concomitant procedures were excluded. To evaluate changing patterns of risk factors and outcomes over time, the study population (n ⫽ 25,202) was divided into 4 eras (1990 to 1994, n ⫽ 6,489; 1995 to 1999, n ⫽ 8,175; 2000 to 2004, n ⫽ 6,741, and 2005 to 2009, n ⫽ 3,797). The study was approved by our Institutional Research Ethics board; individual patient consent was waived. Our primary outcome was LCOS, which was diagnosed if the patient required an intraaortic balloon pump (IABP) in the operating room to be weaned from cardiopulmonary bypass or in the intensive care unit because of hemodynamic compromise. The LCOS was also diagnosed if the patient required inotropic medication (either dopamine, dobutamine, milrinone, or epinephrine) to maintain a systolic blood pressure greater than 90 mm Hg and cardiac output greater than 2.2 L · min⫺1 · m⫺2 for at least 30 minutes in the intensive care unit, after optimizing preload, and afterload as well as correcting all 0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.06.017
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1679
Table 1. Demographics and Clinical Profile Variable
1990–1994
1995–1999
2000–2004
2005–2009
No. of patients Age (years) mean ⫾ SD Less than 65 65–74 ⱖ75 Female Diabetes mellitus Hypertension Hyperlipidemia Angina Stable Crescendo ACS Preoperative MI Cardiogenic shock Congestive heart failure Preoperative IABP LVEF ⬎0.60 0.40–0.60 0.20–0.40 ⬍0.20 Left main disease Triple-vessel disease Preoperative stroke PVD Reoperative CABG Previous PCI ⫾ stents Urgency of surgery Elective Semiurgent Urgent Emergency
6,489 61.6 ⫾ 9.7 58.7 33.8 7.5 20 23.7 50.2 52.5
8,175 62.9 ⫾ 9.9 52.3 35.9 11.8 22.2 27.8 53.7 65.9
6,741 64.4 ⫾ 9.9 47.6 36.5 15.9 20.4 33.3 64.5 83
3,797 65.8 ⫾ 10 43.6 34.9 21.5 21.9 39.1 76.4 90.2
p Value ⬍0.0001 ⬍0.0001
0.003 ⬍0.0001 ⬍0.0001 ⬍0.0001
37.1 33.6 27.3 16.3 1.2 8.3 3.5
32.2 30 35.8 20.3 1.2 8.9 3.5
39 33.8 23.3 22 1.1 10.1 3.1
35.1 29.5 25.5 23.7 1.5 10.1 3.1
⬍0.0001
32.2 43.6 21.2 3.1 18.2 72.7 8.3 14 7.2 10.3
31 45.7 20.5 2.9 18.3 74.7 8.1 14.7 5 9.2
40.8 38 18.8 2.5 26.3 80.3 8.5 18.7 3.5 9.1
53.6 31 14.3 1.2 34.9 82 9.5 20.4 2.1 11
⬍0.0001
57.3 20.6 15.7 2
48.1 36.2 11.1 0.7
51.7 36.6 8.5 0.6
49.5 36.4 9.8 0.6
⬍0.0001
⬍0.0001 0.2 0.006 0.4
⬍0.0001 ⬍0.0001 0.1 ⬍0.0001 ⬍0.0001 0.002
ACS ⫽ acute coronary syndrome; CABG ⫽ coronary artery bypass grafting; IABP ⫽ intraaortic balloon pump; LVEF ⫽ left ventricular ejection fraction; MI ⫽ myocardial infarction; PCI ⫽ percutaneous coronary intervention; PVD ⫽ peripheral vascular disease.
electrolyte and blood gas abnormalities. Patients who required a renal dose of dopamine (⬍4 g/kg) or those who received vasoconstricting medications to increase peripheral vascular resistance in the presence of normal or high cardiac output (ⱖ 2.5 L · min⫺1 · m⫺2) were not considered to have LCOS.
Statistical Analysis Statistical analysis was performed with SAS 9.2 software (SAS Institute, Inc, Cary, NC). We used the 2 or Fisher exact test to analyze categoric data. Categoric variables were expressed as percentages. Analysis of variance was used to analyze continuous variables that are normally distributed, and the Kruskal-Wallis test was used for variables with a nonparametric distribution. Continuous variables were expressed as means ⫾ standard deviations. Statistical significance was based on two-sided p values less than 0.05. Variables that had a univariate p
value of less than 0.25 and those of biological importance were selected for inclusion in a multivariable logistic regression model. Multivariable analysis was performed for the entire cohort of 25,176 patients as well as for each era separately to determine the independent predictors of LCOS for the entire cohort and for each era separately. Model discrimination and calibration were evaluated by the area under the receiver operating characteristic curve, and the Hosmer–Lemeshow goodness-of-fit statistic, respectively.
Results Demographics and Baseline Clinical Profile Preoperative patient characteristics are detailed in Table 1. The prevalence of most risk factors increased steadily with time. About 20% of patients were 75 years and older
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1680 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
Ann Thorac Surg 2011;92:1678 – 85
Table 2. Operative Data and Postoperative Outcomes Variable Intraoperative variables LITA use RITA use Radial artery usea No. of grafts CPB time (minutes) Cross-clamping time (minutes) Cardioplegia technique Intermittent cold antegrade Warm or tepid cardioplegia Off pump Postoperative outcomes Mortality Inotrope use ⬎30 minutes Low-output syndrome IABP Myocardial infarction Stroke Renal failure Pulmonary complications Sternal wound infection Superficial Deep Sepsis Duration of ventilation (hours) Intensive care unit stay (hours) Hospital stay (days) a
1990–1994
1995–1999
2000–2004
2005–2009
p Value
83 2.1 — 3.5 ⫾ 0.9 86.9 ⫾ 26.8 60.5 ⫾ 18.4
89.6 3.4 2.5 3.6 ⫾ 0.9 85.2 ⫾ 26 63.5 ⫾ 18.4
88.9 8.1 3.7 3.5 ⫾ 0.9 91.8 ⫾ 25.7 68.8 ⫾ 20.8
91.8 7.2 1.9 3.4 ⫾ 0.9 89 ⫾ 23.3 68.1 ⫾ 21.3
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
76 14.6 0.2
67.6 16.6 3
80.9 4.9 6.2
86.8 0.4 9.5
⬍0.0001
2.5 31.2 5.1 4.9 3.2 1.5 0.8 8.9
1.7 30.8 4 4.9 2.5 1.4 0.9 8.5
1.1 39.7 2.6 3.2 2.9 0.9 0.9 7.5
1.4 37.1 1.8 2.2 1.8 0.9 1.3 10.2
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.001 0.01 0.07 ⬍0.0001
2.1 0.9 0.6 26.7 ⫾ 61.2 56.4 ⫾ 79 10 ⫾ 9.4
1.5 0.9 0.7 14.1 ⫾ 52.1 43.3 ⫾ 90.8 8 ⫾ 8.9
1.5 0.6 1.3 16.9 ⫾ 58 51.9 ⫾ 116 8 ⫾ 9.9
1.2 0.6 1.6 16.8 ⫾ 59.3 53.6 ⫾ 100 7.9 ⫾ 7.6
0.006 0.001 ⬍0.0001 ⬍0.0001 ⬍0.0001
Radial artery data was not collected during the first time cohort (1991–1996).
CPB ⫽ cardiopulmonary bypass;
IABP ⫽ intraaortic balloon pump;
in the latest time cohort versus 7.5% of patients in the first one. The prevalence of left main disease doubled between the first and the last time cohort (18.2% vs 34.9%). In contrast, the prevalence of left ventricular dysfunction and reoperative CABG has declined significantly with time.
