2006 by the Societe Internationale de Chirurgie

Published Online: 16 June 2006

World J Surg (2006) 30: 15921604

DOI: 10.1007/s00268-005-0657-x

Postsurgical Infections are Reduced with

Specialized Nutrition Support
Dan L. Waitzberg, MD,1 Hideaki Saito, MD,2 Lindsay D. Plank, PhD,3
Glyn G. Jamieson, MD,4 Palepu Jagannath, MD,5 Tsann-Long Hwang, MD,6
Juan M. Mijares, MD,7 David Bihari, MD8
1

~ Paulo Medical School, Sao Paulo, Brazil

Department of Gastroenterology, LIM 35, University of Sao
Departments of Surgery, University of Tokyo, Tokyo, Japan
3
University of Auckland, Auckland, New Zealand
4
University of Adelaide, Adelaide, Australia
5
Lilavati Hospital, Mumbai, India
6
Chang Gung University, Taipei, Taiwan
7
University of Anahuac, Huixquilucan, Mexico City, Mexico
8
Department of Intensive Care, Prince of Wales Hospital, Sydney, NSW, Australia
2

Abstract
Objective: The objective was to examine the relationship between pre-, peri-, and postoperative
specialized nutritional support with immune-modulating nutrients and postoperative morbidity in
patients undergoing elective surgery.
Methods: Studies were identified by searching MEDLINE, review article bibliographies, and abstracts and proceedings of scientific meetings. All randomized clinical trials in which patients were
supplemented by the IMPACT formula before and/or after elective surgery and the clinical outcomes reported were included in the meta-analysis. Seventeen studies (n = 2,305), 14 published
(n = 2,102), and 3 unpublished (n = 203), fulfilled the inclusion criteria. Ten studies (n = 1,392)
examined the efficacy of pre- or perioperative IMPACT supplementation in patients undergoing
elective surgery, whereas 7 (n = 913) assessed postoperative efficacy. Fourteen of the studies
(n = 2,083) involved gastrointestinal (GI) surgical patients. Postoperative complications, mortality,
and length of stay in hospital (LOS) were major outcomes of interest.
Results: IMPACT supplementation, in general, was associated with significant (39%61%)
reductions in postoperative infectious complications and a significant decrease in LOS in hospital
by an average of 2 days. The greatest improvement in postoperative outcomes was observed in
patients receiving specialized nutrition support as part of their preoperative treatment. In GI surgical patients, anastomotic leaks were 46% less prevalent when IMPACT supplementation was
part of the preoperative treatment.
Conclusion: This study identifies a dosage (0.51 l/day) and duration (supplementation for 5
7 days before surgery) of IMPACT that contributes to improved outcomes of morbidity in elective
surgery patients, particularly those undergoing GI surgical procedures. The cost effectiveness of
such practice is supported by recent health economic analysis. Findings suggest preoperative
IMPACT use for the prophylaxis of postoperative complications in elective surgical patients.

Correspondence to: Dan L. Waitzberg, MD, Department of Gastroenterology, LIM 35, University of S~
ao Paulo Medical School, Sao Paulo,
Brazil, e-mail: [email protected]

mmune suppression is a direct consequence of invasive procedures and as a result, surgical patients are

Waitzberg et al.: Nutrition for Elective Surgery

at a high risk of infection. For example, 55% of patients

will acquire infection following major cardiovascular surgery.1 Postsurgical complications including infections
burden the health care system, slowing recovery,
extending hospital stays, and increasing hospital expenses. A recent focus in medical practice is the proactive management of infection. The approach involves
both pre- and postoperative treatments and has been
shown to effectively reduce complications and decrease
the overall cost of care. The provision of specialized
nutrition support is one component of the proactive approach to infection control.
The specialized nutritional products currently available
contain a blend of nutrients with immune-modulating potential including the amino acids arginine and glutamine,
omega-3 fatty acids, and nucleotides/RNA. As four systematic reviews recently identified,25 specialized nutritional support is of particular benefit to hospitalized
patients undergoing surgical procedures. Hospital-acquired infections and length of stay (LOS) in both the ICU
and hospital were significantly decreased with the provision of specialized nutrition support. However, the former
systematic reviews group various types of enteral products together despite considerable differences in the
quality and quantity of immune-modulating nutrients.6
Considerable evidence supports that IMPACT specialized nutrition support (Novartis Consumer Health,
Nyon, Switzerland) positively influences inflammatory,
metabolic, and immune responses to major surgery.712
Of the various nutrients suggested to offer immunemodulating benefits, IMPACT contains a consistent mixture of arginine, omega-3 fatty acids from fish oil, and
nucleotides in the form of RNA. Nucleotides are important
during immunological challenge to support the development and activation of specialized immune cells.13 Arginine, the precursor of nitric oxide, is an amino acid that
becomes conditionally essential during trauma and sepsis. Following surgery, arginine improves wound healing
and protects against infection and ischemiareperfusion
tissue injury by restoring macrophage function and lymphocyte responsiveness.1417 Omega-3 fatty acids from
fish oil alter the phospholipid content of the cell membrane and shift the balance of leukotriene and prostaglandin production to less inflammatory and less
immunosuppressive mediators.1720 Furthermore, recent
attention has been focused on the synergistic interaction
of omega-3 fatty acids from fish oil and arginine to modulate immune function after surgery.21
The aim of this review was to compare the effect of
IMPACT specialized nutritional support initiated before
and/or after elective surgery with that of standard post-

