Preoperative renal function as a predictor of survival after coronary artery bypass grafting: comparison with a matched general population.

Abstract

Preoperative renal dysfunction is an established risk factor for early and late mortality after revascularization. We studied how renal function affects long-term survival of patients after coronary artery bypass grafting.
Early and late mortality were determined retrospectively among consecutive patients having isolated coronary bypass at a single Dutch institution between January 1998 and December 2007. Patients were stratified into 4 groups according to preoperative renal function. Expected survival was gauged using a general Dutch population group that was obtained from the database of the Dutch Central Bureau for Statistics; for each of our renal function groups, a general population group was assembled by matching for age, gender, and year of operation.
After excluding 122 patients lost to follow-up, 10,626 patients were studied; in 10,359, preoperative creatinine clearance could be calculated. Multivariate logistic regression and Cox regression analysis identified renal dysfunction as a predictor for early and late mortality. When long-term survival of patient groups was compared with expected survival, only patients with a creatinine clearance less than 30 mL x min(-1) showed a worse outcome. Patients with a creatinine clearance between 60 and 90 mL x min(-1) had a long-term survival exceeding the expected survival.
Severity of renal dysfunction was related to poor survival. When compared with expected survival, however, patients having coronary bypass had a worse outcome only when severe preoperative renal dysfunction was present.

Full text

Page 1

Preoperative renal function as a predictor of survival after
coronary artery bypass grafting: Comparison with a matched
general population
Albert H. M. van Straten, MD,aMohamed A. Soliman Hamad, MD,aAndre ´ A. J. van Zundert, MD, PhD,
FRCA,bElisabeth J. Martens, PhD,cJacques P. A. M. Scho ¨nberger, MD, PhD,aand Andre M. de Wolf, MDd
Objective: Preoperative renal dysfunction is an established risk factor for early and late mortality after revas-
cularization. We studied how renal function affects long-term survival of patients after coronary artery bypass
grafting.
Methods: Early and late mortality were determined retrospectively among consecutive patients having isolated
coronary bypass at a single Dutch institution between January 1998 and December 2007. Patients were stratified
into 4 groups according to preoperative renal function. Expected survival was gauged using a general Dutch pop-
ulation group that was obtained from the database of the Dutch Central Bureau for Statistics; for each of our renal
function groups, a general population group was assembled by matching for age, gender, and year of operation.
Results: After excluding 122 patients lost to follow-up, 10,626 patients were studied; in 10,359, preoperative
creatinine clearance could be calculated. Multivariate logistic regression and Cox regression analysis identified
renal dysfunction as a predictor for early and late mortality. When long-term survival of patient groups was com-
pared with expected survival, only patients with a creatinine clearance less than 30 mL $ min?1showed a worse
outcome. Patients with a creatinine clearance between 60 and 90 mL $ min?1had a long-term survival exceeding
the expected survival.
Conclusions: Severity of renal dysfunction was related to poor survival.When compared with expected survival,
however, patients having coronary bypass had a worse outcome only when severe preoperative renal dysfunction