Operative and Hospital Outcomes Operative profiles and postoperative outcomes are detailed in Table 2. The overall prevalence of LCOS for the entire cohort was 5.7% (n ⫽ 1,431). The prevalence of LCOS declined from 9.1% (n ⫽ 588) in the first time cohort to 2.4% (n ⫽ 91) in the last one (p ⬍ 0.0001). In-hospital mortality declined from 2.5% in the first era to 1.4% in the last one (p ⬍ 0.0001). The prevalence of most postoperative complications declined significantly with time. However, there was a trend of increasing prevalence of sepsis with time. The duration of ventilation and intensive care unit stay has fluctuated with time. Nonetheless, the length of hospital stay decreased significantly with time.
Postoperative Outcomes Stratified by LCOS Table 3 details the postoperative outcomes in patients who developed LCOS compared with those who did not.
LITA ⫽ left internal thoracic artery;
RITA ⫽ right internal thoracic artery.
Across the four eras, operative mortality was higher in patients with LCOS (17.7% vs 0.9%, 18.5% vs 0.6%, 12.8 vs 0.6%, and 24% vs 0.8%), respectively (p ⬍ 0.0001). Similarly, patients with LCOS had significantly longer ventilatory support, intensive care unit stay, hospital stay, and more blood transfusion. Moreover, the risk of mortality in patients with LCOS increased significantly in the latest era, where 22 out of 91 patients with LCOS died in the latest era (24%) compared with mortality of 17.7%, 18.5%, and 12.8% for the LCOS group in the first, second, and third eras, respectively. Thus LCOS was associated with a 29-fold increase in mortality in the latest era, compared with a 17-fold increase in the first era of this study.
Predictors of LCOS Predictors of LCOS are detailed in Table 4. Reoperative CABG (odds ratio [OR], 4.1) and the first era of surgery (OR, 4.1) carried the highest odds of LCOS, followed by LVEF less than 0.20 (OR, 3.5) and then emergency CABG (OR, 2.7). Table 5 presents the odds ratios for independent predictors of LCOS across the four eras of the study. The predictors of LCOS have changed significantly with time. In the latest era, only four variables emerged as significant predictors of LCOS; LVEF less than 0.20 (OR, 7.8),
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1681
Table 3. Outcomes Stratified by Low Cardiac Output Syndrome Statusa Variable Mortality LCOS % (n) No LCOS % (n) Intensive care unit stay (hours) LCOS No LCOS Duration of ventilation (hours) LCOS No LCOS Postoperative hospital stay (days) LCOS No LCOS a
1990–1994
1995–1999
2000–2004
2005–2009
17.7 (588) 95 (5,892)
18.5 (487) 61 (7,677)
12.8 (265) 56 (6,471)
24.2 (91) 84 (3,705)
111.3 ⫾ 134 50.8 ⫾ 68.5
141.7 ⫾ 224 37.3 ⫾ 70.9
205 ⫾ 440 45.6 ⫾ 73.2
200 ⫾ 251 49.9 ⫾ 90
66.8 ⫾ 111 22.8 ⫾ 52.3
75.3 ⫾ 152.2 10.3 ⫾ 37.7
109 ⫾ 165.4 13.2 ⫾ 45.1
124.6 ⫾ 194 14 ⫾ 47.5
13.4 ⫾ 12.7 9.6 ⫾ 8.9
15.1 ⫾ 20.8 7.5 ⫾ 7.1
18.2 ⫾ 37.7 7.5 ⫾ 6.2
16.1 ⫾ 21.8 7.6 ⫾ 6.7
p ⬍ 0.001 for each individual grouping variable and across the four eras.
LCOS ⫽ low cardiac output syndrome.
reoperative CABG (OR, 5.4), LVEF 0.20 to 0.39 (OR, 4), and emergent-urgent surgery (OR, 2.5). Figure 1 shows that the observed frequency of LCOS declined with time in all high-risk categories except patients with LVEF less than 0.20. In these patients with severe LV dysfunction, the prevalence of LCOS increased from 10.7% in the third era of this study to 18.2% in the latest era.
LVEF Less Than 0.20 Unlike other predictive variables, the prevalence of LCOS increased significantly in patients with LVEF less than 0.20 in the most recent era (18.2%) compared with the third era of this study (10.7%). We used a Cochrane Mantel-Haenszel test to compare the odds of developing LCOS in patients with LVEF less than 0.20 across the four eras. The odds ratio for LCOS in patients with LVEF less
Table 4. Multivariable Predictors of Low Cardiac Output Syndrome (n ⫽ 25,176) Variable Constant Agea Female gender Diabetes Cardiogenic shock Congestive heart failure Peripheral vascular disease Reoperative CABG Left main disease LVEFb ⬍ 0.20 0.20–0.40 0.40–0.60 Urgency of surgery Same hospitalization Urgent Emergent Year of operation 1991–1996c 1997–2002c 1997–2002c
n
Estimate ⫾ SE
Odds Ratio
95% CI
p Value
— 5,322 7,539 305 2,325 4,136 1,259 5,775
⫺70.26 0.015 ⫾ 0.003 0.70 ⫾ 0.06 0.17 ⫾ 0.06 0.58 ⫾ 0.16 0.63 ⫾ 0.08 0.15 ⫾ 0.07 1.41 ⫾ 0.09 0.26 ⫾ 0.07
1.016 2 1.2 2.3 1.7 1.2 4.1 1.3
1.01–1.02 1.8–2.3 1.1–1.3 1.7–3.2 1.4–2.3 1.01–1.3 3.5–4.8 1.1–1.5
⬍0.0001 ⬍0.0001 0.006 ⬍0.0001 ⬍0.0001 0.04 ⬍0.0001 ⬍0.0001
646 4,847 10,281
1.27 ⫾ 0.14 0.71 ⫾ 0.09 0.23 ⫾ 0.08
3.5 2.0 1.3
2.7–4.6 1.7–2.4 1.1–1.5
⬍0.0001 ⬍0.0001 0.003
751 2,622 247
0.41 ⫾ 0.14 0.53 ⫾ 0.1 1 ⫾ 0.2
1.5 1.7 2.7
1.1–2 1.4–2.1 1.8–4
0.005 ⬍0.0001 ⬍0.0001
6,489 8,175 6,741
1.4 ⫾ 0.12 0.96 ⫾ 0.12 0.54 ⫾ 0.13
4.1 2.6 1.7
3.2–5.3 2.1–3.3 1.3–2.2
⬍0.0001 ⬍0.0001 ⬍0.0001
b c Age is a continuous variable. Odds ratios were calculated by comparing against LVEF ⬎60%. Odds ratios were calculated by comparing against the 2005 to 2009 time cohort. The area under the receiver operating characteristic curve was 0.77. The Hosmer–Lemeshow goodness-of-fit p value was 0.35.
a
CABG ⫽ coronary artery bypass grafting surgery;
CI ⫽ confidence interval;
LVEF ⫽ left ventricular ejection fraction;
SE ⫽ standard error.
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1682 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
Table 5. Trends in Odds Ratios for Predictors of Low Cardiac Output Syndrome Variable LVEF ⬍ 0.20 LVEF 0.20–0.40 Reoperative CABG Emergency/ urgent surgery Female sex Hypertension Age Left main disease Triple vessels disease Diabetes
1991–1994a 1995–1999b 2000–2004c 2005–2009d 3.1 1.7 4.3
4.3 2.2 4.3
3.2 2.2 4.1
7.8 4 5.4
3.5
7.9
9.5
2.5
2.4 1.5 1.01 1.4
2 1.8 1.03 1.3
1.9 1.5 — —
— — — —
1.4
1.5
—
—
1.5
—
—
—
a
C-statistic is 0.75, the Hosmer-Lemeshow goodness-of-fit p value was b 0.1. C-statistic is 0.77, the Hosmer-Lemeshow goodness-of-fit c p value is 0.88. C-statistic is 0.76, the Hosmer-Lemeshow goodnessd of-fit p value is 0.40. C-statistic is 0.72. The Hosmer-Lemeshow goodness-of-fit p value is 0.77. The numbers in the last four columns represents odds ratios. CABG ⫽ coronary artery bypass grafting surgery; tricular ejection fraction.
