Cardiopulmonary Bypass Increases Postoperative
Glycemia and Insulin Consumption After Coronary
Surgery
Piotr Knapik, MD, PhD, Paweł Nadziakiewicz, MD, Ewa Urbanska, MD,
Wojciech Saucha, MD, Miroslawa Herdynska, MD, and Marian Zembala, MD, PhD
Departments of Cardiac Anesthesia and Cardiac Surgery, Silesian Centre for Heart Diseases, Zabrze, Poland
Background. Perioperative hyperglycemia should be
avoided in patients undergoing coronary surgery. The aim
of our study was to find out what the influence of cardio-
pulmonary bypass is on postoperative glycemia and insulin
consumption in patients with and without diabetes melli-
tus undergoing coronary artery surgery and whether a
marked hyperglycemia in the early postoperative period is
among the factors associated with early mortality and
morbidity.
Methods. We retrospectively reviewed all patients who
underwent first-time coronary artery surgery in our insti-
tution during the 11-month period. Among 814 patients,
239 patients (29.4%) had diabetes and 575 patients (70.6%)
were nondiabetic. Blood glucose levels were registered
every 2 hours in all patients during the first 24 postoper-
ative hours. Outcomes were difficult glycemic control
(postoperative blood glucose levels >11.0 mmol/L de-
spite aggressive insulin treatment), hospital mortality,
and morbidity (defined as any postoperative complica-
tion such as stroke, renal failure, wound infection, peri-
operative myocardial infarction, ventilation > 24 hours,
sepsis, and multiorgan failure).
Results. Glycemic control was significantly worse in
patients who underwent coronary artery bypass grafting, in
H yperglycemia is one of the independent factors
worsening the prognosis in patients with both
acute coronary syndromes and those undergoing cor-
onary artery surgery as diabetes increases the risk of
cardiac surgery nearly twofold [1, 2]. A negative influ-
ence of hyperglycemia in the perioperative period has
been confirmed in both diabetic and nondiabetic pa-
tients [2, 3]. Prolonged periods of hyperglycemia cause
higher mortality and morbidity, more frequent infec-
tions, hemodynamic impairment, and increased plate-
let aggregation as well as longer and more expensive
hospital stays [4 – 8]. Maintenance of blood glucose
levels below 150 mg% in the postoperative period was
reported to decrease early perioperative mortality and
morbidity after cardiac operations [3]. There is cur-
Accepted for publication Feb 23, 2009.
Address correspondence to Dr Knapik, Silesian Centre for Heart Dis-
eases, ul. Szpitalna 2, Zabrze, 41-800, Poland; e-mail: kardanest@sum.
edu.pl.
© 2009 by The Society of Thoracic Surgeons
Published by Elsevier Inc
ADULT CARDIAC
comparison with off-pump coronary artery bypass grafting
surgery, particularly in nondiabetic patients. Patients with
difficult glycemic control had more serious postoperative
complications resulting in higher mortality (2.5% versus
0.4%; p ⴝ 0.02). In the multivariate analysis, difficult glyce-
mic control was significantly associated with a female sex
(odds ratio [OR], 2.36), presence of diabetes (OR, 2.22), and
the usage of cardiopulmonary bypass (OR, 1.81). Mortality
was significantly associated with the left ventricular ejec-
tion fraction less than 0.35 (OR, 7.38), difficult glycemic
control (OR, 7.06), and previous stroke (OR, 5.66). Difficult
glycemic control was also significantly associated with
postoperative morbidity (OR, 1.87).
Conclusions. Cardiopulmonary bypass increases postop-
erative glycemia and insulin consumption in both diabetic
and nondiabetic patients. The use of cardiopulmonary
bypass during coronary artery surgery in diabetic women is
associated with a more difficult glycemic control in the early
postoperative period. Difficult glycemic control is signifi-
cantly associated with early mortality and morbidity in
patients undergoing coronary artery surgery.
(Ann Thorac Surg 2009;87:1859 – 65)
© 2009 by The Society of Thoracic Surgeons
rently no doubt that effective postoperative glycemic
control is extremely important to avoid postoperative
complications [3, 9, 10].
Coronary artery surgery may be performed with or
without cardiopulmonary bypass. Coronary artery by-
pass graft surgery with cardiopulmonary bypass (CABG)
and off-pump coronary artery bypass graft surgery (OP-
CAB) have been already compared in several clinical
trials. The outcomes are generally similar; however the
avoidance of cardiopulmonary bypass seems to be ben-
eficial for certain subgroups of patients [11, 12].
The association between the use of cardiopulmonary
bypass and perioperative hyperglycemia is not well de-
fined. We therefore aimed to answer the question, whether
the use of cardiopulmonary bypass has an influence on
postoperative glycemia and insulin consumption in pa-
tients with and without diabetes mellitus after coronary
artery surgery. We also aimed to establish whether the
appearance of marked hyperglycemia in the postoperative
0003-4975/09/$36.00
doi:10.1016/j.athoracsur.2009.02.066
 1860 KNAPIK ET AL
GLYCEMIA AFTER CORONARY SURGERY
period (despite strict control by insulin infusion) is associ-
ated with early mortality and morbidity.
Material and Methods
ADULT CARDIAC
We performed a retrospective review of data of all patients
who underwent first-time coronary artery surgery during
the 11-month period. The study was performed in a tertia-
ry-care university hospital, performing approximately 1,700
cardiac surgical procedures per year. The study was ap-
proved by the institutional ethics committee, and the indi-
vidual patient’s consent was not required.