1593

operative nutritional support on postoperative morbidity

outcomes. Based on recent evidence the hypothesis was
formulated that specialized nutritional support may be
considered a form of prophylaxis against postoperative
morbidity. The present meta-analysis eliminates formula
heterogeneity by exclusively analyzing studies using IMPACT specialized nutritional support. Furthermore, to
improve upon previous systematic reviews, this analysis
evaluates clinical outcomes associated with the initiation
and duration of specialized nutrition support; and it
identifies the various types of infectious comorbidities for
surgical patients and gastrointestinal (GI) surgical patients in detail.

MATERIALS AND METHODS

Study Identification
An electronic search for relevant articles was performed using the MEDLINE database. The keywords
enteral nutrition, surgical patients, arginine, fish
oil, omega-3 fatty acids, nucleotides, glutamine,
immunonutri*, preoperative, perioperative, and
postoperative were targeted. The search was limited to
human clinical trials published between January 1985
and December 2003. In addition, reference lists of reviews and original articles were searched manually, and
investigators identified by the manufacturer of IMPACT
were contacted for unpublished data.

Study Selection Criteria

The following criteria were used to select studies for
this review:
1. Study type: randomized clinical trial
2. Type of participants: surgical patients undergoing
major elective operations
3. Type of intervention: enteral nutrition and/or oral supplementation with IMPACT before and/or after surgery
4. Type of outcome measures: defined postoperative
infectious complications, mortality, length of hospital
stay, and cost of in-hospital care
5. Publication languages: English, German, French,
Spanish, Portuguese, Japanese, and Chinese
To comply with the gold standard for systematic reviews,22 our analysis included all studies published in
English and non-English literature as well as unpublished
trials. Furthermore, to select studies with the highest
validity in terms of a relative treatment effect, we included

1594

Waitzberg et al.: Nutrition for Elective Surgery

only randomized clinical trials.4,23 Studies reporting only

nutritional or immunological outcomes were excluded,
and clinical outcome was considered the primary outcome measure. Authors of multiple studies were contacted to obtain supplemental information not included in
published articles and to avoid the use of duplicate data.
A methodological review was conducted by two
investigators to assess study quality as defined by the
scoring system described by Jadad.24

postsurgical (postoperative) period, or both (perioperative). Strength was assessed by comparing our findings
with those achieved after analyses were confined to
published studies or those associated with GI surgery.

Primary Statistical Analysis

Search methods identified 58 citations and 4 unpublished randomized clinical trials that supplemented with
IMPACT specialized nutritional support. Abstracts were
analyzed with regard to the defined selection criteria.
Studies found to meet inclusion criteria and those lacking
sufficient data in the abstract were reevaluated using the
full-text publication. Protocols or draft manuscripts were
analyzed when data from trials were unpublished. Following full-text evaluation, 18 studies in total, 14 published 1,7,283139 and 4 unpublished (Berne, Amsterdam,
Sydney, and Open), met all the inclusion criteria. The test
diet of all 18 studies had equal proportions of the immune
modulating nutrients arginine, omega-3 fatty acids from
fish oil, and nucleotides.

Primary outcomes of interest were the number of patients with one or more postoperative-acquired infection(s), the LOS in hospital, and hospital mortality.
Individual studies defined and described the occurrence of
infectious complications including wound infection, intraabdominal abscess, pneumonia, urinary tract infection
(UTI), and sepsis. Infection rates and the frequently
encountered noninfectious surgical complication, anastomotic leak, were evaluated as secondary outcomes.
Statistical analyses utilized Comprehensive META
ANALYSIS software, version 2.2.021, June 20, 2005
(Biostat, Englewood, NJ, USA [www.Meta-Analysis.
com]). Intent-to-treat analysis used summary data from
published and unpublished studies. The most inclusive
category was used when individual study results were not
reported as intent-to-treat. Infection was treated as a
binary variable. Results of the meta-analysis are expressed in terms of relative risk (RR) for the treatment
group vs. the control group, such that an RR < 1 favors
the treatment group and an RR > 1 favors the control
group. Combined data from all 17 studies were used to
estimate an overall RR and associated 95% confidence
interval (CI). The Mantel-Haenszel method was used to
test the significance of treatment effect25 and a random
effects model estimated the overall RR.26 For LOS
analysis, effect size (ES) was used to describe the
standardized difference between means from treatment
and control. Hedges method was used to estimate the
individual ES and pooled ES between two treatments.27
The pooled simple difference between 2 group means is
reported to provide the estimate of treatment effect in
days. Only statistically significant ESs are analyzed in this
manner. P < 0.05 was considered statistically significant.