LVEF ⫽ left ven-
than 0.20 compared with those with LVEF 0.20 or less was 3.7 (CI, 2.7 to 5.2), 5.2 (CI, 3.8 to 7.1), 3.1 (CI, 1.9 to 5.1), and 9.8 (CI, 4.4 to 21.7) in the first, second, third, and fourth eras, respectively (p ⫽ 0.04). This indicates that the odds of developing LCOS in patients with LVEF less than 0.20 increased by a relative 80% to 200% in the latest era (OR, 9.8) compared with the previous three eras. This difference was statistically significant by univariate analysis (p ⫽ 0.04). Furthermore, multivariable analysis (Table 5) confirms this finding, where the OR increased to 7.8 in the most recent era compared with an OR of 3.1, 4.3, and 3.2 in the first three eras, respectively. Figure 2 presents the risk profile of patients with LVEF less than 0.20 across the four eras of this study. The prevalence of elderly Fig 1. Prevalence of low cardiac output syndrome by high-risk factors. Left ventricular ejection fraction (LVEF) less than 0.20, reoperative coronary artery bypass grafting surgery (redo CABG), emergent and urgent surgery (Emerg/urg), elderly patients 75 years and older (Age 75⫹), left main coronary stenosis (Left main), and female gender.
Ann Thorac Surg 2011;92:1678 – 85
patients, preoperative myocardial infarction, left main disease, and congestive heart failure increased markedly in patients with LVEF less than 0.20 over the time span of this study.
Comment In the present study, we examined the trends in prevalence and the changes in predictors of LCOS over time in a large series of patients who underwent isolated CABG surgery at our institution over 2 decades (1990 to 2009). The significance of most predictors of LCOS has declined over time, whereas the impact of severe LV dysfunction (LVEF ⬍ 0.20) on the development of postoperative LCOS has increased significantly in the latest era of this study. The prevalence of LCOS has declined significantly with time; however, its’ associated morbidity and mortality has increased. Maganti and associates [5] found similar trends in patients undergoing isolated mitral valve surgery, where LCOS-associated mortality increased from 23% (23 of 100) in the earliest era of their study (1990 to 1995) to 35% (16 of 46) in the last era (2002 to 2008). In their study, LCOS was associated with a 30-fold increase in mortality [4]. Further, LCOS portended even a higher risk of mortality in patients undergoing isolated aortic valve surgery, a 38-fold increase in mortality [5]. This makes LCOS perhaps the most serious complication after cardiac surgery. Indeed, in a multicenter prospective study performed by the Northern New England Cardiovascular Disease Study Group, 384 deaths in 8,641 consecutive patients undergoing isolated CABG between 1990 and 1995 were analyzed with respect to the mode of death. Persistent low cardiac output was judged to be the primary mode of death in 65% of the patients [6]. We believe that the trend of increasing mortality with time in patients with LCOS is partially driven by the worsening risk profile of these patients with time. Indeed, about 20% of patients with LCOS were 75 years or older in the latest era compared with 11.5% in the first one (p ⫽ 0.008). Similarly, the prevalence of preoperative
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1683
Fig 2. Risk profile of patients with left ventricular ejection fraction less than 0.20 by era of surgery. (CHF ⫽ congestive heart failure; LMD ⫽ left main coronary disease; Preop MI ⫽ preoperative myocardial infarction; Preop shock ⫽ preoperative cardiogenic shock. Numbers over the bars are the p values.)
myocardial infarction (39% vs 28%, p ⫽ 0.02), left main disease (46.2% vs 23.6%, p ⬍ 0.001), and requirement of preoperative IABP (20% vs 9.2, p ⫽ 0.001) were all higher in patients with LCOS during the latest era of this study compared with the first one. Moreover, another potential driver of increasing mortality and morbidity in patients with LCOS in the latest era is the higher proportion of LCOS in patients with severe LV dysfunction, representing a very sick population of patients with multiple comorbidities and dismal LV function.
LV Dysfunction Contrary to our initial assumptions, the prevalence of patients with moderate to severe LV dysfunction (LVEF ⬍ 0.40) decreased steadily with time. We hypothesized that this decline might be related to a shift of patients with poor LV function from isolated CABG to combined CABG and valvular surgery, particularly mitral surgery. Exploratory analyses of our institutional data demonstrated an increase in the prevalence of LVEF less than 0.40 among patients who underwent combined mitral valve surgery (24.4%, 32.5%, 38.5%, and 39.1% for the first, second, third, and forth eras, respectively). This shift of patients with LVEF less than 0.40 from the isolated CABG group to the combined CABG and mitral valve surgery group over the last two decades appears to explain most of the reduced prevalence of moderate to severe LV systolic dysfunction in our isolated CABG patients. However, our data cannot clarify whether there was a simultaneous decrease in the referral of patients with left ventricular dysfunction for CABG in favor of percutaneous coronary intervention. Severe LV dysfunction (LVEF ⬍ 0.20) was the most significant predictor of LCOS in the most recent era of this study. Several studies have shown that over time, the relative influence of left ventricular dysfunction on postoperative adverse outcomes has declined [7–10]. This was true during the first three eras of our study; however, the influence of severe LV dysfunction on development of LCOS has increased significantly in the most recent era of our study compared with the first three eras. There are
several potential explanations for this trend. First, this trend can be explained partly by the worsening risk profile of patients with LVEF less than 0.20 over time. Also, surgeons might have had a lower threshold in the latest era to operate on patients with dismal LV function given advances in circulatory support. Additionally, it is possible that the degree of severity of LV dysfunction was higher in the latest era of this study, leading to the higher prevalence of LCOS in the latest era among patients with LVEF less than 0.20. However, we can only speculate on this last explanation, as our data do not capture the exact LV ejection fraction. The increasing prevalence of LCOS in patients with LVEF less than 0.20 puts these patients at high risk of mortality and morbidity. Therefore, prudent preoperative planning is critical in this high-risk group. Preoperative prophylactic insertion of an IABP is justifiable in this high-risk group and may reduce the risk of morbidity and mortality [11]. Additionally, patients with LVEF less than 0.20 should be screened and evaluated preoperatively for potential requirement of circulatory support or consideration of heart transplantation. If a patient with LCOS fails to respond to IABP and inotropic support, implantation of ventricular assist devices as a bridge either to recovery or to transplantation has been shown to be beneficial, often resulting in successful discharge from the hospital [12–14]. Indeed, early implementation of mechanical circulatory assistance in the setting of postoperative LCOS to provide mechanical unloading of the ventricle and rapid restoration of normal end-organ perfusion is associated with improved survival rates [15–17]. The multivariable analysis showed that the impact of most risk factors on the development of LCOS has declined with time, which we believe is the result of refinements and advancements in perioperative surgical and anesthetic techniques. Importantly, cardioplegia techniques have evolved significantly in our institution over the span of this study. Diluted 4:1 blood-tocrystalloid cardioplegia was employed during the first era. Less diluted 8:1 blood cardioplegia was used in the
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1684 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
second and the third eras. In the latest era, undiluted 66:1 blood cardioplegia (microplegia) was used exclusively, and may have accounted in part for the reduction in the prevalence of LCOS and probably lessened the impact of several risk factors on the development of postoperative LCOS. Indeed, previous studies have shown that microplegia is associated with reduced myocardial edema due to a reduction in the volume of crystalloid delivered to the heart, and that this reduced edema may improve myocardial recovery after cardioplegic arrest [18 –21]. Our study also has some limitations. Although the data were prospectively collected, this study remains a retrospective review of a single-center experience. Additionally, two decades of coronary surgery were included in this study, during which both perioperative and postoperative management for coronary disease have changed significantly. However, we tried to account for the time effect by dividing our cohort into four equivalent eras. In conclusion, the significance of most predictors of LCOS has declined over time. Importantly, the impact of severe LV dysfunction (LVEF ⬍ 0.20) on the development of postoperative LCOS has increased significantly in the latest era of this study. Patients with severe LV dysfunction should be the focus of further research, especially in the areas of preoperative risk stratification, perioperative myocardial protection, and postoperative circulatory support. Terrence M. Yau holds the Angelo & Lorenza DeGasperis Chair in Cardiovascular Surgery Research at the University of Toronto.