Patients
All 848 consecutive, first-time coronary artery patients
who underwent first-time coronary revascularization
were extracted from the hospital database. After analysis
of the clinical data, 34 patients were excluded: 4 were in
the critical preoperative condition (cardiogenic shock), 19
patients had incomplete medical records, 10 patients
underwent a separate fast-track program and stayed in
the intensive care unit less than 12 hours, and 1 patient
underwent urgent reoperation and died in the operating
theater on the first postoperative day.
Among the remaining 814 patients, 239 patients (29.4%)
had diabetes and 575 patients (70.6%) were nondiabetic
preoperatively. We further subclassified 239 diabetic pa-
tients on the basis of their glucose control at the time of
surgery into 112 patients taking insulin (46.9%), 93 pa-
tients taking oral medications (38.9%), and 34 patients
using dietary control only (14.2%).
Preoperative risk assessment was performed on the
basis of EuroSCORE classification, and patients were
classified as having low risk (0 to 2 points), moderate risk
(3 to 5 points), or high risk (6 or more points) [13].
Blood Glucose Levels
Blood glucose levels were determined by automated
analyzer (Bayer M865, Germany) and were registered
every 2 hours in all patients during the first 24 postoper-
ative hours. Additional blood samples were taken in
between these measurements if necessary, but these
results were not included in the statistical analysis. In a
given period, 9,076 measurements were performed
(mean, 11.2 measurements; 7 to 12 measurements per
patient).
All blood glucose measurements in the first 24 hours
after admission to the postoperative intensive care unit
were collected from the case notes. Intensive care unit
charts were analyzed, and the overall insulin require-
ment for the first 24 postoperative hours (including
infusion and bolus doses) was calculated for each patient.
From these data, the peak and lowest glucose levels
were identified for each patient. Mean glucose levels
(from all 13 measurements) and individual glycemia
range (per patient) were also separately calculated for
each patient. In the first 24 hours glycemia is managed in
our institution with intravenous insulin infusion only,
according to the protocol proposed by Furnary and
colleagues [3]. Starting from postoperative day 2, glyce-
Ann Thorac Surg
2009;87:1859 – 65
mia is managed with further insulin infusion (if neces-
sary) or either subcutaneous insulin, oral hypoglycemic
medication, or diet.
Data Analysis
Perioperative morbidity and mortality was analyzed. Cu-
mulative incidence of the most common complications
(stroke, renal failure, wound infection, perioperative
myocardial infarction, ventilation ⬎ 24 hours, and sepsis
or multiple-organ dysfunction syndrome) was defined as
morbidity; for the calculation of early mortality all deaths
in the first 30 postoperative days were noted.
Stroke was defined as a new focal or global dysfunction
of cerebral function lasting longer than 24 hours. Renal
failure was defined as renal dysfunction requiring renal
replacement therapy. Wound infection was recognized
on the basis of the US Centers for Disease Control and
Prevention definition [14]. Myocardial infarction was
defined as any new Q wave or disappearance of R wave
on postoperative electrocardiogram or troponin I level of
3.9 ␮g/L or greater within 24 hours of the operation [15].
Sepsis was defined as the clinical signs describing sys-
temic inflammatory response syndrome together with
definitive evidence of infection whereas multiple-organ
dysfunction syndrome was defined as a severe dysfunc-
tion of at least two organ systems lasting for more than 24
hours, as stated by the Consensus Conference of the
American College of Chest Physicians and the Society of
Critical Care Medicine [16].
The decision about which data were going to be ex-
tracted and analyzed was made before the start of the
trial. All data were extracted from the case notes directly
into the computer in a standardized manner. All data of
40 randomly chosen patients (5%) were extracted by a
second, independent investigator with the achievement
of 94% agreement. The accuracy of data extraction re-
garding the usage of definitions for perioperative mor-
bidity was assessed in 16 randomly chosen patients (2%)
by one of the main authors of the study, who at the time
of the assessment, was blinded to the information of
which patients were diabetic or nondiabetic and which
patients were operated on with cardiopulmonary bypass
or off-pump. These data were found to be in 90% agree-
ment with the initial assessment.
Patients who underwent their operation on-pump
(CABG) and off-pump (OPCAB) were compared. Analysis
was performed separately for nondiabetic and diabetic
patients. Comparisons included demographic and clinical
data as well as early postoperative mortality, morbidity, and
glycemic control. Mean blood glucose levels were also
compared between CABG and OPCAB patients for the
subsets of nondiabetic and diabetic patients.
Further, patients were stratified according to whether
they had elevated blood glucose levels in the early
postoperative period with at least one blood glucose in
excess of 11 mmol/L during the first postoperative day.
This level was chosen because it has been previously
confirmed that any blood glucose concentration above
this level correlates with an increased mortality risk in
diabetic patients after acute coronary artery syndromes
Ann Thorac Surg KNAPIK ET AL 1861
2009;87:1859 – 65 GLYCEMIA AFTER CORONARY SURGERY