was present.
Severe renal dysfunction, particularly dysfunction requiring
renalreplacementtherapy,hasbeenidentifiedasa riskfactor
for worse outcome after coronary artery bypass grafting
(CABG).1-3Renal dysfunction not requiring dialysis also
has been identified as a predictor of poor outcome.4-8Re-
cently, the effect of preoperative and postoperative renal
function on late survival has been described in large patient
groups.9,10In these studies, a strong correlation was found
between age and deterioration of renal function. Because ad-
vanced age has a strong influence on late survival after
CABG, interpretation of long-term survival curves compar-
ing groups of patients stratified by renal function might be
difficult, despite appropriate statistical tests. Postoperative
survivalvariesnotonlywithagebutalsoaccordingtogender,
year of surgery, and country. We therefore studied early and
late survival of our Dutch patients who were post-CABG,
stratifiedbypreoperativecreatinineclearance(CrCl),incom-
parisonwithageneralDutchpopulationthatwasmatchedfor
age, gender, and year of operation.
PATIENTS AND METHODS
Patients
We analyzed data concerning consecutive patients having isolated
CABG in a single center between January 1998 and December 2007. Ap-
proval for this study was obtained from the institution’s research review
board. Starting in 1998, clinical data including demographic information,
risk factors, and complications were collected prospectively in a database.
Preoperativeserum creatinineconcentrationswere retrievedfromthe hospi-
tal laboratory database. To obtain survival information for general popula-
tion matched with our patient groups for age, gender, and year of
operation, we used the database of the Dutch Central Bureau for Statistics,
which can be downloaded online (www.CBS.nl). We considered this gen-
eral population data to represent expected survival for our patient groups.
Operative Techniques
All patients received short-acting anesthetic drugs to facilitate early extu-
bation.Extracorporealcirculation(ECC) wasperformedusingnormothermic
nonpulsatile flow. Cold antegrade crystalloid cardioplegia (St Thomas’ Hos-
pitalsolution)orwarmintermittentantegradebloodcardioplegiawasusedac-
cordingtothesurgeon’spreferencetoinduceandmaintaincardioplegicarrest.
All patients having CABG with use of ECC received a low dose of aprotinin
From the Department of Cardio-Thoracic Surgery,aCatharina Hospital, Eindhoven,
The Netherlands; Department of Anesthesiology,bCatharina Hospital–Brabant
Medical School, Eindhoven, The Netherlands, and University Hospital Ghent,
Ghent, Belgium; Department of Education and Research,cCatharina Hospital,
Eindhoven, The Netherlands, and Center of Research on Psychology in Somatic
Diseases,DepartmentofMedicalPsychology,TilburgUniversity,TheNetherlands;
and Department of Anesthesiology,dFeinberg School of Medicine, Northwestern
University, Chicago, Ill.
Received for publication Dec 24, 2008; revisions received March 24, 2009; accepted
for publication May 20, 2009; available ahead of print July 6, 2009.
Address for reprints: Mohamed A. Soliman Hamad, MD, Department of Cardiotho-
racic Surgery, Catharina Hospital, Michelangelolaan 2, Postbus 1350, 5602 ZA
Eindhoven, The Netherlands (E-mail: aasmsn@cze.nl).
J Thorac Cardiovasc Surg 2009;138:971-6
0022-5223/$36.00
Copyright ? 2009 by The American Association for Thoracic Surgery
doi:10.1016/j.jtcvs.2009.05.026
The Journal of Thoracic and Cardiovascular SurgerycVolume 138, Number 4 971
PM
van Straten et alPerioperative Management