Ann Thorac Surg 2011;92:1678 – 85
7. 8.
9.
10.
11.
12.
13. 14.
15.
References 1. Rao V, Ivanov J, Weisel RD, Ikonomidis JS, Christakis GT, David TE. Predictors of low cardiac output syndrome after coronary artery bypass. J Thorac Cardiovasc Surg 1996;112: 38 –51. 2. Hogue CW Jr, Sundt T 3rd, Barzilai B, Schectman KB, Dávila-Román VG. Cardiac and neurologic complications identify risks for mortality for both men and women undergoing coronary artery bypass graft surgery. Anesthesiology 2001;95:1074 – 8. 3. Baillot RG, Joanisse DR, Stevens LM, Doyle DP, Dionne B, Lellouche F. Recent evolution in demographic and clinical characteristics and in-hospital morbidity in patients undergoing coronary surgery. Can J Surg 2009;52:394 – 400. 4. Maganti M, Badiwala M, Sheikh A, et al. Predictors of low cardiac output syndrome after isolated mitral valve surgery. J Thorac Cardiovasc Surg 2010;140:790 – 6. 5. Maganti MD, Rao V, Borger MA, Ivanov J, David TE. Predictors of low cardiac output syndrome after isolated aortic valve surgery. Circulation 2005;112:448 – 452. 6. O’Connor GT, Birkmeyer JD, Dacey LJ, et al. Results of a regional study of modes of death associated with coronary
16.
17.
18. 19. 20.
21.
artery bypass grafting. Northern New England Cardiovascular Disease Study Group. Ann Thorac Surg 1998;66:1323– 8. Topkara VK, Cheema FH, Kesavaramanujam S, et al. Coronary artery bypass grafting in patients with low ejection fraction. Circulation 2005;112:344 –50. Davierwala PM, Maganti M, Yau TM. Decreasing significance of left ventricular dysfunction and reoperative surgery in predicting coronary artery bypass grafting–associated mortality: a twelve-year study. J Thorac Cardiovasc Surg 2003;126:1335– 44. Pasquet A, Lauer MS, Williams MJ, Secknus MA, Lytle B, Marwick TH. Prediction of global left ventricular function after bypass surgery in patients with severe left ventricular dysfunction: Impact of preoperative myocardial function, perfusion and metabolism. Eur Heart J 2000; 21:125–36. Appoo J, Norris C, Merali S, et al. Long-term outcome of isolated coronary artery bypass surgery in patients with severe left ventricular dysfunction. Circulation 2004;110: 13–7. Santarpino G, Onorati F, Rubino AS, et al. Preoperative intraaortic balloon pumping improves outcomes for highrisk patients in routine coronary artery bypass graft surgery. Ann Thorac Surg 2009;87:488. Leshnower BG, Gleason TG, O’Hara ML, et al. Safety and efficacy of left ventricular assist device support in postmyocardial infarction cardiogenic shock. Ann Thorac Surg 2006; 81:1365–71. Goldstein DJ, Oz MC. Mechanical support for postcardiotomy cardiogenic shock. Semin Thorac Cardiovasc Surg 2000;12:220 – 8. Pae WE Jr, Miller CA, Matthews Y, Pierce WS. Ventricular assist devices for postcardiotomy cardiogenic shock. A combined registry experience. J Thorac Cardiovasc Surg 1992; 104:541–52. Mehta SM, Aufiero TX, Pae WE Jr, Miller CA, Pierce WS. Results of mechanical ventricular assistance for the treatment of post cardiotomy cardiogenic shock. ASAIO J 1996; 42:211– 8. DeRose JJ Jr, Umana JP, Argenziano M, et al. Improved results for postcardiotomy cardiogenic shock with the use of implantable left ventricular assist devices. Ann Thorac Surg 1997;64:1757– 62. Helman DN, Morales DL, Edwards NM, et al. Left ventricular assist device bridge-to-transplant network improves survival after failed cardiotomy. Ann Thorac Surg 1999;68: 1187–94. Menasché P. Blood cardioplegia: do we still need to dilute? Ann Thorac Surg 1996;62:957– 60. Machiraju VR, Lima CA, Culig MH. The value of minicardioplegia in the clinical setting. Ann Thorac Surg 1997;64: 887–94. McCann UG 2nd, Lutz CJ, Picone AL, et al. Whole blood cardioplegia (minicardioplegia) reduces myocardial edema after ischemic injury and cardiopulmonary bypass. J Extra Corpor Technol 2006;38:14 –21. 15. Velez DA, Morris CD, Budde JM, et al. All-blood (Miniplegia) versus dilute cardioplegia in experimental surgical revascularization of evolving iInfarction. Circulation 2001;104:I-296.
INVITED COMMENTARY Algarni and colleagues [1] present a 20-year analysis of a large cohort of patients undergoing coronary artery bypass grafting, in which they document a decrease in the prevalence of low cardiac output syndrome, improved in-hospital mortality, and a reduction in major postoper© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc
ative complications over this 20-year span. Two unexpected findings in the current era were an increase in mortality among patients suffering low cardiac output syndrome and an increase in the likelihood of low cardiac output syndrome among patients with an ejection frac0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.08.073
Predictors of Low Cardiac Output Syndrome After Isolated Coronary Artery Bypass Surgery: Trends Over 20 Years Khaled D. Algarni, MD, MHS, Manjula Maganti, MS, and Terrence M. Yau, MD, MS Division of Cardiovascular Surgery, Peter Munk Cardiac Center, Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada
Background. Postoperative low cardiac output syndrome (LCOS) is associated with high morbidity and mortality after coronary artery bypass grafting (CABG). We sought to examine trends in predictors of LCOS after isolated CABG. Methods. A total of 25,176 consecutive patients who underwent isolated CABG between 1990 and 2009 were included. To examine trends over time, we divided patients into four equivalent eras (1990 to -1994, n ⴝ 6,489; 1995 to 1999, n ⴝ 8,175; 2000 to 2004, n ⴝ 6,741; 2005 to 2009, n ⴝ 3,797). We used multivariable analysis to identify predictors of LCOS. Results. The prevalence of LCOS declined from 9.1% (1990 to 1994) to 2.4% (2005 to 2009, p < 0.001). The following were the major independent predictors of LCOS for the entire cohort (odds ratios in parentheses): reoperative CABG (4.1); earlier year of operation (4.1, 2.6, 1.7 for the first, second, and third eras, respectively); left
ventricular ejection fraction (LVEF) less than 0.20 (3.5), emergency surgery (2.7), cardiogenic shock (2.3), female gender (2), and LVEF 0.20 to 0.39 (2). Unlike other risk factors, the impact of LVEF less than 0.20 on development of postoperative LCOS increased substantially in the latest era (odds ratio, 7.8) compared with (odds ratios, 3.1, 4.3, and 3.2) the first, second, and third eras, respectively. Conclusions. The impact of LVEF less than 0.20 on development of postoperative LCOS has increased markedly in the latest era of our study. Prudent preoperative evaluation in patients with severe left ventricular dysfunction is critical. Further innovative research in myocardial protection and circulatory support is warranted in patients with severe left ventricular dysfunction.
