1

EFFECTS - IMPACT OF DIABETES MELLITUS ON THE FREQUENCY OF POST-

OPERATIVE COMPLICATIONS AFTER CAROTID ENDARTERECTOMY
Short title: Influence of Diabetes on Carotid Endarterectomy
Abstract
Background: There are significant controversies surrounding the connexion of insulin-
dependent and insulin-non-dependent diabetes mellitus with post-operative complications
after carotid endarterectomy. The aim of this study is identification of possible effects of
diabetes on the frequency of post-operative complications after CEA, as well as evaluation of
30-day and 1-year long outcomes.
Methods: Prospective study has been conducted at the IKVB "Dedinje" clinic for vascular
surgery, Belgrade. The patients who underwent CEA in the time period from December 2012
until December 2013 were divided into two groups: Group A (37.7 %) consisted of 279
diabetics, including 98 (35,1%) insulin-dependent (IDDM) and 181 (64,9%) insulin-
independent ones (IIDM); and Group B (62,3%) consisted of 461 non-diabetics.
Results: Pre-operative characteristics have mostly been similar, except for a greater
prevalence of dyslipidemia in patients with diabetes (Hi-square=5,330; p=0,021), more
persistent coronary disease (Hi-square=15,090; p<0,001), and more persistent peripheral
arterial disease (Hi-square=20,607; p<0,001). Post-operative cardiac events (myocardial
infarction, coronary insufficiency) occurred (3,6%) in patients with diabetes (1,1%) and non-
diabetic patients (p= 0,039); post-operative neurological events (TIA) among patients with
diabetes (3,6%) and among non-diabetics (0,9%, p=0,009). Perioperative mortality rate has
been 2,5% in the diabetics' group (4 cardiac deaths + 3 strokes) and 0,9% in the non-diabetics'
group (3 myocardial infarctions + 1 stroke, value p=0,113). The total percentage of post-
operative complications was two or more times higher in the diabetics' group - 8,5% vs
18,3%, (p<0,001). In total, post-operative complications occurred in 25,4% of patients on oral
antidiabetic medications, as compared to 8,2% of patients on insulin (p<0,001).
Conclusions: Diabetes mellitus significantly increases surgical risks of CEA. Higher rate of
mortality and post-operative complications has been observed in patients being treated with
oral antidiabetics than in those on insulin.
Keywords: CEA, diabetes mellitus, post-operative complications.
Introduction
Stenotic occlusive diseases of carotid arteries are some of the most frequent surgical vascular
diseases, with a tendency to increase in numbers, and carotid endarterectomy (CEA) is the
most frequent operation carried out at clinics for vascular surgery (1). It is well-known that
diabetes mellitus (DM) is a progressive disease which significantly contributes to degradation
of atherosclerotic changes in arteries (2).
CEA represents the first choice of treatment of high-level carotid stenosis with a low rate of
post-operative morbidity and mortality (3,4). According to the literature data, the percentage
of patients with DM, within the group of patients who have undergone CEA, has been on the
 2

rise, ranging from 10% to over 25% (5,6). On the other hand, diabetes is an important risk
factor for occurrences of myocardial infarction and stroke in general population, and it can
also impact the outcome of CEA (7). Some authors think that the patients suffering from DM
are at up to two times greater risk of overall post-operative complications as compared to
patients with no diabetes (7, 8,9,10,11,12). Contrary to these statements, some other authors
observe that the patients suffering from DM are exposed to the same risk of cardiological
morbidity, mortality, and stroke after CEA, just as the patients with no diabetes (13, 14, 15).
It has been proven that there is a risk of stroke in diabetics, and it is related to the level of
hyperglycaemia; it is assumed that controlled glycaemia may lower the risk of stroke (16).
Since most of those studies have not taken into consideration type of an applied antidiabetic
therapy, it could explain contradictory results amongst studies on the influence of DM on the
outcome of CEA. This paper's aim is to determine the frequency of post-operative
complications after CEA in patients with or without diabetes mellitus, as well as identify the
type of applied antidiabetic therapy, in order to define the safety of the operation concerned.