Table 1. Glycemia Control in Patients Undergoing Coronary Artery Bypass Grafting and Off-Pump Coronary Artery Bypass
Grafting Surgery
CABG OPCAB
Variable (n ⫽ 404) (n ⫽ 171) p Value

ADULT CARDIAC
Patients without diabetes
Initial ICU glucose (mmol/L) 7.5 ⫾ 2.1 7.0 ⫾ 1.4 0.02
Mean glucose (mmol/L) 8.4 ⫾ 1.0 8.0 ⫾ 0.9 ⬍0.01
Peak ICU glucose (mmol/L) 10.6 ⫾ 1.9 9.8 ⫾ 1.6 ⬍0.01
Lowest ICU glucose (mmol/L) 6.1 ⫾ 1.3 6.1 ⫾ 1.1 0.64
Glucose range (mmol/L) 4.5 ⫾ 2.3 3.7 ⫾ 1.7 ⬍0.01
Peak ICU glucose ⬎ 11 mmol/L 126 (31.2%) 36 (21.1%) 0.01
Patients requiring insulin infusion 195 (48.3%) 62 (36.3%) ⬍0.01
Insulin/24 h, if used (U) 24.1 (2–251) 15.7 (4–70) 0.02

CABG OPCAB
Variable (n ⫽ 159) (n ⫽ 80) p Value

Diabetic patients
Initial ICU glucose (mmol/L) 7.7 ⫾ 2.4 7.2 ⫾ 2.2 0.16
Mean glucose (mmol/L) 8.4 ⫾ 1.2 8.2 ⫾ 1.1 0.34
Peak ICU glucose (mmol/L) 11.7 ⫾ 2.6 11.0 ⫾ 2.4 0.04
Lowest ICU glucose (mmol/L) 5.4 ⫾ 1.3 5.4 ⫾ 1.2 0.69
Glucose range (mmol/L) 6.3 ⫾ 2.8 5.6 ⫾ 2.6 0.04
Peak ICU glucose ⬎ 11 mmol/L 85 (53.5%) 31 (38.8%) 0.03
Patients requiring insulin infusion 157 (98.7%) 74 (92.5%) 0.03
Insulin/24 h, if used (U) 54.3 (4–150) 49.9 (4–145) 0.29

CABG ⫽ coronary artery bypass grafting; ICU ⫽ intensive care unit; OPCAB ⫽ off-pump coronary artery bypass grafting.