Subgroup and Sensitivity Analysis

Subgroup analyses were carried out for IMPACT supplementation during the presurgical (preoperative) or

RESULTS
Trials Identified

Eighteen out of 58 Published Studies Met

Inclusion Criteria
Published studies were excluded for the following
reasons: clinical outcome was not an endpoint in 15,
the study was not randomized in 1, only a single immune modulating nutrient was evaluated in 5, the
specialized nutritional formula was not IMPACT in 1,
and surgical patients were not analyzed individually
within the study population in 16. In addition, patient
populations from 6 publications were compiled and included in subsequent publications to justify their
exclusion from analyses. Of the 4 unpublished studies
identified, the principal investigators provided adequate
data for analysis of 3, including clinical outcome results
(for each), study protocols (for 2), and a draft manuscript (for 1). The fourth study, although complete, was
excluded because the data bank was not yet closed
due to unresolved inconsistencies (Open). Table 1 details the characteristics of the 17 trials, comprising
2,305 patients, included in the analyses. Studies were
generally of acceptable quality, 10 with a methodological score of 3 or above were considered moderate to
high quality and 7 with a score of less than 3 were
rated low quality.24

Waitzberg et al.: Nutrition for Elective Surgery

Descriptions of the Unpublished Trials Used in

Analyses Follow
A prospective trial carried out in Switzerland (Berne)
included 29 patients, split into 2 groups, who underwent
major elective surgery for GI and/or pancreatic cancer. As
detailed in Table 1, preoperative supplementation of IMPACT was the experimental variable. Both groups received IMPACT postoperatively. A separate trial of doubleblind design was carried out in the Netherlands (Amsterdam) in a group of cardiac patients identified as high risk
and undergoing cardio-pulmonary bypass according to a
previously described protocol.1 Patients, 48 in total, were
randomized into 2 groups, differing in supplementation of
IMPACT nutrition support prior to surgery (Table 1). Patients received no enteral feeding following surgery. The
third study of double-blind design originated in Australia
(Sydney), and followed 126 patients undergoing major
surgery for colorectal cancer. Patients randomly received
either perioperative IMPACT or an isocaloric, isonitrogenous diet, and outcomes were evaluated (Table 1).

Surgical and Nutritional Interventions

All trials included patients undergoing major elective
surgery. Fourteen studies described patients with GI
malignancy; patients underwent upper GI surgery in 8
studies, lower GI surgery in 2, and upper or lower GI
surgery in 4. Two studies included patients undergoing
bypass and valve replacement surgery, and 1 included
patients undergoing surgery for head or neck malignancy.
Ten studies, 7 published1,2830,3739 and 3 unpublished
(Amsterdam, Sydney, and Berne), supplemented with
IMPACT before surgery in efficacy trials of preoperative or
perioperative specialized nutrition support. A variety of
study designs were used to describe the efficacy of preoperative IMPACT supplementation. In 3 studies1,28
(Amsterdam), treatment was exclusively preoperative and
IMPACT was compared with a standard formula that was
isonitrogenous and isocaloric. A set of studies provided
consistent postoperative support with IMPACT (Berne), a
standard formula,30 or intravenous (i.v.) electrolytes (5%
glucose),29 and compared preoperative supplementation
with IMPACT versus no nutrition support.
A set of studies investigated the efficacy of perioperative IMPACT specialized nutritional support. Patients in
the treatment group were supplemented with IMPACT
both pre- and postoperatively. For comparison, patients in
the control group received either an isocaloric, isonitrogenous formula30,37,39 (Sydney), a standard formula,38 or
no nutritional support,28 preoperatively. The control group

1595

from another study28 received no nutritional support during both the pre- and postoperative treatment periods.
Seven studies investigated the immune-modulating
effect of postoperative IMPACT supplementation,7,3136
and 5 initiated patient feedings within 24 hours of surgery.7,31,3335 Among the 7 postoperative studies, 1
compared IMPACT specialized nutrition with both a
standard enteral formula and a low calorie, low fat i.v.
solution.32 Another evaluated specialized nutritional
support against isocaloric, isonitrogenous enteral and
parenteral support.35 In summary, 6 of the postoperative
efficacy trials compared IMPACT with a control enteral
feed7,3132,34,36 and 3 evaluated IMPACT against an i.v.
solution or feed.32,34,35

Effect of IMPACT on Cost, Infectious Complication Rates, LOS in Hospital, and Mortality
Cost data were available from only two studies8,37 and
were not analyzed further. Sixteen studies (out of 50)
reported postoperative infectious complications as shown
in Tables 24. The aggregated results of these studies
demonstrate that the use of IMPACT was associated with
significantly fewer postoperative infectious complications
(RR 0.49, 95% CI 0.420.58, P < 0.0001) and the test for
heterogeneity was not significant (P = 0.95; results not
shown).
Sixteen of the 17 studies reported LOS in hospital (Tables 24). Mean and SD data were available for all but one
study.25 Overall, patients receiving IMPACT had a significantly shorter LOS in hospital (ES ())0.66, 95% CI
())0.86())0.45, P < 0.0001) and the test for heterogeneity
was significant (P < 0.0001; results not shown). The pooled
difference between the group means measured a reduction in LOS in hospital of 3.1 days (95% CI ())3.9())2.3
days) with IMPACT supplementation (results not shown).
Mortality was low in the patient population. Analysis of
aggregate data did not detect an improvement in mortality
rates with IMPACT supplementation (RR 0.72, 95% CI
0.391.31, P = 0.28) and the test for heterogeneity was
not significant (P = 0.98; results not shown).