Materials and methods

The prospective non-randomised examination has included all the patients treated at IKVB
"Dedinje", Belgrade, who, in the time period of December 2012-December 2013, underwent
CEA, and they have all been entered into the prospective vascular registry. The patients were
divided into two groups: Group A (37.7 %) consisted of 279 diabetics, including 98 (35,1%)
being insulin-dependent and 181 (64,9%) taking oral hypoglycemics; Group B (62,3%)
consisted of 461 non-diabetics. A complete comparison of demographic characteristics and
risk factors has been done (sex, age, hypertension, hyperlipidaemia, smoking, pre-operative
IM, concomitant vascular diseases), and after the operation we compared post-operative
complications (transient ischaemic attack (TIA), cerebro-vascular infarction (CVI),
myocardial infarction (IM), neurological, cardiological and other morbidities, and the total
mortality). The results are shown in Tables. Controlling CDS of neck artery was carried out
one month, six months, and one year later. Our primary goal was a 30-day morbidity and
mortality, and is defined as any death or post-operative morbidity within 30 days after the
treatment, as well as a percentage of carotid restenosis up to one year post-operatively.
Indication for CEA was set after CDS of neck artery, and multislice (MSCT) angiography of
supra-aortic branches. Carotid stenosis has been ascertained according to ESCT criteria, as
well as by means of criteria described by AbuRahma et al. Carotid stenosis has been
determined as significant (>70% of constriction) if systolic velocity was >150 cm/s, and
diastolic velocity >90 cm/s. (17). Perioperative neurological morbidity has been classified as
TIA (lasting less than 24 hours) or permanent stroke (deficiency present on discharge). The
adverse post-operative cardiac event has been designated as a post-operative IM and
congestive heart failure (CHF).
Post-operative restenosis has been defined as an ultrasound-verified stenosis of carotid artery
on the operative side larger than 50%.
Patients with a history of disease of coronary artery have undergone stress tests and, in some
cases, coronarography. Peripheral vascular disease has been proven by means of CDS of
 3

lower extremities and, if deemed necessary, by using angiography. For evaluation of pre-
operative neurological disorder, as well as post-operative neurological condition, a modified
scale of Rankin scores (mRS) has been used, with a neurological damage estimate degree
ranging from 0 to 5 (18).
The patients were operated on in the general endotracheal anesthesia. CEA eversion technique
has been applied on all patients. All the patients with post-operative complications have
undergone a CT of endocranium post-operatively.

Statistical analysis
The testing of statistical hypotheses has made use of: t-test for two independent samples,
Mann Whitney test, Hi-square test and Fisher test of accurate probability. Logistic regression
has been used for analysing the relationship between binary outcomes and potential
predictors. Statistical hypotheses have been tested at the level of statistical significance (alpha
level) of 0,05.

Data availability
Data availability statement: the data related to this paper are not publicly available, but they
are available upon a sensible request to the corresponding author.

DISCUSSION
Although previous studies had been evaluating the connexion between diabetes and a greater
operative risk during a carotid endarterectomy, there appear various or conflicting results in
many studies. Most studies introduce CEA as a well-known and permanent procedure for
prevention of TIA and cerebro-vascular infarction in patients with a significant stenosis of the
carotid artery (19,20).
In previous studies, which included patients operated on for stenotic occlusive disease of
carotid arteries, the percentage of diabetics in most cases was ranging from 13% to 23,6%
(20, 7). Our series includes the percentage of patients with diabetes that reaches 37,7%, and is
considerably higher as compared to the study done by Ahari et al, whose percentage reaches
13% (6). Dorigo et al report that the percentage of those with DM is 20,05%, whereas the
Rockman et al study reports the same percentage as 23.5% (7,14). The study done by Min Jae
et al reports on a high percentage of diabetics in Asia, going up to 39% (21). The high
percentage of diabetics in our study can be explained by the fact that DM has been showing
epidemic proportions in the general population for the last two decades, and especially in the
group of patients with atherosclerotic disease of blood vessels.
Our study shows that pre-operative factors believed to increase the risk of death and stroke in
diabetics include greater values of LDL (p=0,021), coronary disease (p<0,001), and peripheral
arterial disease (p<0,001). The use of statins and a prompt diagnostics of these comorbidities
should have an influence on reduction of mortality and stroke after CEA.
Barret et al report that in perioperative morbidity and mortality cerebro-vascular infarction
and cardiological complications are significantly more persistent in diabetics, if there is a pre-
operative difference in values of cholesterol and cardiovascular diseases (22), while Dorigo et
 4