[17]. Patients with elevated blood glucose levels were
compared with patients who had all blood glucose levels
less than 11 mmol/L with regard to their demographic
data, mode of operation, and diabetes status, as well as
early postoperative mortality and morbidity.
Finally, a multiple logistic regression model was used
to investigate the relationship of the outcomes (difficult
glycemic control, mortality, and morbidity) with the pre-
operative and clinical variables. Independent variables
that might influence outcomes were cardiopulmonary
bypass usage, diabetes mellitus, insulin use, age older
than 65 years, left ventricular ejection fraction less than
0.35, EuroSCORE greater than 5, Canadian Cardiovascu-
lar Society score of 4, female sex, previous myocardial
infarction, previous stroke, peripheral vascular disease,
previous percutaneous transluminal coronary angio-
plasty or stent, arterial hypertension, obesity (body mass
index ⬎ 30) and—for morbidity and mortality—also dif-
ficult postoperative glycemic control (at least one blood
glucose ⬎ 11 mmol/L during the first postoperative day).
Statistical Analysis
Power analysis was made on the basis of the first 20
studied patients (10 patients in each group). It was
calculated that it should be at least 20 patients studied in
each group to achieve a power greater than 0.8 and a
significance level of 0.05, and to detect 20% difference
between groups in terms of the presence of any compli-
cation (morbidity) and death in the early postoperative
period.
Depending on the statistical distribution, numeric data
are shown as either mean and standard deviation or
median values and their range, and then compared with
the Mann-Whitney U test. Binary data were shown as the
number and a percentage and compared with the use of
the ␹2 test. Independent variables that might influence
dependent variables (morbidity and mortality) were
identified. The effect of independent variables on the
outcome variables of interest (difficult glycemic control,
morbidity, and mortality) was then calculated by means
of univariate logistic regression, and variables with a
probability value less than 0.1 were then included in the
multivariate logistic regression analysis. For the final
analysis a probability value less than 0.05 was considered
significant.
Results
Patients who underwent their operation on-pump
(CABG) and off-pump (OPCAB) had similar demo-
graphic and clinical data, with only a few exceptions.
Diabetic patients undergoing CABG had a higher mean
left ventricular ejection fraction (0.510 ⫾ 0.104 versus
0.486 ⫾ 0.104; p ⫽ 0.03). Among nondiabetic patients,
fewer patients with a high preoperative risk (EuroSCORE
⬎ 5) were found in the CABG group (23.3% versus 38.8%;
p ⫽ 0.01).
In a comparison of CABG and OPCAB patients, glyce-
mic control (expressed by mean postoperative glycemia,
insulin consumption, glucose range, peak glucose levels,
 1862 KNAPIK ET AL Ann Thorac Surg
GLYCEMIA AFTER CORONARY SURGERY 2009;87:1859 – 65

Table 2. Preoperative Clinical Data in Patients With Elevated Blood Glucose Level (⬎11 mmol/L) and the Remaining Patients
At Least One BGL ⬎ 11 mmol/L All BGL ⬍ 11 mmol/L
Variable (n ⫽ 278) (n ⫽ 536) p Value

Age (y) 63.5 ⫾ 8.3 61.1 ⫾ 9.2 ⬍0.01

ADULT CARDIAC

Body mass index (kg/m2) 28.4 ⫾ 4.3 27.8 ⫾ 3.8 0.25

Obesity (BMI ⬎ 30 kg/m2) 89 (32.0%) 163 (30.4%) 0.64
Ejection fraction 0.496 ⫾ 0.108 0.499 ⫾ 0.103 0.52
EuroSCORE
Low risk (0–2) 75 (27%) 204 (38.1%) ⬍0.01
Moderate risk (3–5) 139 (50%) 236 (44.3%) 0.11
High risk (⬎6) 64 (23%) 96 (17.9%) 0.08
Canadian Cardiovascular Society score 2.7 ⫾ 0.7 2.7 ⫾ 0.7 0.96
Female sex 97 (34.9%) 89 (16.6%) ⬍0.01
Previous myocardial infarction 151 (54.3%) 338 (63.1%) 0.02
Previous stroke 15 (5.4%) 19 (3.5%) 0.21
Peripheral vascular disease 58 (20.9%) 100 (18.7%) 0.45
Previous PTCA or stent 76 (27.3%) 163 (30.4%) 0.36
Arterial hypertension 221 (79.5%) 402 (75.0%) 0.15
No diabetes 162 (58.2%) 413 (80.5%) ⬍0.01
Diabetes treatment
Diet 14 (12.1%) 20 (16.3%) 0.35
Oral 35 (30.2%) 58 (47.2%) ⬍0.01
Insulin 67 (57.8%) 45 (36.6%) ⬍0.01

BGL ⫽ blood glucose level; BMI ⫽ body mass index; PTCA ⫽ percutaneous transluminal coronary angioplasty.