Page 2

Abbreviations and Acronyms
CABG ¼ coronary artery bypass grafting
CBS
¼ Central Bureau for Statistics
CKD
¼ chronic kidney disease
COPD ¼ chronic obstructive pulmonary disease
CrCl
¼ creatinine clearance
ECC
¼ extracorporeal circulation
EF
¼ ejection fraction
PVD
¼ peripheral vascular disease
(2 millionkallikrein-inactivatingunits) administeredintheprimingsolution
for ECC. Patients having off-pump surgery did not receive aprotinin.
Estimation of Renal Function
Renal function was determined by calculating CrCl using the Cockcroft
and Gault formula,11which differs between men [(140?age) 3 weight/(se-
rum creatinine372)] and women [(140 ? age) 3 weight/(serum crea-
tinine 372)30.85]. The latest available preoperative serum creatinine
concentration was used for calculation. Patients were grouped according
to chronic kidney disease (CKD) class as defined by the National Kidney
Foundation Disease Outcome Quality Initiative Advisory Board.12Patients
withCrCl above90 mL$ min?1wereplaced inclass1(normal function);60
to 90 mL $ min?1, class 2 (mild dysfunction); 30 to 59 mL $ min?1, class 3
(moderate dysfunction); and 15 to 29 mL $ min?1, class 4 (severe dysfunc-
tion). Patients with CrCl below 15 mL $ min?1and those treated with dial-
ysis were excluded from this study.
Follow-up
Follow-up data concerning mortality were gathered using databases of
health insurance companies. Initially, data for 9% of our total patient group
could not be retrieved from these databases; we then contacted patients’
generalpractitionerstoobtaininformationaboutmortalityand,ifnecessary,
government authorities in the cities where patients lived at the time of the
operation.Earlymortalitywasdefinedasdeathoccurringupto30dayspost-
operatively for any cause, and late mortality was defined as death from any
cause occurring after 30 days.
Statistical Analyses
Discrete variables were compared with the chi-square test and are pre-
sented as numbers and percentages of patients. Continuous variables were
compared with t test and analysis of variance and are presented as the
mean ? standard deviation. Univariate and multivariate logistic regression
analyses were performed to investigate the impact of biomedical variables
on early mortality. Univariate analyses were used to test for the potentially
confounding effect of biomedical and demographic factors on outcome.
Cox proportional hazard regression analyses, which are designed for anal-
ysis of time until an event, were performed to identify risk factors for late
mortality. If P<.05, confounding variables were included in the multivar-
iate logistic and Cox regression analyses to predict mortality. The cumula-
tive incidence of death was depicted using the Kaplan–Meier method,
comparingdifferencesbetweenpreoperativeCKDclasseswithlog-ranksta-
tistics.‘‘Timezero’’wasusedtodesignatethetimeofCABG.Hazardratios
are reported with 95% confidence intervals. All statistical analyses were
performed using SSPS version 15.0 (Chicago, Ill).
RESULTS
Between January 1998 and December 2007, 10,626
patients had isolated CABG at our hospital. In 267 patients,
either preoperative CrCl or weight was lacking. After ex-
cluding patients who were lost to follow up (n ¼ 122),
10,057 patients were evaluated. Almost all patients lost to
follow-up were citizens of foreign countries or were living
abroad. The minimum follow-up interval for surviving
patients was 60 days. The mean follow-up period was
1696 ? 1026 days (range, 0–3708, with 0 days representing
operative death).Sixty-sixpatients(0.6%)neededpostoper-
ative dialysis.
Baseline characteristics stratified by preoperative CKD
class are shown in Table 1. Patients with impaired preoper-
ative renal function (classes 2–4) were older; more often fe-
male; more likely to have diabetes, hypertension, chronic
obstructive pulmonarydisease(COPD),orperipheralvascu-
lar disease (PVD); more likely to have a left ventricular ejec-
tion fraction (EF) below 35%; and more likely to require
perioperative intra-aortic balloon pump support. The num-
ber of off-pump coronary artery bypass operations was
less in patients with renal impairment.
Earlyandlatemortalitiesweregreaterinpatientsinhigher
CKD classes (Table 2). Risk factors for early mortality iden-
tified by univariate and multivariate logistic regression anal-
yses are shown in Table 3. Univariate logistic regression
analysis identified preoperative CrCl, considered as a con-
tinuous variable, as a risk factor for early mortality. Other
patient-related risk factors indicated by univariate analysis
included, age, female gender, COPD, diabetes, EF below
35%, PVD, previous cardiac surgery, year of operation,
andemergencycircumstancesofCABG.Perioperative com-
plications, such as myocardial infarction, need for intraoper-
ative intra-aortic balloon pump support, and reexploration,
also were identified as risk factors for early mortality. On
the other hand, hypertension, use of extracorporeal circula-
tion, and cold crystalloid versus warm intermittent blood
cardioplegia were not identified as risk factors for early mor-
tality.
Allpreoperativerisk factors identified by univariatelogis-
tic regression analysis were entered in the multivariate logis-
tic regression model. Preoperative CrCl, considered as
a continuous variable, proved to be an independent risk fac-
tor for early mortality; others were age, COPD, diabetes, EF
below 35%, previous cardiac surgery, and emergency
CABG but not PVD or the year of operation.
Results of Cox regression analyses to identify risk factors
for late mortality are shown in Table 4. Univariate analysis
identified preoperative CrCl as a risk factor for late mortal-
ity. Other patient-related risk factors identified included
age, year of operation COPD, diabetes, EF below 35%,
PVD, female gender, hypertension, previous cardiac sur-
gery, and emergency CABG. Perioperative complications
representing risk factors for early mortality also were identi-
fied as risk factors for late mortality.
All preoperative risk factors identified by univariate anal-
ysis were entered into the multivariate Cox regression
Perioperative Managementvan Straten et al
972 The Journal of Thoracic and Cardiovascular SurgerycOctober 2009
PM