Effect of IMPACT on Postoperative Morbidity

Meta-analysis of postoperative morbidity reported in
these 17 trials indicated that IMPACT supplementation
was associated with significantly fewer outcomes of
morbidity (Table 5), including wound infections (RR 0.63,
95% CI 0.460.87, P = 0.005), abdominal abscesses (RR
0.46, 95% CI 0.290.72, P = 0.001), pneumonia (RR
0.53, 95% CI 0.390.71, P < 0.0001), UTIs (RR 0.53, 95%

Study year, country

19982000, Italy

, USA
19881990, USA

No

2
2

No
No

Yes
No
No
Yes
No

4
1
2
5
2

No
No

3
3
No
Yes
Yes
Yes

Yes

3
4
5
3

No
Yes

3
5

No

No

Yes

Blinding

Jadad
scale

ITT
ITT

PP
ITT
ITT
PP
PP

ITT
ITT
ITT
PP

ITT
ITT

PP

ITT
PP

ITT

ITT

ITT

PP

Study
analysis

60
85

118
260
195
154
41

100
206
129
154

100
203

126

100
45

204

100

29

48

Patient
number

UGS
UGS

UGS
UGS
UGS
UGS
UGS+LGS

LGS
UGS+LGS
H+NCS
UGS

UGS+LGS
UGS+LGS

LGS

LGS
cardiac

UGS+LGS

UGS+LGS

UGS

Cardiac

Patient
group

IN, IC
IN, IC + parenteral
i.v. crystalloids
IN, IC
Standard formula +
i.v. solution
Standard formula
Standard formula

IN, IC
Preop: nil Postop:
i.v. solution
Nil
IN, IC
Standard formula
IN, IC

IN, IC

Preop: nil Postop:

impact
Preop: nil Postop:
IN, IC
Preop: nil Postop:
i.v. solution
IN, IC
IN, IC

IN, IC

Control
feed(s)

days/6 hours
days/6 hours
days/
days/12 hours

/first postop day

/first postop day

/third postop day

/6 hours
/within 24 hours
/12 hours
/as early as possible

5
7
5
5

7 days/within 12 hours
5 days/within 12 hours

5 days/6 hours

5 days/
5 days/

5 days/

7 days/

5 days/6 hours

5 days/

Initiation of feed
before/after surgery

UGS: upper GI surgery; LGS: lower GI surgery; H+NCS: head and neck cancer surgery; PP: Per protocol; ITT: intent to treat; IN: isonitrogenous; IC: isocaloric; : no
standard protocol/not reported.

31

39

, Italy
, Italy
38
19941996, USA
37
19941997, Germany
Postoperative IMPACT studies
36
, China
35
19931997, Italy
34
19941996, USA
33
19921994, Germany
32
, Switzerland

30

, Italy
19961997, The Netherlands
Perioperative IMPACT studies
Sydney study
19992001, Australia
(unpublished)
29
19982000, Italy
28
19982000, Italy

30

28

Preoperative IMPACT studies

Amsterdam study
19982000, The Netherlands
(unpublished)
Berne study
19981999, Switzerland
(unpublished)
29
19982000, Italy

Reference

Table 1.
Randomized studies evaluating IMPACT as a pre-, peri-, or postoperative treatment in elective surgery
1596
Waitzberg et al.: Nutrition for Elective Surgery

Waitzberg et al.: Nutrition for Elective Surgery

1597

Table 2.
Randomized studies in elective surgery patients evaluating the effect of preoperative application of IMPACT on mortality, rate of
infectious complications, and length of stay in hospital (LOS)
Mortality, number/
total (%)
Study
Amsterdam study (unpublished)
Berne study (unpublished)b
29
c
28
a
30a
1

Preoperative*
a

0/23
1/14
1/50
1/102
0/50
1/23

(0)
(7)
(2)
(1)
(0)
(4)

Patients with infectious

complications, number/total (%)

Control
1/24
0/15
2/50
1/102
1/50
1/22

(4)
(0)
(4)
(1)
(0)
(5)

Preoperative**
4/23
2/14
8/50
14/102
6/50
4/23

(17)
(14)
(16)
(13)
(12)
(17)

Control
12/24
10/15
12/50
31/102
15/50
12/22

(50)
(67)
(24)
(30)
(30)
(55)

Hospital days,
mean (SD)
Preoperative***
7.6
19.7
13.2
11.6
9.5
9.6

(3.0)
(2.3)
(3.5)
(4.7)
(2.9)
(5.3)d

Control
9.0
29.1
15.3
14.0
12.2
11.7

(4.5)
(3.6)
(4.1)
(7.7)
(3.9)
(12.0)d

*RR 0.74, 95% CI 0.232.35, P =0.609 vs. control.