and amount of patients who reached very high glucose
levels) was significantly worse in patients who under-
went CABG, particularly in a group of patients without
diabetes (Table 1).
Comparison of demographic and clinical data of pa-
tients with elevated (at least one result ⬎ 11 mmol/L) and
not elevated blood glucose levels is shown in Table 2.
Patients with difficult glycemic control (elevated glucose
levels) were more likely to have diabetes (41.7% versus
22.9%; p ⬍ 0.01), be of the female sex (34.9% versus 16.6%;
p ⬍ 0.01), and undergo CABG (75.9% versus 65.7%; p ⬍
0.01). In addition, patients with elevated glucose levels
clearly had worse postoperative glucose control (Table 3)
and more serious postoperative complications such as
perioperative myocardial infarction, intraaortic balloon
pump use, or prolonged mechanical ventilation, resulting
in higher mortality (Table 4). Multivariate analysis re-
vealed that difficult glycemic control after coronary sur-
gery was significantly associated with the female sex
(odds ratio [OR], 2.36), the presence of diabetes (OR,
2.22), and the use of cardiopulmonary bypass (OR, 1.81;
Table 5).
Identification of independent variables influencing
other outcomes of interest was also started with the
univariate analysis. For the mortality, univariate analysis
selected elevated blood glucose levels, insulin use, left
ventricular ejection fraction less than 0.35, EuroSCORE
greater than 5, and previous stroke as suitable variables
for the multivariate analysis. For the morbidity, univari-
ate analysis selected elevated blood glucose levels, insu-
lin use, age older than 65 years, EuroSCORE greater than
5, female sex, peripheral vascular disease, and arterial
hypertension as suitable variables for the multivariate
analysis. Elevated blood glucose levels, however, were
strongly correlated with the insulin use, and the Euro-
SCORE variable was clearly dependent on age, sex, and

Table 3. Glycemia Control in Patients With Elevated (⬎11 mmol/L) and Not Elevated Blood Glucose Levels
At Least One BGL ⬎ 11 mmol/L All BGL ⬍ 11 mmol/L
Variable (n ⫽ 278) (n ⫽ 536) p Value

Initial ICU glucose (mmol/L) 8.2 ⫾ 2.5 7.0 ⫾ 1.6 ⬍0.01

Mean glucose (mmol/L) 9.1 ⫾ 1.0 7.9 ⫾ 0.8 ⬍0.01
Peak ICU glucose (mmol/L) 12.9 ⫾ 1.9 9.5 ⫾ 1.0 ⬍0.01
Lowest ICU glucose (mmol/L) 5.9 ⫾ 1.5 5.9 ⫾ 1.2 0.90
Glucose range (mmol/L) 7.1 ⫾ 2.5 3.6 ⫾ 1.5 ⬍0.01
Patients requiring insulin infusion 246 (88.5%) 242 (45.2%) ⬍0.01
Insulin/24 h, if used (U) 45.3 (4–251) 27.9 (2–110) ⬍0.01

BGL ⫽ blood glucose level; ICU ⫽ intensive care unit.

Ann Thorac Surg KNAPIK ET AL 1863
2009;87:1859 – 65 GLYCEMIA AFTER CORONARY SURGERY

Table 4. Operative Data and Complications in Patients With Elevated (⬎11 mmol/L) and Not Elevated Blood Glucose Levels
At Least One BGL ⬎ 11 mmol/L All BGL ⬍ 11 mmol/L
Variable (n ⫽ 278) (n ⫽ 536) p Value

Procedure characteristics

ADULT CARDIAC
Cardiopulmonary bypass (min) 85.6 ⫾ 34.3 82.0 ⫾ 33.6 0.15
Aortic cross-clamp time (min) 49.8 ⫾ 18.1 47.6 ⫾ 19.9 0.05
Postoperative events
Reoperation for bleeding 17 (6.1%) 10 (1.9%) ⬍0.01
Epinephrine ⬎ 0.1 ␮g · kg⫺1 · min⫺1 28 (10.1%) 13 (2.4%) ⬍0.01
IABP 33 (11.9%) 29 (5.4%) ⬍0.01
Isolated complications
Stroke 4 (1.4%) 14 (2.6%) 0.41
Renal failure 5 (1.8%) 6 (1.1%) 0.63
Wound infection 5 (1.8%) 4 (0.8%) 0.31
Perioperative MI 22 (7.9%) 9 (1.7%) ⬍0.01
Ventilation ⬎ 24 h 26 (9.4%) 30 (5.6%) 0.04
Sepsis 6 (2.2%) 9 (1.7%) 0.63
MOF 4 (1.4%) 6 (1.1%) 0.95
Any complications 46 (16.6%) 44 (8.2%) ⬍0.01
Death 7 (2.5%) 2 (0.4%) 0.02
Mean ICU stay (days) 2.1 (1–18) 1.6 (1–39) ⬍0.01
ICU stay ⬎ 5 days 23 (8.3%) 16 (3.0%) ⬍0.01
Mean hospital stay (days) 7.7 (1–57) 6.6 (4–40) ⬍0.01
Hospital stay ⬎ 10 days 36 (13.0%) 26 (4.9%) ⬍0.01

BGL ⫽ blood glucose level; IABP ⫽ intraaortic balloon pump; ICU ⫽ intensive care unit; MI ⫽ myocardial infarction; MOF ⫽ multiorgan
failure.