T
Patients and Methods
he development of low cardiac output syndrome (LCOS) after coronary artery bypass grafting (CABG) surgery is associated with a 10-fold to 17-fold increase in mortality and markedly increased morbidity [1].The reported prevalence of LCOS after isolated CABG is 3% to 14% [1–3]. The prevalence and the independent predictors of LCOS were described previously by Rao and associates [1]. However, no studies have examined temporal changes in predictors of LCOS in patients undergoing isolated CABG surgery. The purpose of this study was to examine trends over time in the prevalence of LCOS in patients undergoing CABG surgery over a 20-year period, as well as to identify changes in the independent predictors of LCOS with time. These findings may assist in identification of patients at high risk of postoperative LCOS and therefore facilitate preoperative planning and patient counselling, and guide potential preoperative interventions for optimization of high-risk subgroups.
Accepted for publication June 8, 2011. Address correspondence to Dr Yau, Division of Cardiovascular Surgery, Toronto General Hospital, 4N-470, 200 Elizabeth St, Toronto, Ontario, Canada, M5G 2C4; e-mail: [email protected].
© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2011;92:1678 – 85) © 2011 by The Society of Thoracic Surgeons
Data Collection and Study Outcomes Clinical, operative, and outcome data were collected prospectively in a computerized database for 25,176 consecutive patients undergoing isolated CABG at our institution between January 1, 1990 and December 31, 2009. Patients undergoing CABG with concomitant procedures were excluded. To evaluate changing patterns of risk factors and outcomes over time, the study population (n ⫽ 25,202) was divided into 4 eras (1990 to 1994, n ⫽ 6,489; 1995 to 1999, n ⫽ 8,175; 2000 to 2004, n ⫽ 6,741, and 2005 to 2009, n ⫽ 3,797). The study was approved by our Institutional Research Ethics board; individual patient consent was waived. Our primary outcome was LCOS, which was diagnosed if the patient required an intraaortic balloon pump (IABP) in the operating room to be weaned from cardiopulmonary bypass or in the intensive care unit because of hemodynamic compromise. The LCOS was also diagnosed if the patient required inotropic medication (either dopamine, dobutamine, milrinone, or epinephrine) to maintain a systolic blood pressure greater than 90 mm Hg and cardiac output greater than 2.2 L · min⫺1 · m⫺2 for at least 30 minutes in the intensive care unit, after optimizing preload, and afterload as well as correcting all 0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.06.017
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1679
Table 1. Demographics and Clinical Profile Variable
1990–1994
1995–1999
2000–2004
2005–2009
No. of patients Age (years) mean ⫾ SD Less than 65 65–74 ⱖ75 Female Diabetes mellitus Hypertension Hyperlipidemia Angina Stable Crescendo ACS Preoperative MI Cardiogenic shock Congestive heart failure Preoperative IABP LVEF ⬎0.60 0.40–0.60 0.20–0.40 ⬍0.20 Left main disease Triple-vessel disease Preoperative stroke PVD Reoperative CABG Previous PCI ⫾ stents Urgency of surgery Elective Semiurgent Urgent Emergency
6,489 61.6 ⫾ 9.7 58.7 33.8 7.5 20 23.7 50.2 52.5
8,175 62.9 ⫾ 9.9 52.3 35.9 11.8 22.2 27.8 53.7 65.9
6,741 64.4 ⫾ 9.9 47.6 36.5 15.9 20.4 33.3 64.5 83
3,797 65.8 ⫾ 10 43.6 34.9 21.5 21.9 39.1 76.4 90.2
p Value ⬍0.0001 ⬍0.0001
0.003 ⬍0.0001 ⬍0.0001 ⬍0.0001
37.1 33.6 27.3 16.3 1.2 8.3 3.5
32.2 30 35.8 20.3 1.2 8.9 3.5
39 33.8 23.3 22 1.1 10.1 3.1
35.1 29.5 25.5 23.7 1.5 10.1 3.1
⬍0.0001
32.2 43.6 21.2 3.1 18.2 72.7 8.3 14 7.2 10.3
31 45.7 20.5 2.9 18.3 74.7 8.1 14.7 5 9.2
40.8 38 18.8 2.5 26.3 80.3 8.5 18.7 3.5 9.1
53.6 31 14.3 1.2 34.9 82 9.5 20.4 2.1 11
⬍0.0001
57.3 20.6 15.7 2
48.1 36.2 11.1 0.7
51.7 36.6 8.5 0.6
49.5 36.4 9.8 0.6
⬍0.0001
⬍0.0001 0.2 0.006 0.4
⬍0.0001 ⬍0.0001 0.1 ⬍0.0001 ⬍0.0001 0.002
ACS ⫽ acute coronary syndrome; CABG ⫽ coronary artery bypass grafting; IABP ⫽ intraaortic balloon pump; LVEF ⫽ left ventricular ejection fraction; MI ⫽ myocardial infarction; PCI ⫽ percutaneous coronary intervention; PVD ⫽ peripheral vascular disease.
electrolyte and blood gas abnormalities. Patients who required a renal dose of dopamine (⬍4 g/kg) or those who received vasoconstricting medications to increase peripheral vascular resistance in the presence of normal or high cardiac output (ⱖ 2.5 L · min⫺1 · m⫺2) were not considered to have LCOS.
Statistical Analysis Statistical analysis was performed with SAS 9.2 software (SAS Institute, Inc, Cary, NC). We used the 2 or Fisher exact test to analyze categoric data. Categoric variables were expressed as percentages. Analysis of variance was used to analyze continuous variables that are normally distributed, and the Kruskal-Wallis test was used for variables with a nonparametric distribution. Continuous variables were expressed as means ⫾ standard deviations. Statistical significance was based on two-sided p values less than 0.05. Variables that had a univariate p
value of less than 0.25 and those of biological importance were selected for inclusion in a multivariable logistic regression model. Multivariable analysis was performed for the entire cohort of 25,176 patients as well as for each era separately to determine the independent predictors of LCOS for the entire cohort and for each era separately. Model discrimination and calibration were evaluated by the area under the receiver operating characteristic curve, and the Hosmer–Lemeshow goodness-of-fit statistic, respectively.
Results Demographics and Baseline Clinical Profile Preoperative patient characteristics are detailed in Table 1. The prevalence of most risk factors increased steadily with time. About 20% of patients were 75 years and older
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1680 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
Ann Thorac Surg 2011;92:1678 – 85
Table 2. Operative Data and Postoperative Outcomes Variable Intraoperative variables LITA use RITA use Radial artery usea No. of grafts CPB time (minutes) Cross-clamping time (minutes) Cardioplegia technique Intermittent cold antegrade Warm or tepid cardioplegia Off pump Postoperative outcomes Mortality Inotrope use ⬎30 minutes Low-output syndrome IABP Myocardial infarction Stroke Renal failure Pulmonary complications Sternal wound infection Superficial Deep Sepsis Duration of ventilation (hours) Intensive care unit stay (hours) Hospital stay (days) a
1990–1994
1995–1999
2000–2004
2005–2009
p Value
83 2.1 — 3.5 ⫾ 0.9 86.9 ⫾ 26.8 60.5 ⫾ 18.4
89.6 3.4 2.5 3.6 ⫾ 0.9 85.2 ⫾ 26 63.5 ⫾ 18.4
88.9 8.1 3.7 3.5 ⫾ 0.9 91.8 ⫾ 25.7 68.8 ⫾ 20.8
91.8 7.2 1.9 3.4 ⫾ 0.9 89 ⫾ 23.3 68.1 ⫾ 21.3
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
76 14.6 0.2
67.6 16.6 3
80.9 4.9 6.2
86.8 0.4 9.5
⬍0.0001
2.5 31.2 5.1 4.9 3.2 1.5 0.8 8.9
1.7 30.8 4 4.9 2.5 1.4 0.9 8.5
1.1 39.7 2.6 3.2 2.9 0.9 0.9 7.5
1.4 37.1 1.8 2.2 1.8 0.9 1.3 10.2
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.001 0.01 0.07 ⬍0.0001
2.1 0.9 0.6 26.7 ⫾ 61.2 56.4 ⫾ 79 10 ⫾ 9.4
1.5 0.9 0.7 14.1 ⫾ 52.1 43.3 ⫾ 90.8 8 ⫾ 8.9
1.5 0.6 1.3 16.9 ⫾ 58 51.9 ⫾ 116 8 ⫾ 9.9
1.2 0.6 1.6 16.8 ⫾ 59.3 53.6 ⫾ 100 7.9 ⫾ 7.6
0.006 0.001 ⬍0.0001 ⬍0.0001 ⬍0.0001
Radial artery data was not collected during the first time cohort (1991–1996).