**RR 0.42, 95% CI 0.230.59, P < 0.0001 versus control.
***Effect size (ES) ())0.71, 95% CI ())1.14())0.28, P =0.001 vs. control, pooled difference ())3.4 days, 95% CI ())5.6
())1.2 days.
Postoperative feed: aNo artificial feeding; bIMPACT; cStandard formula; dMeans SD provided by main investigator.
Table 3.
Randomized studies in elective surgery patients evaluating the effect of perioperative application of IMPACT on mortality, rate of
infectious complications, and LOS in hospital
Mortality, number/
total (%)
Study

Perioperative

Sydney study (unpublished)

29
28
30
39
38
37

0/61
0/50
2/101
1/50
0/102
0/82
0/78

(0)
(2)
(2)
(0)
(0)
(0)

Patients with infectious

complications, number/total (%)

Control
3/65
2/50
1/102
0/50
1/104
0/47
0/76

(5)
(4)
(1)
(0)
(1)
(0)
(0)

Perioperative*
15/61
5/50
16/101
5/50
14/102
19/82
10/78

(25)
(10)
(16)
(10)
(14)
(25)
(13)

Control
32/65
12/50
31/102
16/50
31/104
19/47
18/76

(49)
(24)
(30)
(32)
(30)
(41)
(24)

Hospital days,
mean (SD)
Perioperative**
11.4
12.0
12.2
9.8
11.1
15.3
22.2

(4.0)
(3.8)
(4.1)
(3.1)
(4.4)
(9.1)
(4.1)

Control
12.7
15.3
14.0
12.0
12.9
17.4
25.8

(5.8)
(4.1)
(7.7)
(4.5)
(4.6)
(11.9)
(3.8)

*RR 0.49, 95% CI 0.390.62, P < 0.0001 vs. control.

**ES ())0.48, 95% CI ())0.68())0.28, P < 0.0001 vs. control, pooled difference ())2.4 days, 95% CI ())3.1())1.7 days.

CI 0.320.87, P = 0.011), and anastomotic leaks (RR

0.56, 95% CI 0.370.83, P = 0.004), whereas the number
of septic episodes was also reduced by more than 40%,
but not statistically significantly (RR 0.56, 95% CI 0.30
1.03, P = 0.060) (Table 6). The test for heterogeneity for
all parameters was not significant (P = 0.810.99).

Subgroup Analysis
The effect of the pre-, peri-, and postoperative application of IMPACT was examined. Compared with their
control counterparts, the rate of infectious complications
was significantly lower in patients receiving IMPACT
preoperatively (RR 0.42, 95% CI 0.300.59, P < 0.0001)
and perioperatively (RR 0.49, 95% CI 0.390.62, P <
0.0001), as well as with postoperative supplementation
alone (IMPACT vs. enteral or i.v. control) (RR 0.55, 95%

CI 0.410.75, P = 0.0001; Tables 24). Separate evaluation of the postoperative use of IMPACT against a
standard enteral formula or i.v. solution (Table 4) showed
that infectious complications were significantly reduced
with IMPACT supplementation in both groups (RR 0.56,
95% CI 0.390.82, P = 0.003 in standard enteral controls;
RR 0.53, 95% CI 0.310.90, P = 0.02 in i.v. controls).
Grouping studies by their initiation of IMPACT treatment identified that preoperative supplementation (ES
())0.71, 95% CI ())1.14())0.28, P = 0.001) and perioperative supplementation (ES ())0.48, 95% CI ())0.68
())0.28, P < 0.0001) were associated with significantly
shorter LOS in hospital (Tables 2, 3). The pooled difference between group means was ())3.4 days, 95% CI
())5.6())1.2 days preoperatively and ())2.4 days, 95%
CI ())3.1())1.7 days perioperatively (Tables 2, 3). In
comparison, postoperative supplementation of IMPACT

1598

Waitzberg et al.: Nutrition for Elective Surgery

Table 4.
Randomized studies in elective surgery patients evaluating the effect of the postoperative application of IMPACT on mortality, rate
of infectious complications, and LOS in hospital
Mortality, number/
total (%)
Study
36
35
33
32
31
7
35
34
32

Postoperative
0/60 (0)
1/87 (1.1)
3/77 (3.9)
0/14 (0)
1/30 (3.3)
2/41 (4.9)
1/87 (1.1)
2/97 (2.1)
0/14 (0)

Control
0/58 (0)
2/87 (2.3)
2/77 (2.6)
0/14 (0)
2/30 (6.6)
0/44 (0)
2/86 (2.3) (i.v.)
3/98 (3.1) (i.v.)
0/14 (0) (i.v.)

Patients with infectious

complications, number/total (%)
Postoperative
,

Control

0/60 (0)* **
13/87 (15)*,**
14/77 (18)*,**
3/14 (21)*,**
1/30 (3)*,**
5/41 (12)*,**
13/87 (15)*,***
NR
3/14 (21)*,***

Hospital days,
mean (SD)
Postoperative

2/58 (3)
20/87 (23)
19/77 (25)
6/14 (43)
11/30 (37)
13/44 (30)
24/86 (28) (i.v.)
NR
6/14 (43) (i.v.)