Page 3

model. Preoperative CrCl considered as a continuous vari-
ablewasidentifiedasanindependentriskfactorforlatemor-
tality. Other risk factors were age, COPD, year of operation
diabetes,EFbelow35%,PVD,malegender,andemergency
operation. Hypertension and previous cardiac surgery were
not identified as independent risk factors.
Figure 1 shows long-term survival stratified by preopera-
tive CKD class. The log-rank test yielded a P value below
.000, indicating significant differences in long-term survival
between all groups. When long-term survival of patients in
classes 1 and 2 was compared with the expected survival
(Figure 2), survival of patients with normal function (class
1) was similar to that expected, with a P value of .323. Sur-
vival of patients with mild impairment (class 2) was better
than expected, with a P value of .002. Compared with the
expected survival, long-term survival of patients with mod-
erate impairment (class 3; Figure 3) was similar, with a P
value of .385. However, survival of patients with severe im-
pairment (class 4; Figure 3) was worse than the expected
survival (P<.0001; log-rank test).
Table 5 shows 1-, 5-, and 10-year survival rates corre-
sponding to the Kaplan-Meier curves.
DISCUSSION
Preoperative renal impairment is a well-established pre-
dictor of adverse outcome in patients having CABG,4-9as
was confirmed in this study. However, when compared
with expected survival, only severe renal impairment led
to worse outcome, and patients with moderate renal impair-
ment had an outcome similar to that expected.
Renal Impairment and Early Mortality
Renal impairment is a well-established risk factor for
early mortality. Currently used risk stratification algorithms
use a serum creatinine concentration exceeding 200 mmol/L
as a cutoff point,13but recent studies suggest that even mild
renal dysfunction predicts poor early outcome after
CABG.4-9This finding was confirmed in our study. How re-
nal dysfunction is related to poor early outcome after CABG
has not been well established. Possibly, renal dysfunction
could be a result of myocardial dysfunction or other sys-
temic diseases. Earlier reports14stated that patients with re-
nal dysfunction have a high prevalence of traditional
cardiovascular risk factors. In our study, the percentages of
patients with impaired left ventricular function, PVD, and
diabetes weremuch higher inpatients inclass 4thanpatients
in other classes. The need for postoperative dialysis in the
entire study population was 0.6%. This finding is in a accor-
dancewithpreviousreports.14Newriskstratificationmodels
should incorporate preoperative CrCl as a continuous but
nonlinear variable.
Renal Impairment and Late Mortality
In agreement with others,9we noted that patients in
a higher preoperative CKD class had poorer late survival.
Renal impairment has been described as a strong risk factor
TABLE 1. Baseline characteristics stratified by preoperative CKD class
Class 1 (n ¼ 2324)
55.7 ? 8.3
2113 (90.9)
486 (20.9)
946 (40.7)
231 (9.9)
201 (8.6)
46 (2.0)
97 (4.2)
298 (12.4)
797 (33.8)
1184 (50.2)
98 (4.2)
33 (1.4)
102 (4.4)
52 (2.2)
Class 2 (n ¼ 4819)
64.4 ? 7.8
3929 (81.5)
932 (19.3)
1885 (39.1)
578 (12.0)
496 (10.3)
139 (3.0)
165 (3.4)
411 (8.5)
2035 (41.8)
2212 (45.4)
294 (6.1)
83 (1.7)
253 (5.3)
144 (3.0)
Class 3 (n ¼ 2798)
71.8 ? 6.4
1660 (59.3)
650 (23.2)
1279 (45.7)
420 (15.0)
439 (15.7)
133 (4.9)
98 (3.5)
186 (6.6)
1368 (48.5)
1154 (40.9)
205 (7.3)
86 (3.1)
181 (6.5)
87 (3.1)
Class 4 (n ¼ 116)
74.8 ? 6.0
51 (44.0)
42 (36.2)
73 (62.9)
17 (14.7)
26 (22.4)
12 (10.6)
8 (6.9)
4 (3.4)
55 (46.6)
54 (45.8)
10 (8.6)
5 (4.3)
4 (3.4)
6 (5.2)
P value
Age, years (mean ? SD)
Male
Diabetes
Hypertension
COPD
PVD
LVEF<35%
Emergency CABG
Off-pump
Crystalloid cardioplegia
Blood cardioplegia
Previous cardiac surgery
IABP
Reexploration
Periop MI
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
.007
.096
Values are expressed as number of patients (percentage) unless otherwise mentioned. CKD, Chronic kidney disease; COPD, chronic obstructive pulmonary disease; PVD, periph-
eralvascular disease;LVEF,leftventricularejectionfraction;CABG,coronaryartery bypassgrafting;IABP,intra-aorticballoonpumpsupport;PeriopMI,perioperativemyocardial
infarction.
TABLE 2. Early and late mortality stratified by preoperative CKD class
Class 1 (n ¼ 2324)
14 (0.6%)
85 (3.7%)
Class 2 (n ¼ 4819)
86 (1.8%)
438 (9.1%)
Class 3 (n ¼ 2798)
112 (4.0%)
535 (19.1%)
Class 4 (n ¼ 116)
13 (11.2%)
52 (44.8%)
P value
Early mortality
Late mortality
<.0001
<.0001
CKD, Chronic kidney disease.
van Straten et alPerioperative Management
The Journal of Thoracic and Cardiovascular SurgerycVolume 138, Number 4 973
PM