al state that patients with diabetes were predominantly women, who suffered from coronary
disease, peripheral arterial disease, and hyperlipidaemia (7). Similar to our results, Pistolese et
al report that the only difference occurring pre-operatively is a greater prevalence of
peripheral vascular disease, coronary disease and dyslipidemia in the diabetics' group (11).
Some other authors claim that pre-CEA diagnosed risk factors like atherosclerosis and
diabetes have an effect on 30-day mortality and stroke, but they do not record a significant
influence of dyslipidemia (21).
Our research, however, has identified most frequent post-operative complications, such as
post-operative myocardial infarction, coronary insufficiency, TIA, CVI, respiratory
insufficiency, post-operative bleeding, wound infection - all these complications have been
significantly more persistent in patients with DM.
Post-operative TIA was present in 0,9% of patients without DM and in 3,6% of patients with
DM (p=0,009), while post-operative CVI was 1,3% vs 3,2% (p=0,072). Patients with DM
suffer significantly more often from early post-operative TIA and CVI. A group of authors
reports that transient ischaemic attack and cerebro-vascular infarction mostly occur in
diabetics, the percentage being 3,2% vs 2,4% (24). The greater risk of cerebro-vascular
infarction in patients with DM was reported in the recent Hussain et al review, wherein they
determined that DM was associated with 1,5 times greater risk of stroke after CEA (25).
Contrary to these statements, Ballota et al suggest that there is no important difference in the
frequency of these post-operative complications between diabetics and non-diabetics (14).
Post-operative cardiological complications, IM included, occurred in 1,1% of patients without
DM, and in 3,6% of patients with DM (p=0,039). However, that can be explained by the fact
that our study pool consisted of a group of diabetics whose pre-operative cardiological
complications were usually more persistent. Akbari et al report that diabetic patients
experience greater negative cardiac events than non-diabetics, and point out that the diabetic
patients in their series had a pre-operative history of coronary artery disease (8). Those
authors suggest no significant difference in post-operative cardiac event, and note that after
CEA a consequent coronary insufficiency occurs in 2,3% of cases, with no difference
observed between the groups concerned (26). Ombrellaro et al have recently observed in
post-CEA patients an undesirable cardiac event (MI,CHF) in 14,3% amongst the patients with
diabetes, and with 16% of incidence in the non-diabetics' group, the difference not being
statistically important (27).
Our study has noted 0,7% of non-DM patients, and 2,9% of DM ones suffering from post-
operative complications. As opposed to our findings, Eliasziw et al reach a conclusion that
respiratory complications occur in 0,8% of patients, with no differences noted between
diabetics and non-diabetics (28)
The frequency of haematoma of the operated wound occurred in 2,4% of non-DM patients,
and in 3,6% of DM patients, while the wound infection was present only in patients with DM
(1,8%). Zhao k. wu et al state that the incidence of haematoma was relatively high in
diabetics, the cause being pre-operative high doses of heparin (1 mg/kg), as well as a double
anti-aggregation therapy; the same authors specify that diabetes may increase the possibility
of wound and systemic infection, and that pre-operative glycosylated haemoglobin should be
 5