#,

13.1 (2.5)
16.1 (6.2)#,
27.0 (2.3)#,
14.5 (8)#,
16 (0.9)#,
18.8 (11.1)#,
16.1 (6.2)#,
11 (441)#,,a
14.5 (8)#,

Control
14.5 (3.0)
19.2 (7.9)
30.6 (3.1)
14.0 (19)
22 (2.9)
20.4 (9.6)
21.6 (8.9) (i.v.)
10 (675)a (i.v.)
14.0 (10.3) (i.v.)

NR: Not reported.

*RR 0.55, 95% CI 0.410.75, P = 0.0001 vs. enteral + i.v. control.
**RR 0.56, 95% CI 0.390.82, P = 0.003 vs. enteral control.
***RR 0.53, 95% CI 0.310.91, P = 0.02 vs. i.v. control.
#ES ())0.73, 95% CI ())1.18())0.29, P = 0.001 vs. enteral + i.v. control, pooled difference ())3.5 days, 95% CI ())5.1
())1.9 days.
ES ())0.84, 95% CI ())1.43())0.24, P = 0.006 vs. enteral control.
ES ())0.47, 95% CI ())1.10.16, P = 0.145 vs. i.v. control.
a
Median (range).

was also associated with a significant decrease in LOS in

hospital (ES ())0.73, 95% CI ())1.18())0.29,
P = 0.001), a pooled difference of ())3.5 days, 95% CI
())5.1())1.9 days. Comparison of postoperative IMPACT supplementation with a standard enteral formula
showed that LOS in hospital again decreased significantly
(ES ())0.84, 95% CI ())1.43())0.25, P = 0.006),
whereas postoperative IMPACT supplementation vs. i.v.
control showed a decrease that did not reach statistical
significance (ES ())0.47, 95% CI ())1.10.16, P = 0.145;
Table 4).
There was no significant difference in mortality for the 3
subgroups, preoperative (RR 0.74, 95% CI 0.232.35,
P = 0.51), perioperative (RR 0.48, 95% CI 0.141.61,
P = 0.23), or postoperative (RR 0.87, 95% CI 0.362.10,
P = 0.75) IMPACT supplementation (Tables 24).
As shown in Table 7, when specific postoperative
complications were analyzed in GI surgical patients preoperative IMPACT supplementation resulted in fewer
abdominal abscesses (RR 0.44, 95% CI 0.161.19,
P = 0.107), fewer wound infections (RR 0.56, 95% CI
0.291.07, P = 0.077), significantly less pneumonia (RR
0.38, 95% CI 0.180.83, P = 0.015), fewer UTIs (RR
0.64, 95% CI 0.281.47, P = 0.142), fewer episodes of
sepsis (RR 0.29, 95% CI 0.051.71, P = 0.169), and
fewer anastomotic leaks (RR 0.51, 95% CI 0.241.05,
P = 0.069). Perioperative application of IMPACT was
associated with significantly fewer abdominal abscesses

(RR 0.43, 95% CI 0.210.91, P = 0.027) and wound

infections (RR 0.61, 95% CI 0.380.96, P = 0.033), less
pneumonia (RR 0.54, 95% CI 0.340.87, P = 0.011),
fewer UTIs (RR 0.53, 95% CI 0.231.19, P = 0.122) and
episodes of sepsis (RR 0.53, 95% CI 0.221.27,
P = 0.153), as well as significantly fewer anastomotic
leaks (RR 0.52, 95% CI 0.280.95, P = 0.034). The test
for heterogeneity for all parameters was not significant
(P = 0.580.99). Significantly fewer abdominal abscesses (RR 0.48, 95% CI 0.240.98, P = 0.044), but no
significant improvement in the other types of postoperative complications was found in the subset of GI patients
supplemented postoperatively (Table 7).

Sensitivity Analysis
The status of publication did not affect the overall outcome. When limiting analyses to the 14 published studies, IMPACT supplementation continued to significantly
reduce infectious complications (RR 0.50, 95% CI 0.42
0.60, P < 0.0001), and LOS in hospital (ES ())0.63, 95%
CI ())0.83())0.43, P < 0.0001), resulting in a pooled
difference between group means of ())3.0 days, 95% CI
())3.7())2.3 days (data not shown). Again, no significant effect was detected with regard to mortality (RR
0.75, 95% CI 0.401.42, P = 0.37) in the patient population (data not shown). When measures of postoperative
morbidity were analyzed in the published studies alone,