peripheral vascular disease, so insulin use and Euro-
SCORE were not included in the multivariate analysis.
Results of multivariate analysis for mortality and mor-
bidity are also shown in Table 5. Early mortality was
significantly associated with the left ventricular ejection
fraction less than 0.35 (OR, 7.38), difficult glycemic control
(OR, 7.06), and previous stroke (OR, 5.66). Postoperative
morbidity was significantly associated with the presence
of peripheral vascular disease (OR, 1.89), difficult glyce-
mic control (OR, 1.87), age older than 65 years (OR, 1.84),
and female sex (OR, 1.80).
Comment
glucose homeostasis is already disturbed in many non-
diabetic patients undergoing coronary artery bypass
graft surgery [18].
Vermes and associates [21] indicated that the situation
may be totally opposite and OPCAB surgery may be
associated with an increased early postoperative morbid-
ity in patients with diabetes. The results of this paper
could have been influenced by the fact that more high-
Table 5. Predictors of Difficult Glycemia Control, Morbidity,
and Mortality After Coronary Surgery
Variable OR 95% CL p Value

Our results confirm previous suggestions that cardiopul- Difficult glycemic control
monary bypass has a negative influence on glucose Female sex 2.36 1.65–3.36 ⬍0.001
homeostasis [18]. The use of hypothermia may exacer- Presence of diabetes 2.22 1.60–3.07 ⬍0.001
bate this effect [19]. Our results are also in agreement Cardiopulmonary bypass 1.81 1.29–2.56 ⬍0.001
with the recent study (with a sample size similar to ours) Mortality
confirming that avoidance of cardiopulmonary bypass Left ventricular ejection 7.38 1.86–29.22 0.004
significantly reduced the number of postoperative com- fraction ⬍ 0.35
plications in diabetic patients [20]. According to our data, Elevated blood glucose levels 7.06 1.42–35.04 0.017
various factors indicating overall quality of glycemic Previous stroke 5.66 1.04–30.93 0.045
control (such as glucose range, peak glucose levels, or the Morbidity
percentage of patients who reached very high glucose Peripheral vascular disease 1.89 1.15–3.12 0.012
levels) were worse if cardiopulmonary bypass was used. Elevated blood glucose levels 1.87 1.18–2.96 0.007
This difference was even more obvious in nondiabetic Age ⬎ 65 years 1.84 1.17–2.89 0.008
patients—this could probably be caused by the perma- Female sex 1.80 1.11–2.91 0.018
nently increased glucose levels in the diabetic group.
Apart from that, it has been previously confirmed that 95% CL ⫽ 95% confidence limits; OR ⫽ odds ratio.
 1864 KNAPIK ET AL
GLYCEMIA AFTER CORONARY SURGERY
risk patients were scheduled for OPCAB surgery. In our
study, the situation was similar— diabetic patients with
high preoperative EuroSCORE were scheduled more
often to the OPCAB technique (39% versus 23%; p ⫽
0.01), but patients operated on without the use of cardio-
ADULT CARDIAC
pulmonary bypass still had a similar, but slightly lower,
number of postoperative complications (12.1% versus
8.8%; p ⫽ 0.21).
Despite all these facts, the mean blood glucose level
curves were usually comparable in our CABG and OP-
CAB patients. This could be partially explained by the
fact that comparison of the whole curves (instead of
individual time points) usually “blunts” the statistical
difference. Even if the curves were similar, mean blood
glucose levels were consistently and repeatedly lower in
the OPCAB group.
In our study, patients with elevated postoperative
levels of blood glucose (defined as at least one result ⬎ 11
mmol/L) had strikingly more postoperative complica-
tions (particularly those associated with low cardiac out-
put: adrenaline use, intraaortic balloon pump use, or
perioperative myocardial infarction) and much higher
mortality (2.5% versus 0.4%; p ⫽ 0.02). The negative effect
of high glucose levels on postoperative complications
was previously highlighted in many studies, including
the study published by Van den Berghe and associates
[10], publications by the Furnary group [3, 4], and many
others [1, 9, 22, 23].
There are, however, some important limitations to our
study. On the basis of our data it may not be concluded
how hyperglycemia affects the prognosis of patients with
specific complications, as the majority of these complica-
tions were associated with low cardiac output. Retrospec-
tive data extraction is another limitation to this study;
however, in a set proportion of patients this process was
repeated to confirm the acceptable quality of the hospital
database. The arbitrary decision to use a cutoff point of 11
mmol/L was made because this value is easy to remem-