about 7%, and thus reduce possible infections (23). Most authors report that wound infection
is an extremely rare complication, and one such study found 0,2% of wound infection cases,
with no difference between the two groups concerned (29).
Post-operative 50% carotid restenosis in the course of one year was not significant - it
occurred in 2,2% of non-DM patients, and in 1,8% of DM patients (Hi-square=0,124;
p=0,724). Other investigations produced similar results. The studies on carotid restenosis
published in the last 17 years (16 studies) report the incidence of symptomatic restenosis as
ranging from 1% to 8%, without difference between the investigated groups (2% in average).
(30)
Total post-operative mortality (neurological and cardiological) was present in 0,9% of non-
diabetics and in 2,5% of diabetics (p=0,113). Ahari et al report in their study that diabetics
had higher 30-day mortality (3,2% vs 1,4%; p = 0,02) (6). In a study by Akbari et al, the total
mortality with diabetics is significantly higher (2,1% of diabetics vs 1,1% of non-diabetics; p
= 0,28) (8).
The total post-operative complications are observed in 8,5% of non-diabetics, and in 18,3% of
DM patients, (p<0,001). DM patients are at more than two times greater risk of suffering from
post-operative complications. Dorigo et al find that the risk of post-operative complications is
even two times greater in patients suffering from DM (7). The Min jao et al study concludes
that DM patients are not at greater risk of 30-day morbidities and mortality after CEA than
those without DM (21).
Our patients on oral antidiabetics considerably more often suffer from post-operative
complications (25,4% vs 8,2%), (p<0,001). In the study done by Axelrod et al there appears a
somewhat higher percentage of post-operative complications in diabetics, although without
difference between the patients on insulin and those on oral antidiabetics (31). Dimić et al
have shown the incidence of cumulative rate of TIA/cerebro-vascular infarction (P= 0,02) to
be greater in insulin-dependent diabetics than in those who are insulin-independent (32).
Vallaert et al conclude that patients at the greatest risk are those with risk factors, including
age ≥ 80 and IDDM (33). Similarly, Bennett et al have stated that insulin-necessitated DM is
one of independent predictors of high morbidity and mortality amongst the patients who have
undergone CEA (34). However, Dorigo et al discover that patients with diabetes were at
greater risk of death, but with no difference between the patients with insulin-controlled
diabetes and the ones on oral medications (7). Parlani et al report that patients with IDDM
had higher rates of cerebro-vascular infarction and death (6,5% vs 1,7%; P ¼.02) than non-
diabetics (11). Comparably, Alexander et al specify that patients with IDDM have higher rates
of 30-day cerebro-vascular infarction and death than those without DM (3,4% vs 1,5%; P
<.001) (35)
In our series, a more significant rate of mortality and post-operative complications occurs in
diabetics being treated with oral antidiabetics, as compared to those being treated with insulin,
which conflicts other studies that report a higher frequency of complications in groups on
insulin therapy. To some extent, this can be explained by extremely lengthy and irregular
therapy of patients on oral antidiabetic agents, which again leads to chronic, atherosclerotic
 6