31

34

35

31

32

33

35

36

37

38

39

30

28

29

Preoperative

1
3
10
3
1
2
NA
3
Control Perioperative
4
5

NA

Control

Preoperative

UTI

6
2
8
4
4
2
9
2
Control Perioperative
9
0

10

Control

Pneumonia

NR

Preoperative

2
0
5
0
3
NR
1
NR
Control Perioperative
4
0

Control

Sepsis

NA

Preoperative

1
3
2
4
NR
3
NR
NA
Control Perioperative
4
1

NR

Control

5
10
5
NA
Control
0

NA

control

Anastomotic leak

2
4
0
1
3
6
1
2
0
1
3
5
7
11
4
10
6
8
2
5
1
2
5
10
2
4
0
1
3
5
1
5
NR
NR
3
6
4
6
2
4
4
10
3
3
2
5
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0
1
0
1
4
9
2
1
4
4
3
8
Postoperative Control Postoperative Control Postoperative Control Postoperative Control Postoperative Control Postoperative Control
NR
NR
0
2
NR
NR
NR
NR
NR
NR
NR
NR
4
5
4
8
2
3
1
1
NR
NR
NR
NR
1
2
0
2
9
8
1
6
2
4
7
10
1
2
NR
NR
1
0
0
1
1
3
NR
NR
0
1
0
2
0
6
NR
NR
0
2
0
1
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
Postoperative Control Postoperative Control Postoperative Control Postoperative Control Postoperative Control Postoperative Control
(i.v.)
(i.v.)
(i.v.)
(i.v.)
(i.v.)
(i.v.)
4
6
4
8
3
5
1
2
NR
NR
NR
NR
5
3
2
1
3
7
1
3
1
1
3
4
1
1
NR
NR
1
1
0
0
4
1
NR
NR

4
1
11
4
5
0
2
NA
Control Perioperative
17
2

NA

Preoperative

Abdominal abscess

NA: not applicable; NR: not reported or no events occurred; UTI urinary tract infection.

Sydney
(unpublished)

30

28

2
7
3
0
Perioperative
10

29

Amsterdam
(unpublished)
Bern
(unpublished)

Control

Preoperative

Study

Wound infection

Type of infection

Table 5.
Number of patients with complications reported in 17 randomized surgical trials using IMPACT

Waitzberg et al.: Nutrition for Elective Surgery

1599

Infectious complications
Abdominal
0.40
abscess
(0.220.73)
Wound
0.58
infection (0.370.90)a
Pneumonia
0.54
(0.380.76)b
UTI
0.61
(0.351.08)a
Sepsis
0.53
(0.251.13)b
Noninfectious complication
Anastomotic
0.53
leak
(0.350.81)
0.99
0.99
0.91
0.74
0.96

0.99

0.003

0.015

0.0001

0.090

0.099

0.003

0.56
(0.271.19)

0.41
(0.151.08)
0.59
(0.331.04)
0.54
(0.350.82)
0.55
(0.201.54)
0.54
(0.231.26)
0.131

0.156

0.253

0.004

0.069

0.072

0.42

0.59

0.29

0.55

0.94

0.96

Reflects an 50% reduction in incidence with IMPACT supplementation vs. controls.

Reflects an 60% reduction in incidence with IMPACT supplementation vs. controls.

Studies evaluated on ITT

0.53
(0.330.86)

0.47
(0.270.81)b
0.65
(0.430.97)a
0.47
(0.310.72)
0.57
(0.321.01)
0.39
(0.160.97)

Reflects an 35%40% reduction in incidence with IMPACT supplementation vs. controls.

Blinded studies

Studies with isonitrogenous,

isocaloric control feed

0.009

0.042

0.055

0.001

0.036

0.006

0.99

0.99

0.94

0.90

0.98

0.99

0.61
(0.341.09)

0.43
(0.210.87)c
0.61
(0.380.97)
0.54
(0.370.79)b
0.61
(0.291.26)
0.53
(0.231.19)

0.096

0.122

0.180

0.002

0.036

0.020

0.78

0.74

0.60

0.85

0.99

0.99

Relative risk P value

P value
Relative risk P value
P value
Relative risk P value
P value
Relative risk P value
P value
(95% CI)
heterogeneity (95% CI)
heterogeneity (95% CI)
heterogeneity (95% CI)
heterogeneity

Studies with moderate to high quality

(Jadad score > 3)

Table 6.
Relative risk (95% CI) of infectious and noninfectious postoperative complications in 17 randomized surgical trials with IMPACT (sensitivity analysis with regard to
methodological quality)
1600
Waitzberg et al.: Nutrition for Elective Surgery

0.90
0.325
0.69 (0.321.46)
0.87
0.069

Reflects an 50% reduction in incidence with IMPACT supplementation vs. control.

Postoperative IMPACT vs. postoperative control enteral feed/i.v. control combined.
b

0.034
0.52 (0.280.95)a

0.81

0.79
0.94
0.56
0.89
0.39
0.044
0.516
0.290
0.065
0.521
(0.240.98)
(0.421.54)
(0.411.30)
(0.121.06)
(0.271.94)
0.48
0.81
0.73
0.38
0.73
0.99
0.99
0.97
0.58
0.74
0.027
0.033
0.011
0.122
0.153
(0.210.91)
(0.380.96)
(0.340.87)
(0.231.19)
(0.221.27)
0.43
0.61
0.54
0.53
0.53
0.82
0.99
0.59
0.69
0.96
0.107
0.077
0.015
0.293
0.169

Infectious complications
Abdominal abscess
0.44 (0.161.19)
Wound infection
0.56 (0.291.07)
Pneumonia
0.38 (0.180.83)
UTI
0.64 (0.281.47)
Sepsis
0.29 (0.051.71)
Noninfectious complication
Anastomotic leak
0.51 (0.241.05)a