Page 4

for development of cardiovascular disease.15,16It occurs
more often in advanced age and in patients with a number
of other important cardiovascular risk factors including hy-
percoagulability,17endothelial dysfunction,18arteriosclero-
sis,19and left ventricular hypertrophy.20This means
increased incidence of new cardiovascular events for those
who survive the postoperative period. Holzmann and col-
leagues21showed an 8% incidence of myocardial infarction
in patients with renal dysfunction within 5 years after
CABG. In 35% of the cases, death was the outcome within
1 month. It is also possible that in the presence of renal dys-
function, traditional risk factors such as smoking, hyperlip-
idemia, hypertension, and diabetes will have a different
qualitative relationship with cardiovascular events.22Our
data show that higher CKD class is associated with greater
mean age.
In ourstudy,wefound that no patients inthe class 4group
survived 10 years, and less than 50% survived 5 years, al-
though the same operative and perioperative techniques
were adopted in all groups. These results must be weighed
TABLE 3. Predictors of early mortality (?30 days after CABG) by univariate and multivariate logistic regression analyses
Univariate analysis Multivariate analysis
Risk factorHR (95% CI)P valueHR (95% CI)P value
Preop CrCl*
Year of operation
Age*
COPD
Diabetes
LVEF<35%
PVD
Previous cardiac surgery
Emergency
Periop MI
Reexploration
IABP
Male gender
Hypertension
Crystalloid cardioplegia
Off-pump
0.965 (0.958–0.972)
0.944 (0.901–0.989
1.083 (1.065–1.102)
2.06 (1.49–2.84)
1.51 (1.12–2.02)
6.15 (4.18–0.04)
1.62 (1.14–2.30)
4.26 (3. 11–5.83)
6.56 (4.21–8.70)
6.42 (4.35–9.49)
5.29 (3.81–7.36)
14.41 (10.01–20.74)
0.74 (0.55–0.99)
0.97 (0.74–1.27)
1.18 (0.89–1.57)
0.67 (0.39–1.17)
<.0001
.015
<.0001
<.0001
.006
<.0001
.007
<.0001
<.0001
<.0001
<.0001
<.0001
.047
.745
.242
.163
0.980 (0.971–0.990)
0.965 (0.917–1.016)
1.055 (1.031–1.079)
1.78 (1.26–2.50)
1.49 (1,08–2.04)
4.19 (2.76–6.36)
1.30 (0.89–1.90)
2.49 (1.66–3.74)
3.73 (2.30–6.04)
<.0001
.175
<.0001
.001
.013
<.0001
.171
<.0001
<.0001
1.15 (0.83–1.61).385
CABG, Coronary artery bypass grafting; COPD, chronic obstructive pulmonary disease; HR, hazard ratio; CI, confidence interval; IABP, intra-aortic balloon pump support; LVEF,
left ventricular ejection fraction; Periop MI, perioperative myocardial infarction; Preop CrCl, preoperative creatinine clearance; PVD, peripheral vascular disease. *Entered as
a continuous variable.
TABLE 4. Predictors of late mortality (>30 days after CABG) by univariate and multivariate Cox regression analyses
Univariate analysisMultivariate analysis
Risk factorHR (95% CI)P value HR (95% CI)P value
Preop CrCl*
Year of operation
Age*
COPD
Diabetes
LVEF<35%
PVD
Male gender
Hypertension
Previous cardiac surgery
Emergency CABG
Off-pump
Periop MI
Reexploration
Crystalloid cardioplegia
IABP
0.966 (0.964–0.969)
0.954 (0.931–0.978)
1.092 (1.084–1.099)
1.95 (1.71–2.23)
1.68 (1.49–1.89)
3.15 (2.59–3.83)
2.12 (1.85–2.43)
0.84 (0.75–0.95)
1.17 (1.04–1.30)
1.95 (1.66–2.31)
1.95 (1.58–2.41)
0.69 (0.54–0.88)
2.61 (2.09–3.27)
1.96 (1.64–2.35)
1.10 (0.98–1.23)
3.69 (2.96–4.61)
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
.008
.005
<.0001
<.0001
.003
<.0001
<.0001
.097
<.0001
0.984 (0.980–0.988)
0.952 (0.928–0.977)
1.067 (1.057–1.077)
1.66 (1.45–1.90)
1.52 (1.34–1.72)
2.58 (2.12–3.15)
1.61 (1.40–1.86)
1.39 (1.22–1.59)
1.07 (0.96–1.20)
1.47 (1.23–1.20)
1.92 (0.50–2.47)
<.0001
<0.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
.201
<.0001
<.0001
CABG, Coronary artery bypass grafting; COPD, chronic obstructive pulmonary disease; HR, hazard ratio; CI, confidence interval; IABP, intra-aortic balloon pump support; LVEF,
left ventricular ejection fraction; Periop MI, perioperative myocardial infarction; Preop CrCl, preoperative creatinine clearance; PVD, peripheral vascular disease. *Entered as
a continuous variable.
Perioperative Managementvan Straten et al
974The Journal of Thoracic and Cardiovascular SurgerycOctober 2009
PM