changes on blood vessels, whereas patients on insulin have been more eager in the application
of therapy.
Our study has found that the average duration of DM in patients without post-operative
complications was 9,3±4,1 years, while the average duration of DM in patients with post-
operative complications was 12,5±4,8 years (t=4,967; p<0,001). Parlani et al specify that the
longer the duration of DM, and the more numerous insulin-dependent DM patients, the
greater the risks of post-operative complications and death (11).
In terms of a long-term regulation of glycaemia, our patients with post-operative
complications have significantly higher values of HbA1c (t=5,010; p<0,001). Tanashian et al
report that the presence of DM is associated with an increased risk of ischaemic lesions in the
brain and a higher percentage of post-operative complications, associated with increased
values of glycaemia (8,0 mmol / l) and glycosylated haemoglobin (7,8-8%) in the pre-
operative phase (36). Dimić et al state that the group of diabetics with HbA1c> 7 % has had a
greater cumulative rate of TIA/cerebro-vascular infarction (P = 0,03) (32). Parr et al report
that patients with IDDM, when compared to those with IIDM, have higher rates of cerebro-
vascular infarction (3,27%, 0,93% and 0,94%; P <.0001), MI (3,35%, 1,10% and 0,87%, P
<.0001) and an increased hospital mortality (P <, 0001) (37). Min jao et al have shown that
insulin use is associated with a higher rate of mortality and morbidity. The absence of data on
serial measurements of HbA1c level in the analysis was the reason why they could not explain
differences in glycaemic control (21).
It is certain that high concentrations of lipoprotein and a chronic hyperglycaemia, proven by
high values of HbA1c, help develop atherosclerosis that sets in earlier and is more
comprehensive in diabetics. On the basis of the results of our investigation, it is believed that
the percentage of post-operative complications may be reduced by means of a better
regulation of glycaemia, lower values of HbA1c, a prompt discovery of glucose intolerance,
and a regular application of adequate antidiabetic therapy. Considering the small number of
studies which have dealt with this kind of investigation - complications related to applied
therapy, the assumption remains to be definitively proven in similar future and randomised
studies.
Conclusions
In comparison to all the studies known to us, this paper includes: a large number of patients
for a short time period; a very high percentage of diabetics; an investigation conducted with
diabetics depending on the applied antidiabetic therapy; as well as occurrences of
complications depending on the values of glycosylated haemoglobin, and regarding the
duration of diabetes - which has not been done yet, or only a few studies have examined the
issues.
Diabetes is, it appears, an independent risk factor for cardiac or neurological events (fatal and
non-fatal) after CEA, because it causes 2,5 times greater post-operative complications in the
diabetics' group.
 7

Our study records a higher rate of mortality and post-operative complications occurring in the
diabetics being treated with oral antidiabetics, and having higher values of glycosylated
haemoglobin than in the diabetics on insulin therapy, thus conflicting similar studies.
In the end, this paper's results show that CEA is a reliable and efficient method of surgical
treatment of patients with significant carotid stenosis, and of patients with concomitant
diabetes mellitus, regardless of the type of antidiabetic therapy.

Remarks
Statement on conflict of interest: This article has not been submitted to any other journal, and
all authors have become familiar with the content of this manuscript. The authors declare that
there is no conflict of interest associated with publication of this article.
Approval of the manuscript by all authors
All authors have read and approved of the final version of the manuscript. Its prospective
publication has been approved of by all co-authors.
Contribution of authors
All the named authors have given an important contribution in data collection, analysis, and
commenting. All authors have taken part in composition and peer review of the paper. All of
them have approved of the final version of the paper, and consent to being accountable for all
aspects of the paper; they are also concordant that issues relating to accuracy and integrity of
any part of the paper may be examined and resolved.

RESULTS - TABLES
Table 1 Pre-operative characteristics of diabetics (group A) and non-diabetics (group B)

Group B non-
Characteristics Group A diabetics P Value
diabetics
N 279 (37,7%) 461 (62,3%) ---
Average age 67,5±7,2 66,8±7,5 NS
Male 165 pts (59.1%) 278 pts (60.3%) NS
Female 114 pts (40.9%) 183 pts (39.7%) NS
Smoking 169 pts (95.7%) 257 pts (92.4%) NS
Hypertension 267 pts (61.3%) 426 pts (64.4%) NS
Dyslipidemia 265 pts (95.0%) 416 pts (90.2%) =0,021
Concomitant coronary
89 pts (31.9%) 89 pts (19.3%) <0.001
disease
 8