CPB ⫽ cardiopulmonary bypass;
IABP ⫽ intraaortic balloon pump;
in the latest time cohort versus 7.5% of patients in the first one. The prevalence of left main disease doubled between the first and the last time cohort (18.2% vs 34.9%). In contrast, the prevalence of left ventricular dysfunction and reoperative CABG has declined significantly with time.
Operative and Hospital Outcomes Operative profiles and postoperative outcomes are detailed in Table 2. The overall prevalence of LCOS for the entire cohort was 5.7% (n ⫽ 1,431). The prevalence of LCOS declined from 9.1% (n ⫽ 588) in the first time cohort to 2.4% (n ⫽ 91) in the last one (p ⬍ 0.0001). In-hospital mortality declined from 2.5% in the first era to 1.4% in the last one (p ⬍ 0.0001). The prevalence of most postoperative complications declined significantly with time. However, there was a trend of increasing prevalence of sepsis with time. The duration of ventilation and intensive care unit stay has fluctuated with time. Nonetheless, the length of hospital stay decreased significantly with time.
Postoperative Outcomes Stratified by LCOS Table 3 details the postoperative outcomes in patients who developed LCOS compared with those who did not.
LITA ⫽ left internal thoracic artery;
RITA ⫽ right internal thoracic artery.
Across the four eras, operative mortality was higher in patients with LCOS (17.7% vs 0.9%, 18.5% vs 0.6%, 12.8 vs 0.6%, and 24% vs 0.8%), respectively (p ⬍ 0.0001). Similarly, patients with LCOS had significantly longer ventilatory support, intensive care unit stay, hospital stay, and more blood transfusion. Moreover, the risk of mortality in patients with LCOS increased significantly in the latest era, where 22 out of 91 patients with LCOS died in the latest era (24%) compared with mortality of 17.7%, 18.5%, and 12.8% for the LCOS group in the first, second, and third eras, respectively. Thus LCOS was associated with a 29-fold increase in mortality in the latest era, compared with a 17-fold increase in the first era of this study.
Predictors of LCOS Predictors of LCOS are detailed in Table 4. Reoperative CABG (odds ratio [OR], 4.1) and the first era of surgery (OR, 4.1) carried the highest odds of LCOS, followed by LVEF less than 0.20 (OR, 3.5) and then emergency CABG (OR, 2.7). Table 5 presents the odds ratios for independent predictors of LCOS across the four eras of the study. The predictors of LCOS have changed significantly with time. In the latest era, only four variables emerged as significant predictors of LCOS; LVEF less than 0.20 (OR, 7.8),
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1681
Table 3. Outcomes Stratified by Low Cardiac Output Syndrome Statusa Variable Mortality LCOS % (n) No LCOS % (n) Intensive care unit stay (hours) LCOS No LCOS Duration of ventilation (hours) LCOS No LCOS Postoperative hospital stay (days) LCOS No LCOS a
1990–1994
1995–1999
2000–2004
2005–2009
17.7 (588) 95 (5,892)
18.5 (487) 61 (7,677)
12.8 (265) 56 (6,471)
24.2 (91) 84 (3,705)
111.3 ⫾ 134 50.8 ⫾ 68.5
141.7 ⫾ 224 37.3 ⫾ 70.9
205 ⫾ 440 45.6 ⫾ 73.2
200 ⫾ 251 49.9 ⫾ 90
66.8 ⫾ 111 22.8 ⫾ 52.3
75.3 ⫾ 152.2 10.3 ⫾ 37.7
109 ⫾ 165.4 13.2 ⫾ 45.1
124.6 ⫾ 194 14 ⫾ 47.5
13.4 ⫾ 12.7 9.6 ⫾ 8.9
15.1 ⫾ 20.8 7.5 ⫾ 7.1
18.2 ⫾ 37.7 7.5 ⫾ 6.2
16.1 ⫾ 21.8 7.6 ⫾ 6.7
p ⬍ 0.001 for each individual grouping variable and across the four eras.
LCOS ⫽ low cardiac output syndrome.
reoperative CABG (OR, 5.4), LVEF 0.20 to 0.39 (OR, 4), and emergent-urgent surgery (OR, 2.5). Figure 1 shows that the observed frequency of LCOS declined with time in all high-risk categories except patients with LVEF less than 0.20. In these patients with severe LV dysfunction, the prevalence of LCOS increased from 10.7% in the third era of this study to 18.2% in the latest era.
LVEF Less Than 0.20 Unlike other predictive variables, the prevalence of LCOS increased significantly in patients with LVEF less than 0.20 in the most recent era (18.2%) compared with the third era of this study (10.7%). We used a Cochrane Mantel-Haenszel test to compare the odds of developing LCOS in patients with LVEF less than 0.20 across the four eras. The odds ratio for LCOS in patients with LVEF less
Table 4. Multivariable Predictors of Low Cardiac Output Syndrome (n ⫽ 25,176) Variable Constant Agea Female gender Diabetes Cardiogenic shock Congestive heart failure Peripheral vascular disease Reoperative CABG Left main disease LVEFb ⬍ 0.20 0.20–0.40 0.40–0.60 Urgency of surgery Same hospitalization Urgent Emergent Year of operation 1991–1996c 1997–2002c 1997–2002c
n
Estimate ⫾ SE
Odds Ratio
95% CI
p Value
— 5,322 7,539 305 2,325 4,136 1,259 5,775
⫺70.26 0.015 ⫾ 0.003 0.70 ⫾ 0.06 0.17 ⫾ 0.06 0.58 ⫾ 0.16 0.63 ⫾ 0.08 0.15 ⫾ 0.07 1.41 ⫾ 0.09 0.26 ⫾ 0.07
1.016 2 1.2 2.3 1.7 1.2 4.1 1.3
1.01–1.02 1.8–2.3 1.1–1.3 1.7–3.2 1.4–2.3 1.01–1.3 3.5–4.8 1.1–1.5
⬍0.0001 ⬍0.0001 0.006 ⬍0.0001 ⬍0.0001 0.04 ⬍0.0001 ⬍0.0001
646 4,847 10,281
1.27 ⫾ 0.14 0.71 ⫾ 0.09 0.23 ⫾ 0.08
3.5 2.0 1.3
2.7–4.6 1.7–2.4 1.1–1.5
⬍0.0001 ⬍0.0001 0.003
751 2,622 247
0.41 ⫾ 0.14 0.53 ⫾ 0.1 1 ⫾ 0.2
1.5 1.7 2.7
1.1–2 1.4–2.1 1.8–4
0.005 ⬍0.0001 ⬍0.0001
6,489 8,175 6,741
1.4 ⫾ 0.12 0.96 ⫾ 0.12 0.54 ⫾ 0.13
4.1 2.6 1.7
3.2–5.3 2.1–3.3 1.3–2.2
⬍0.0001 ⬍0.0001 ⬍0.0001
b c Age is a continuous variable. Odds ratios were calculated by comparing against LVEF ⬎60%. Odds ratios were calculated by comparing against the 2005 to 2009 time cohort. The area under the receiver operating characteristic curve was 0.77. The Hosmer–Lemeshow goodness-of-fit p value was 0.35.
a
CABG ⫽ coronary artery bypass grafting surgery;
CI ⫽ confidence interval;
LVEF ⫽ left ventricular ejection fraction;
SE ⫽ standard error.