ber for the staff (“by all means, a blood glucose of 200
mg% should not be exceeded!”) and—as previously ex-
plained—this level significantly increases the risk of
death in diabetic patients with acute coronary syndromes
[17].
No propensity score analysis was performed in our
study. This is important, as for various comparisons we
performed the groups of patients compared were usually
not similar with respect to potential risk factors, but most
important comparisons were performed separately for
diabetic and nondiabetic patients.
Tight glucose control is limited clinically by the fact
that postoperative hypoglycemia may be easily induced
[24]. This is extremely dangerous in the first 24 postop-
erative hours, when each cardiac surgical patient is not
fully alert for at least part of the time. Therefore, we
always observe some patients with elevated blood glu-
cose levels in the postoperative period.
Multivariate analysis of all significant preoperative and
operative variables of both diabetic and nondiabetic
patients was able to show that such factors as danger-
ously high blood glucose levels are more likely to happen
Ann Thorac Surg
2009;87:1859 – 65
in patients with diabetes or of the female sex, and after
the operation with the use of cardiopulmonary bypass.
Moreover, dangerously high blood glucose levels are
among the few factors significantly associated with early
mortality and morbidity in patients undergoing coronary
artery surgery. This finding is supported by many previ-
ous papers, as diabetes significantly increases the risk of
cardiac surgery [25, 26] and increased glucose levels are
harmful if they appear in a preoperative [27], intraoper-
ative [28], or postoperative period [5]. According to one
study, the worst option is to perform coronary artery
surgery in patients with undiagnosed diabetes, as these
patients more frequently require resuscitation, reintuba-
tion, and prolonged postoperative ventilation [29].
The fact that female sex is a significant risk factor for
perioperative morbidity was quite surprising to us; how-
ever, for some reason early results of coronary surgery
are significantly worse in women [30] and women are at
greater risk for acquiring surgical site infections in com-
parison to their male counterparts [31].
The results of our study indicate that cardiopulmonary
bypass has a negative influence on postoperative glycemia
and insulin consumption in both diabetic and nondiabetic
patients. Postoperative complications are more common if
elevated glucose levels appear despite aggressive control of
glycemia by insulin infusion. This could happen particu-
larly to women with diabetes, so the use of cardiopulmo-
nary bypass in this group requires strict glycemic control in
the early postoperative period. Elevated blood glucose
levels in the postoperative period should be aggressively
treated, as they are among the few factors significantly
associated with early mortality and morbidity in patients
undergoing coronary artery surgery.
References
1. Malmberg K, Ryden L, Wedel H, et al. Intense metabolic
control by means of insulin in patients with diabetes melli-
tus and acute myocardial infarction (DIGAMI 2): effects on
mortality and morbidity. Eur Heart J 2005;26:650 – 61.
2. Estrada CA, Young JA, Nifong LW, Chitwood WR Jr. Out-
comes and perioperative hyperglycemia in patients with or
without diabetes mellitus undergoing coronary artery by-
pass grafting. Ann Thorac Surg 2003;75:1392–9.
3. Furnary AP, Gao G, Grunkemeier GL, et al. Continuous
insulin infusion reduces mortality in patients with diabetes
undergoing coronary artery bypass grafting. J Thorac Car-
diovasc Surg 2003;125:1007–21.
4. Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous
intravenous insulin infusion reduces the incidence of deep
sternal wound infection in diabetic patients after cardiac
surgical procedures. Ann Thorac Surg 1999;67:352– 60.
5. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Peri-
operative glycemic control and the risk of infectious compli-
cations in a cohort of adults with diabetes. Diabetes Care
1999;22:1408 –14.
6. Ouattara A, Lecomte P, Le Manach Y, et al. Poor intraoper-
ative blood glucose control is associated with a worsened
hospital outcome after cardiac surgery in diabetic patients.
Anesthesiology 2005;103:687–94.
7. Marfella R, Nappo F, De Angelis L, Paolisso G, Tagliamonte
MR, Giugliano D. Hemodynamic effects of acute hypergly-
cemia in type 2 diabetic patients. Diabetes Care 2000;23:658 –
63.
Ann Thorac Surg
2009;87:1859 – 65
8. Gresele P, Guglielmini G, De Angelis M, et al. Acute,