Concomitant peripheral
70 pts (25.1%) 56 pts (12.1%) <0.001
disease
Previous myocardial
38 pts (13.6%) 30 pts (6.5%) =0.001
infarction
Previous TIA 10 pts (3.6%) 4 pts (0.9%) NS
Previous CVI 9 pts (3.2%) 6 pts (1.3%) NS
Positive CT of endocranium 12 pts (4.66%) 10 pts (2.17%) NS
The main characteristics are shown in Table 1. Group A (37.7 %) consisted of 279 diabetics,
and Group B (62,3%) consisted of 461 non-diabetic patients. Except for slightly greater
prevalence of dyslipidemia in patients with diabetes (Hi-square=5,330; p=0,021) - patients
with DM have more frequent coronary disease (Hi-square=15,090; p<0,001), and more
persistent peripheral arterial disease (Hi-square=20,607; p<0,001), other pre-operative
characteristics for the two groups are similar and comparable.

Table 2 Post-operative complications in diabetics and non-diabetics after CEA

Group A Group B
Characteristics P Value
diabetics non-diabetics
N 279 461 ---
Post-operative TIA 10 pts (3,6%) 4 pts (0,9%) =0.009
Post-operative cerebro-vascular
9 pts (3,2%) 6 pts (1,3%) =0.072
infarction – CVI
CT ischaemic brain lesion 11 pts (3,94) 8 pts (1,74%) =0.424
Cranial nerves lesion 2 pts (0,7%) 4 pts (0,9%) NS
Myocardial infarction (IM) 2 pts (0,7%) 1 pts (0,2%) =0,300
Congestive heart failure 8 pts (2,9%) 4 pts (0,9%) =0.039
Post-operative respiratory complications 8 pts (2,9%) 3 pts (0,7%) =0,024
Haematoma of operated wound 10 pts (3,9%) 11 pts (2,4%) =0.341
Infection of operated wound 5 pts (1,8%) 0 pts (0,0%) =0.007
Post-operative 50% restenosis 5 pts (1,8%) 10 pts (2,2%) =0.724

Post-operative neurological events (TIA) among patients with diabetes (3,6%) and among
non-diabetics (0,9%, p=0,009). Post-operative CVI occurred in 1,3% of patients without DM
and in 3,2% of patients with DM (Hi-square=3,241; p=0,072). Cumulative neurological
events TIA/cerebro-vascular infarction were also statistically more numerous in the diabetics'
group (P = 0,02). Adverse post-operative cardiac events (myocardial infarction, congestive
heart failure) occurred in 3,6% of patients with diabetes and in 1,1% of non-diabetic patients
(Fisher's test of accurate probability; p= 0,039). Haematoma of operated wound occurred in
11 patients (2,4%) without DM and in 10 patients (3,6%) with DM, statistically considerably
more frequent (Hi-square=0,905; p=0,341). Infection of operated wound in our study was
present in 1,8% of patients with DM, while none of the patients without DM had wound
infection - statistically important (Fisher's test of accurate probability; p=0,007). We observe
 9

that 0,9% of non-DM patients and 0,7% of DM patients have had symptoms of cranial nerves
lesion - statistically insignificant (Fisher's test of accurate probability; p=1,000). Post-
operative restenosis occurred in 2,2% of patients without DM and in 1,8% of the ones with
DM.

Table 3 Total mortality and morbidity in patients with diabetes (group A) and without
diabetes (group B) who have undergone CEA

Group A Group B non- P Value

Structure of patients Total
diabetics diabetics
N 279 461 740 --
Fatal cardiac event 4 (1,4%) 3 (0,7%) 7 (0,9%) =0.435
Fatal neurological event 3 (1,1%) 1 (0,2%) 4 (0,5%) =0.153
Total mortality 7 (2,5%) 4 (0,9%) 11 (1,5%) =0.113
Total morbidity 51 (18,3%) 39 (8,5%) 90 (12,2%) <0.01

Total rate of complications has been within the recommended limits; as for the rate of
mortality and total morbidity, the two groups have differed from one another considerably.
Operative and post-operative mortality (neurological and cardiological) has been 2,5% in the
diabetics' group (4 cardiac events + 3 cerebro-vascular infarctions) and 0,9% in the non-
diabetics' group (3 myocardial infarctions + 1 cerebro-vascular infarction, value p=0,113).
The patients with DM have statistically significantly higher total mortality.
Total post-operative complications are observed in 8,5% of patients without DM and in 18,3%
of patients with DM (Hi-square=15,688; p<0,001). Post-operative complications occur in DM
patients considerably more frequently (two and more times).