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1682 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
Table 5. Trends in Odds Ratios for Predictors of Low Cardiac Output Syndrome Variable LVEF ⬍ 0.20 LVEF 0.20–0.40 Reoperative CABG Emergency/ urgent surgery Female sex Hypertension Age Left main disease Triple vessels disease Diabetes
1991–1994a 1995–1999b 2000–2004c 2005–2009d 3.1 1.7 4.3
4.3 2.2 4.3
3.2 2.2 4.1
7.8 4 5.4
3.5
7.9
9.5
2.5
2.4 1.5 1.01 1.4
2 1.8 1.03 1.3
1.9 1.5 — —
— — — —
1.4
1.5
—
—
1.5
—
—
—
a
C-statistic is 0.75, the Hosmer-Lemeshow goodness-of-fit p value was b 0.1. C-statistic is 0.77, the Hosmer-Lemeshow goodness-of-fit c p value is 0.88. C-statistic is 0.76, the Hosmer-Lemeshow goodnessd of-fit p value is 0.40. C-statistic is 0.72. The Hosmer-Lemeshow goodness-of-fit p value is 0.77. The numbers in the last four columns represents odds ratios. CABG ⫽ coronary artery bypass grafting surgery; tricular ejection fraction.
LVEF ⫽ left ven-
than 0.20 compared with those with LVEF 0.20 or less was 3.7 (CI, 2.7 to 5.2), 5.2 (CI, 3.8 to 7.1), 3.1 (CI, 1.9 to 5.1), and 9.8 (CI, 4.4 to 21.7) in the first, second, third, and fourth eras, respectively (p ⫽ 0.04). This indicates that the odds of developing LCOS in patients with LVEF less than 0.20 increased by a relative 80% to 200% in the latest era (OR, 9.8) compared with the previous three eras. This difference was statistically significant by univariate analysis (p ⫽ 0.04). Furthermore, multivariable analysis (Table 5) confirms this finding, where the OR increased to 7.8 in the most recent era compared with an OR of 3.1, 4.3, and 3.2 in the first three eras, respectively. Figure 2 presents the risk profile of patients with LVEF less than 0.20 across the four eras of this study. The prevalence of elderly Fig 1. Prevalence of low cardiac output syndrome by high-risk factors. Left ventricular ejection fraction (LVEF) less than 0.20, reoperative coronary artery bypass grafting surgery (redo CABG), emergent and urgent surgery (Emerg/urg), elderly patients 75 years and older (Age 75⫹), left main coronary stenosis (Left main), and female gender.
Ann Thorac Surg 2011;92:1678 – 85
patients, preoperative myocardial infarction, left main disease, and congestive heart failure increased markedly in patients with LVEF less than 0.20 over the time span of this study.
Comment In the present study, we examined the trends in prevalence and the changes in predictors of LCOS over time in a large series of patients who underwent isolated CABG surgery at our institution over 2 decades (1990 to 2009). The significance of most predictors of LCOS has declined over time, whereas the impact of severe LV dysfunction (LVEF ⬍ 0.20) on the development of postoperative LCOS has increased significantly in the latest era of this study. The prevalence of LCOS has declined significantly with time; however, its’ associated morbidity and mortality has increased. Maganti and associates [5] found similar trends in patients undergoing isolated mitral valve surgery, where LCOS-associated mortality increased from 23% (23 of 100) in the earliest era of their study (1990 to 1995) to 35% (16 of 46) in the last era (2002 to 2008). In their study, LCOS was associated with a 30-fold increase in mortality [4]. Further, LCOS portended even a higher risk of mortality in patients undergoing isolated aortic valve surgery, a 38-fold increase in mortality [5]. This makes LCOS perhaps the most serious complication after cardiac surgery. Indeed, in a multicenter prospective study performed by the Northern New England Cardiovascular Disease Study Group, 384 deaths in 8,641 consecutive patients undergoing isolated CABG between 1990 and 1995 were analyzed with respect to the mode of death. Persistent low cardiac output was judged to be the primary mode of death in 65% of the patients [6]. We believe that the trend of increasing mortality with time in patients with LCOS is partially driven by the worsening risk profile of these patients with time. Indeed, about 20% of patients with LCOS were 75 years or older in the latest era compared with 11.5% in the first one (p ⫽ 0.008). Similarly, the prevalence of preoperative
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
1683
Fig 2. Risk profile of patients with left ventricular ejection fraction less than 0.20 by era of surgery. (CHF ⫽ congestive heart failure; LMD ⫽ left main coronary disease; Preop MI ⫽ preoperative myocardial infarction; Preop shock ⫽ preoperative cardiogenic shock. Numbers over the bars are the p values.)
myocardial infarction (39% vs 28%, p ⫽ 0.02), left main disease (46.2% vs 23.6%, p ⬍ 0.001), and requirement of preoperative IABP (20% vs 9.2, p ⫽ 0.001) were all higher in patients with LCOS during the latest era of this study compared with the first one. Moreover, another potential driver of increasing mortality and morbidity in patients with LCOS in the latest era is the higher proportion of LCOS in patients with severe LV dysfunction, representing a very sick population of patients with multiple comorbidities and dismal LV function.