Page 5

against alternative procedures, namely percutaneous coro-
naryintervention(PCI). PriorPCIstudieshavefoundthatal-
though procedural success rates tend to be high in patients
with renal dysfunction, these patients face higher risks for
mortality, morbidity, and restenosis.23,24The risk of revas-
cularization has also to be weighed against the long-term
hazards of conservative treatment in patients with coronary
artery disease. In this regard, Hemmelgarn and colleagues25
reported worse 8-year adjusted survival rates for patients
with kidney disease who were not on dialysis but were med-
ically managed after catheterization compared with those
who had revascularization (54.9% for CABG and 32.7%
for PCI versus 29.7% for no revascularization).
Althoughappropriatestatisticaltestscanbeusedtodistin-
guish between effects caused by age and by renal function,
interpretation of survival curves remains difficult. Life ex-
pectancy also may vary with age, gender, and race. Chang-
ing life expectancy over the years has been well documented
in the Netherlands by the Central Bureau for Statistics
(CBS), which keeps track of mortality rates for the entire
Dutch population. We used the CBS database, which can
be downloaded online (www.CBS.nl), to calculate survival
in the general population groups. We compared late survival
of patients stratified by CKD class with general population
FIGURE 1. Kaplan–Meier survival curves stratified by preoperative
chronic kidney disease (CKD) class.
FIGURE 2. Kaplan–Meier curves for patients with chronic kidney disease
(CKD) class 1 and 2 compared with matched general population groups.
*Expected for patients in CKD class 1; **expected for patients in CKD
class 2.
FIGURE 3. Kaplan–Meier curves for patients with chronic kidney disease
(CKD) class 3 and 4 compared with matched general population groups.
*Expected for patients in CKD class 3; **expected for patients in CKD
class 4.
TABLE 5. Survival rates (%) for 1, 5, and 10 years, stratified by CKD
class
1 y 5 y10 y
Class 1
Class 2
Class 3
Class 4
98.1 ? 0.3
95.0 ? 0.3
89.5 ? 0.6
62.4 ? 4.3
94.9 ? 0.6
86.0 ? 0.2
76.7 ? 1.0
44.3 ? 5.4
90.1 ? 1.4
79.4 ? 1.2
53.2 ? 3.8
*
CKD, Chronic kidney disease. *No patient in CKD class 4 survived for 10 years.
van Straten et alPerioperative Management
The Journal of Thoracic and Cardiovascular SurgerycVolume 138, Number 4 975
PM