Table 4 Distribution of patients depending upon DM therapy, in relation to post-operative

complications

POST-OPERATIVE Oral antidiabetics Insulin Total

COMPLICATIONS N % N % N %
Not present 135 74,6 90 91,8 225 80,6
 10

Present 46 25,4 8 8,2 54 19,4

Total 181 100,0 98 100,0 279 100,0

181 examined DM patients (64,9%) have been using oral antidiabetics, and 98 of them
(35,1%) have been using insulin. Total post-operative complications have occurred in 25,4%
of patients on oral antidiabetics, and in 8,2% of patients on insulin (Hi-square=12,122;
p<0,001). Post-operative complications occur in patients on oral antidiabetics considerably
more frequently.

Table 5 Duration of DM in examined patients

DM duration (years)
N sd Med min Max

Without post-operative
complications
225 9,3 4,1 9,0 1,0 22,0

With post-operative
complications
54 12,5 4,8 11,5 3,0 22,0

Total
279 9,9 4,4 9,0 1,9 22,0

The average duration of DM in patients without post-operative complications has been

9,3±4,1 years, whereas the average duration of DM in patients with post-operative
complications has been 12,5±4,8 years (t=4,967; p<0,001). The patients with post-operative
complications have suffered from DM statistically considerably longer.
The stratification of patients with diabetes according to their levels of HbA1c has shown that
the group with HbA1c> 7,6 % has had higher total morbidity and mortality (Table 6).

Table 6 HbA1c in examined patients

N sd Med min Max

HbA1c
Without post-
operative
225 7,6 1,2 8,0 6,0 9,8
complications
 11

With post-operative
complications 54 8,4 0,9 8,5 6,0 9,5

Total
279 7,8 1,2 8,0 6,0 9,8

The median value of HbA1c in patients without post-operative complications has been
7,6±1,2, whereas the median value of HbA1c in patients with post-operative complications
has been 8,4±0,9 (t=5,010; p<0,001). The patients with post-operative complications have
significantly higher values of HbA1c.
The model of multiple logistic regression, with post-operative complications as a dependent
variable, has been supplemented with those predictors of post-operative complications which
were statistically important in the model of simple logistic regression, the importance level
being 0,05.
Statistically important predictors of early post-operative complications in the model of
multiple logistic regression have been: Age (B=0,069; <0,001), whose odds ratio is OR=1,07,
and it demonstrates that with increase in age by one year there is a 7% greater risk for patients
to develop early post-operative complications. DM (B=0,854; p=0,001), whose odds ratio is
OR=2,35, showing that DM patients are, with all other factors in the model controlled, at 2,35
times greater risk of development of early post-operative complications. Concomitant
coronary disease (B=0,844; p=0,001), whose odds ratio is OR=2,33 - patients with
concomitant coronary disease are at 2,33 times greater risk of development of post-operative
complications, with all other factors in the model controlled. At the level of logistic
regression, the factors identified to increase odds of death and post-operative complications
have been: hyperlipoproteinemia (p=0,021), more persistent coronary disease (p=0,001), and
a higher frequency of peripheral arterial disease (p<0,001), which is shown in Table 1. The
factors determined to increase odds of death and total morbidity have been: higher levels of
HbA1c (P < 0,001), application of peroral antidiabetics for controlling glucose in blood (P <
0,001), and patients who have suffered longer from DM (P < 0,01), which is shown in Tables
4, 5, and 6.