LV Dysfunction Contrary to our initial assumptions, the prevalence of patients with moderate to severe LV dysfunction (LVEF ⬍ 0.40) decreased steadily with time. We hypothesized that this decline might be related to a shift of patients with poor LV function from isolated CABG to combined CABG and valvular surgery, particularly mitral surgery. Exploratory analyses of our institutional data demonstrated an increase in the prevalence of LVEF less than 0.40 among patients who underwent combined mitral valve surgery (24.4%, 32.5%, 38.5%, and 39.1% for the first, second, third, and forth eras, respectively). This shift of patients with LVEF less than 0.40 from the isolated CABG group to the combined CABG and mitral valve surgery group over the last two decades appears to explain most of the reduced prevalence of moderate to severe LV systolic dysfunction in our isolated CABG patients. However, our data cannot clarify whether there was a simultaneous decrease in the referral of patients with left ventricular dysfunction for CABG in favor of percutaneous coronary intervention. Severe LV dysfunction (LVEF ⬍ 0.20) was the most significant predictor of LCOS in the most recent era of this study. Several studies have shown that over time, the relative influence of left ventricular dysfunction on postoperative adverse outcomes has declined [7–10]. This was true during the first three eras of our study; however, the influence of severe LV dysfunction on development of LCOS has increased significantly in the most recent era of our study compared with the first three eras. There are
several potential explanations for this trend. First, this trend can be explained partly by the worsening risk profile of patients with LVEF less than 0.20 over time. Also, surgeons might have had a lower threshold in the latest era to operate on patients with dismal LV function given advances in circulatory support. Additionally, it is possible that the degree of severity of LV dysfunction was higher in the latest era of this study, leading to the higher prevalence of LCOS in the latest era among patients with LVEF less than 0.20. However, we can only speculate on this last explanation, as our data do not capture the exact LV ejection fraction. The increasing prevalence of LCOS in patients with LVEF less than 0.20 puts these patients at high risk of mortality and morbidity. Therefore, prudent preoperative planning is critical in this high-risk group. Preoperative prophylactic insertion of an IABP is justifiable in this high-risk group and may reduce the risk of morbidity and mortality [11]. Additionally, patients with LVEF less than 0.20 should be screened and evaluated preoperatively for potential requirement of circulatory support or consideration of heart transplantation. If a patient with LCOS fails to respond to IABP and inotropic support, implantation of ventricular assist devices as a bridge either to recovery or to transplantation has been shown to be beneficial, often resulting in successful discharge from the hospital [12–14]. Indeed, early implementation of mechanical circulatory assistance in the setting of postoperative LCOS to provide mechanical unloading of the ventricle and rapid restoration of normal end-organ perfusion is associated with improved survival rates [15–17]. The multivariable analysis showed that the impact of most risk factors on the development of LCOS has declined with time, which we believe is the result of refinements and advancements in perioperative surgical and anesthetic techniques. Importantly, cardioplegia techniques have evolved significantly in our institution over the span of this study. Diluted 4:1 blood-tocrystalloid cardioplegia was employed during the first era. Less diluted 8:1 blood cardioplegia was used in the
ADULT CARDIAC
Ann Thorac Surg 2011;92:1678 – 85
1684 ADULT CARDIAC
ALGARNI ET AL LOW OUTPUT SYNDROME AFTER CORONARY BYPASS
second and the third eras. In the latest era, undiluted 66:1 blood cardioplegia (microplegia) was used exclusively, and may have accounted in part for the reduction in the prevalence of LCOS and probably lessened the impact of several risk factors on the development of postoperative LCOS. Indeed, previous studies have shown that microplegia is associated with reduced myocardial edema due to a reduction in the volume of crystalloid delivered to the heart, and that this reduced edema may improve myocardial recovery after cardioplegic arrest [18 –21]. Our study also has some limitations. Although the data were prospectively collected, this study remains a retrospective review of a single-center experience. Additionally, two decades of coronary surgery were included in this study, during which both perioperative and postoperative management for coronary disease have changed significantly. However, we tried to account for the time effect by dividing our cohort into four equivalent eras. In conclusion, the significance of most predictors of LCOS has declined over time. Importantly, the impact of severe LV dysfunction (LVEF ⬍ 0.20) on the development of postoperative LCOS has increased significantly in the latest era of this study. Patients with severe LV dysfunction should be the focus of further research, especially in the areas of preoperative risk stratification, perioperative myocardial protection, and postoperative circulatory support. Terrence M. Yau holds the Angelo & Lorenza DeGasperis Chair in Cardiovascular Surgery Research at the University of Toronto.
Ann Thorac Surg 2011;92:1678 – 85
7. 8.
9.
10.
11.
12.
13. 14.
15.
References 1. Rao V, Ivanov J, Weisel RD, Ikonomidis JS, Christakis GT, David TE. Predictors of low cardiac output syndrome after coronary artery bypass. J Thorac Cardiovasc Surg 1996;112: 38 –51. 2. Hogue CW Jr, Sundt T 3rd, Barzilai B, Schectman KB, Dávila-Román VG. Cardiac and neurologic complications identify risks for mortality for both men and women undergoing coronary artery bypass graft surgery. Anesthesiology 2001;95:1074 – 8. 3. Baillot RG, Joanisse DR, Stevens LM, Doyle DP, Dionne B, Lellouche F. Recent evolution in demographic and clinical characteristics and in-hospital morbidity in patients undergoing coronary surgery. Can J Surg 2009;52:394 – 400. 4. Maganti M, Badiwala M, Sheikh A, et al. Predictors of low cardiac output syndrome after isolated mitral valve surgery. J Thorac Cardiovasc Surg 2010;140:790 – 6. 5. Maganti MD, Rao V, Borger MA, Ivanov J, David TE. Predictors of low cardiac output syndrome after isolated aortic valve surgery. Circulation 2005;112:448 – 452. 6. O’Connor GT, Birkmeyer JD, Dacey LJ, et al. Results of a regional study of modes of death associated with coronary
16.
17.
18. 19. 20.
21.
artery bypass grafting. Northern New England Cardiovascular Disease Study Group. Ann Thorac Surg 1998;66:1323– 8. Topkara VK, Cheema FH, Kesavaramanujam S, et al. Coronary artery bypass grafting in patients with low ejection fraction. Circulation 2005;112:344 –50. Davierwala PM, Maganti M, Yau TM. Decreasing significance of left ventricular dysfunction and reoperative surgery in predicting coronary artery bypass grafting–associated mortality: a twelve-year study. J Thorac Cardiovasc Surg 2003;126:1335– 44. Pasquet A, Lauer MS, Williams MJ, Secknus MA, Lytle B, Marwick TH. Prediction of global left ventricular function after bypass surgery in patients with severe left ventricular dysfunction: Impact of preoperative myocardial function, perfusion and metabolism. Eur Heart J 2000; 21:125–36. Appoo J, Norris C, Merali S, et al. Long-term outcome of isolated coronary artery bypass surgery in patients with severe left ventricular dysfunction. Circulation 2004;110: 13–7. Santarpino G, Onorati F, Rubino AS, et al. Preoperative intraaortic balloon pumping improves outcomes for highrisk patients in routine coronary artery bypass graft surgery. Ann Thorac Surg 2009;87:488. Leshnower BG, Gleason TG, O’Hara ML, et al. Safety and efficacy of left ventricular assist device support in postmyocardial infarction cardiogenic shock. Ann Thorac Surg 2006; 81:1365–71. Goldstein DJ, Oz MC. Mechanical support for postcardiotomy cardiogenic shock. Semin Thorac Cardiovasc Surg 2000;12:220 – 8. Pae WE Jr, Miller CA, Matthews Y, Pierce WS. Ventricular assist devices for postcardiotomy cardiogenic shock. A combined registry experience. J Thorac Cardiovasc Surg 1992; 104:541–52. Mehta SM, Aufiero TX, Pae WE Jr, Miller CA, Pierce WS. Results of mechanical ventricular assistance for the treatment of post cardiotomy cardiogenic shock. ASAIO J 1996; 42:211– 8. DeRose JJ Jr, Umana JP, Argenziano M, et al. Improved results for postcardiotomy cardiogenic shock with the use of implantable left ventricular assist devices. Ann Thorac Surg 1997;64:1757– 62. Helman DN, Morales DL, Edwards NM, et al. Left ventricular assist device bridge-to-transplant network improves survival after failed cardiotomy. Ann Thorac Surg 1999;68: 1187–94. Menasché P. Blood cardioplegia: do we still need to dilute? Ann Thorac Surg 1996;62:957– 60. Machiraju VR, Lima CA, Culig MH. The value of minicardioplegia in the clinical setting. Ann Thorac Surg 1997;64: 887–94. McCann UG 2nd, Lutz CJ, Picone AL, et al. Whole blood cardioplegia (minicardioplegia) reduces myocardial edema after ischemic injury and cardiopulmonary bypass. J Extra Corpor Technol 2006;38:14 –21. 15. Velez DA, Morris CD, Budde JM, et al. All-blood (Miniplegia) versus dilute cardioplegia in experimental surgical revascularization of evolving iInfarction. Circulation 2001;104:I-296.
INVITED COMMENTARY Algarni and colleagues [1] present a 20-year analysis of a large cohort of patients undergoing coronary artery bypass grafting, in which they document a decrease in the prevalence of low cardiac output syndrome, improved in-hospital mortality, and a reduction in major postoper© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc
ative complications over this 20-year span. Two unexpected findings in the current era were an increase in mortality among patients suffering low cardiac output syndrome and an increase in the likelihood of low cardiac output syndrome among patients with an ejection frac0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.08.073