Page 6

groups matched for age, gender, and year of operation. Pa-
tients in CKD class 1 were relatively young, and survival
was similar to that in a matched general population (ex-
pected survival). Kaplan–Meier curves and log-rank tests
suggested that patients in class 2 had better than expected
survival. Patients in class 3 were older than patients in class
2 and showedworse survival, butone that still was similarto
the expected survival (Kaplan-Meyer curves and log-rank
test). For patients in class 4, survival was far worse than
the expected survival.
Caution to interpret these results is needed, because the
CBS database contains data of the total Dutch population
including patients described in this study as well as patients
who were treated in other Dutch cardiac surgery centers. If
differences between groups are found, the results of the gen-
eral population group shift toward the average of the patient
group, making the differences smaller than what our study
shows. In comparing late survival of our patients with out-
come in a matched group from the general population, we
should take into account that a certain but unknown percent-
age of the general population suffers from renal impairment
and coronary disease. One also may assume that our patients
were adequatelytreated for their coronary problem and post-
operatively were treated with aspirin, hypocholesterolemic
drugs, and medications for underlying diseases such as hy-
pertension. This treatment, as well as protection by revascu-
larization, may contribute to improved survival and explain
why patients inCKDclasses1,2,and3hadasimilar oreven
bettersurvival thanmatched general populationgroups.Fur-
ther, before having CABG, patients are screened for severe
underlying disease. If severe underlying disease is present,
treatment alternatives to CABG are considered, thus biasing
the CABG group.
Renal impairment was confirmed to be a risk factor for
early mortality after CABG. Concerning late outcome, low
CrCl was associated with a poor survival. When compared
with groups from the general population matched for age,
gender, and year of CABG, only patients with a CrCl of
15 to 30 mL $ min?1had a worse long-term survival, and
patients with a CrCl of over 30 mL $ min?1had a similar
or better outcome than expected.
Study Limitations
This study had a retrospective design. Some patients were
lost to follow-up and some data were unobtainable, making
results less accurate.
References
1. NakayamaL, Sakata R,Ura M,Miyamoto TA. Coronary artery bypass grafting in
dialysis patients. Ann Thorac Surg. 1999;68:1257-61.
2. Rinehart AL, HerzogCS,Collins AJ, FlackJM, Ma JZ, Opsahl JA.A comparison
of coronary angioplasty and coronary artery bypass grafting outcomes in chronic
dialysis patients. Am J Kidney Dis. 1995;25:281-90.
3. Labrousse L, de Vincentiis C, Madonna F, Deville C, Roques X, Baudet E. Early
and long-term results of coronary artery bypass grafts in patients with dialysis de-
pendent renal failure. Eur J Cardiothorac Surg. 1999;15:691-6.
4. Rao V, Weisel RD, Buth KJ, Cohen G, Borger MA, Shiono N, et al. Coronary
artery bypass grafting in patients with non-dialysis-dependent renal insufficiency.
Circulation. 1997;96(suppl II):38-43.
5. Weerasinghe A, Hornick P, Smith P, Taylor K, Ratnatunga C. Coronary artery
bypass grafting in non-dialysis-dependent mild-to-moderate renal dysfunction.
J Thorac Cardiovasc Surg. 2000;121:1083-9.
6. Nakayama Y, Sakata R, Ura M, Itoh T. Long-term results of coronary artery
bypass grafting in patients with renal insufficiency. Ann Thorac Surg. 2003;75:
496-500.
7. HiroseH, Amano A,TakahashiA, Nagano N. Coronaryartery bypass grafting for
patients with non-dialysis-dependent renal dysfunction (serum creatinine ?2.0
mg/dl). Eur J Cardiothorac Surg. 2000;20:565-72.
8. Szczech LA, Best PJ, Crowley E, Brooks MM, Berger PB, Bittner V, et al. for the
Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Out-
comes of patients with chronic renal insufficiency in the Bypass Angioplasty Re-
vascularization Investigation. Circulation. 2002;105:2253-8.
9. Howell NJ, Keogh BE, Bonser RS, Graham TR, Mascaro J, et al. Mild renal dys-
function predicts in-hospital mortality and post-discharge survival following car-
diac surgery. Eur J Cardiothorac Surg. 2008;34:390-5.
10. Brown JR, Cochran RP, Mackenzie TA, Furnary AP, Kunzelman KS, Ross CS,
et al. Long-term survival after cardiac surgery is predicted by estimated glomer-
ular filtration rate. Ann Thorac Surg. 2008;86:4-11.
11. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creati-
nine. Nephron. 1976;16:31-41.
12. NationalKidneyFoundation.K/DOQIclinicalpracticeguidelinesforchronickid-
ney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;
39(suppl 1):1-266.
13. Nashef SAM, Roques F, Michel P, Gauducheau E, Lemeshow S, Salomon R. Eu-
ropean system for cardiac operative risk evaluation (EuroScore). Eur J Cardio-
thorac Surg. 1999;16:9-13.
14. Zakeri R, Freemantle N, Barnett V, Lipkin GW, Bonser RS, Graham TR, et al.
Relation between mild renal dysfunction and outcomes after coronary artery
bypass grafting. Circulation. 2005;112(9 suppl):I270-5.
15. FoleyRN,MurrayAM,LiS,HerzogCA,McBeanAM,EggersPW,etal.Chronic
kidneydiseaseandtheriskforcardiovasculardisease,renalreplacementanddeath
in the United States Medicare population, 1998 to 1999. J Am Soc Nephrol. 2005;
16:489-95.
16. Go AS, Chertow GM, Fan D, McCullogh CE, Hsu CY. Chronic kidney disease
and the risk of death, cardiovascular events and hospitalization. N Engl J Med.
2004;351:1296-305.
17. Shlipak MG, Fried LF, Crump C, Bleyer AJ, Manolio TA, Tracy RP, et al.
Elevations of inflammatory and procoagulant biomarkers in elderly persons
with renal insufficiency. Circulation. 2003;107:87-92.
18. Blacher J, Safar ME, Guerin AP, Pannier B, Marchais SJ, London GM. Aortic
pulse wave velocity index and mortality in end-stage renal disease. Kidney Int.
2003;63:1852-60.
19. Raggi P, Boulay A, Chasan-Taber S, Amin N, Dillon M, Burke SK, et al. Cardiac
calcification in adult hemodialysis patients. A link between end-stage renal
disease and cardiovascular disease? J Am Coll Cardiol. 2002;39:695-701.
20. Levin A, Thompson CR, Ethier J, Carlisle EJ, Tobe S, Mendelssohn D, et al. Left
ventricular mass index increase in early renal disease: impact of decline in hemo-
globin. Am J Kidney Dis. 1999;34:125-34.
21. Holzmann MJ, Hammar N, Ahnve S, Nordqvist T, Pehrsson K, Ivert T. Renal
insufficiency and long-term mortality and incidence of myocardial infarction in
patients undergoing coronary artery bypass grafting. Eur Heart J. 2007;28:
865-71.
22. Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, et al.
Kidney disease as a risk factor for development of cardiovascular disease: a state-
ment from theAmerican Heart Association CouncilsonKidney inCardiovascular
Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology
and Prevention. Circulation. 2003;108:2154-69.
23. RinehartAL,Herzog CA, CollinsAJ, Flack JM,MaJZ,OpsahlJA. Acomparison
of coronary angioplasty and coronary artery bypass grafting outcomes in chronic
dialysis patients. Am J Kidney Dis. 1995;25:281-90.
24. Reusser LM, Osborn LA, White HJ, Sexson R, Crawford MH. Increased morbid-
ity after coronary angioplasty in patients on chronic hemodialysis. Am J Cardiol.
1994;73:965-7.
25. Hemmelgarn BR, Southern D, Culleton BF, Mitchell LB, Knudtson ML,
Ghali WA. Provincial Project for Outcomes Assessment in Coronary Heart Dis-
ease (APPROACH) Investigators. Survival after coronary revascularization
among patients with kidney disease. Circulation. 2004;110:1890-5.
Perioperative Managementvan Straten et al
976The Journal of Thoracic and Cardiovascular SurgerycOctober 2009
PM