short-term hyperglycemia enhances shear stress-induced
platelet activation in patients with type II diabetes mellitus.
J Am Coll Cardiol 2003;41:1013–20.
9. Jessen ME. Glucose control during cardiac surgery: how
sweet it is. J Thorac Cardiovasc Surg 2003;125:985–7.
10. Van den Berghe G, Wouters PJ, Bouillon R, et al. Outcome
benefit of intensive insulin therapy in the critically ill: insulin
dose versus glycemic control. Crit Care Med 2003;31:359 – 66.
11. Puskas JD, Williams WH, Mahoney EM, et al. Off-pump vs.
conventional coronary artery bypass grafting: early and
1-year graft potency, cost, and quality-of-life outcomes: a
randomized trial. JAMA 2004;291:1841–9.
12. Chamberlain MH, Ascione R, Reeves BC, Angelini GD.
Evaluation of the effectiveness of off-pump coronary artery
bypass grafting in high-risk patients: an observational study.
Ann Thorac Surg 2002;73:1866 –73.
13. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow
S, Salamon R. European system for cardiac operative risk
evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:
9 –13.
14. Swenne CL, Lindholm C, Borowiec J, Carlsson M. Surgical-
site infections within 60 days of coronary artery by-pass graft
surgery. J Hosp Infect 2004;57:14 –24.
15. Carrier M, Pellerin M, Perrault LP, Solymoss BC, Pelletier
LC. Troponin levels in patients with myocardial infarction
after coronary artery bypass grafting. Ann Thorac Surg
2000;69:435– 40.
16. Calandra T, Cohen J, International Sepsis Forum Definition
of Infection in the ICU Consensus Conference. The Interna-
tional Sepsis Forum consensus conference on definitions of
infection in the intensive care unit. Crit Care Med 2005;33:
1538 – 48.
17. Wahab NN, Cowden EA, Pearce NJ, et al. Is blood glucose an
independent predictor of mortality in acute myocardial
infarction in the thrombolytic era? J Am Coll Cardiol 2002;
40:1748 –54.
18. Anderson RE, Brismar K, Barr G, Ivert T. Effects of cardio-
pulmonary bypass on glucose homeostasis after coronary
artery bypass surgery. Eur J Cardiothorac Surg 2005;28:425–
30.
19. Lehot JJ, Piriz H, Villard J, Cohen R, Guidollet J. Glucose
homeostasis: comparison between hypothermic and normo-
thermic cardiopulmonary bypass. Chest 1992;102:106 –11.
20. Srinivasan AK, Grayson AD, Fabri BM. On-pump versus
off-pump coronary artery bypass grafting in diabetic pa-
KNAPIK ET AL 1865
GLYCEMIA AFTER CORONARY SURGERY
tients: a propensity score analysis. Ann Thorac Surg
2004;78:1604 –9.
21. Vermes E, Demaria RG, Martineau R, et al. Increased early
postoperative morbidity with off-pump coronary artery by-
pass grafting surgery in patients with diabetes. Can J Cardiol
2004;20:1461–5.
ADULT CARDIAC
22. Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H,
Apstein CS. Tight glycemic control in diabetic coronary
artery bypass graft patients improves perioperative out-
comes and decreases recurrent ischemic events. Circulation
2004;109:1497–502.
23. McAlister FA, Man J, Bistritz L, Amad H, Tandon P. Diabetes
and coronary artery bypass surgery: an examination of
perioperative glycemic control and outcomes. Diabetes Care
2003;26:1518 –24.
24. Chaney MA, Nikolov MP, Blakeman BP, Bakhos M. At-
tempting to maintain normoglycemia during cardiopulmo-
nary bypass with insulin may initiate postoperative hypo-
glycemia. Anesth Analg 1999;89:1091–5.
25. Bucerius J, Gummert JF, Walther T, et al. Diabetes in
patients undergoing coronary artery bypass grafting. Impact
on perioperative outcome. Z Kardiol 2005;94:575– 82.
26. Carson JL, Scholz PM, Chen AY, Peterson ED, Gold J,
Schneider SH. Diabetes mellitus increases short-term mor-
tality and morbidity in patients undergoing coronary artery
bypass graft surgery. J Am Coll Cardiol 2002;40:418 –23.
27. Guvener M, Pasaoglu I, Demircin M, Oc M. Perioperative
hyperglycemia is a strong correlate of postoperative infec-
tion in type II diabetic patients after coronary artery bypass
grafting. Endocr J 2002;49:531–7.
28. Gandhi GY, Nuttall GA, Abel MD, et al. Intraoperative
hyperglycemia and perioperative outcomes in cardiac sur-
gery patients. Mayo Clin Proc 2005;80:862– 6.
29. Lauruschkat AH, Arnrich B, Albert AA, et al. Prevalence and
risks of undiagnosed diabetes mellitus in patients undergo-
ing coronary artery bypass grafting. Circulation 2005;112:
2397– 402.
30. Czech B, Kucewicz-Czech E, Pacholewicz J, et al. Early
results of coronary artery surgery in women. Kardiol Pol
2007;65:627–33.
31. Bundy JK, Gonzalez VR, Barnard BM, Hardy RJ, DuPont HL.
Gender risk differences for surgical site infections among a
primary coronary artery bypass graft surgery cohort: 1995–
1998. Am J Infect Control 2006;34:114 –21.