P value
heterogeneity
P value
Relative risk
(95% CI)
P value
heterogeneity
P value
Relative risk
(95% CI)
P value
heterogeneity
P value
Relative risk
(95% CI)

Postoperative GI surgery studiesb

Perioperative GI surgery studies
Preoperative GI surgery studies

Table 7.
Relative risk (95% CI) of infectious and noninfectious postoperative complications in 14 randomized gastrointestinal (GI) surgical trials with IMPACT

Waitzberg et al.: Nutrition for Elective Surgery

1601

IMPACT supplementation remained associated with significantly fewer abdominal abscesses (RR 0.45, 95% CI
0.280.72, P = 0.001) and wound infections (RR 0.64,
95% CI 0.450.94, P = 0.021), less pneumonia (RR 0.55,
95% CI 0.390.77, P = 0.001), and fewer anastomotic
leaks (RR 0.54, 95% CI 0.360.81, P = 0.002), whereas
a reduction in UTIs (RR 0.62, 95% CI 0.371.05,
P = 0.076) was no longer statistically significant
(Table 6). The test for heterogeneity for all parameters
was not significant (P = 0.840.99).
When analyses were limited to studies of upper and
lower GI procedures and the results from 3 studies were
excluded (1,37, and Amsterdam), the outcomes of morbidity, LOS in hospital, and mortality did not change.
IMPACT groups had significantly reduced infectious
complications (RR 0.50, 95% CI 0.420.60, P < 0.0001)
and LOS in hospital (ES ())0.72, 95% CI ())0.94
())0.50, P < 0.0001), resulting in a pooled difference
between group means of ())3.2 days, 95% CI ())4.1
())2.4 days (data not shown). No significant effect on
mortality was determined in the patient population (RR
0.72, 95% CI 0.381.37, P = 0.32; data not shown).
Analyses (Table 6) identified comparable reductions in
the various outcomes of postoperative morbidity, and
complications were significantly avoided, with the
exception of septic episodes, in the IMPACT group
(P = 0.0010.013).
Applying above the sensitivity analyses, significant
reductions in infectious complications and LOS in hospital
were observed under pre- and perioperative IMPACT
supplementation (results not shown), whereas significance was lost in the postoperative group.
When limiting analyses to 10 studies of moderate to high
quality (Jadad score 3), IMPACT supplementation continued to significantly reduce infectious complications (RR
0.49, 95% CI 0.410.60, P < 0.0001), and LOS in hospital
(ES ())0.61, 95% CI ())0.79())0.43, P < 0.0001),
resulting in a pooled difference between group means of
())2.6 days, 95% CI ())3.1())2.1 days (data not shown).
Again, no significant effect was detected with regard to
mortality (RR 0.72, 95% CI 0.321.62, P = 0.37) in the
patient population (data not shown). When measures of
postoperative morbidity were analyzed in the moderate to
high quality studies alone, IMPACT supplementation remained associated with significantly fewer abdominal abscesses (RR 0.40, 95% CI 0.220.73, P = 0.003) and
wound infections (RR 0.58, 95% CI 0.370.90, P = 0.015),
less pneumonia (RR 0.54, 95% CI 0.380.76, P = 0.0001),
and fewer anastomotic leaks (RR 0.53, 95% CI 0.350.81,
P = 0.003; Table 6). Identical results were found when
limiting analyses to the 9 studies that were evaluated on an

1602

Waitzberg et al.: Nutrition for Elective Surgery

intent-to-treat basis. IMPACT supplementation showed

significantly reduced infectious complications (RR 0.48,
95% CI 0.390.59, P < 0.0001) and LOS in hospital (ES
())0.76, 95% CI ())1.05())0.47, P < 0.0001), resulting in a
pooled difference between group means of ())3.4 days,
95% CI ())4.6())2.2 days (data not shown) and significantly reduced abdominal abscesses, wound infections,
pneumonia, and anastomotic leaks (Table 6). Restricting
the analyses to 8 blinded studies yielded comparable results with the exception that the reduction in abdominal abscesses was no longer statistically significant
(Table 6). Limiting the analyses to the 10 studies that used
isonitrogenous, isocaloric control feed also yielded comparable results with the exception that the reduction in UTIs
and abdominal abscesses was no longer statistically significant (Table 6).

DISCUSSION
As a result of medical advances, surgical procedures
are generally less invasive. Paradoxically, infection rates
associated with hospital stays are increasing. Furthermore, with extended hospital stays, the risk of infection
increases. The surgical patient is at a particularly high risk
of infection when visceral organ beds are involved.
Immunity is compromised due to the stress of tissue
ischemia and reperfusion combined with blood hemorrhage and transfusion. Furthermore, the GI surgical patient is at risk of the common and costly postoperative
complication of elective surgery, the anastomotic leak.
The present meta-analysis describes the efficacy of
IMPACT specialized nutritional support in a patient population at a high risk of postoperative complications. One
consistent formulation was used in each of the 17 studies
evaluated, providing homogeneity across trials in the
quality and relative quantity of immune modulating nutrients administered. No adverse effects have been observed in the arginine, omega-3 fatty acid and nucleotide
(IMPACT specialized nutrition) supplemented patient
groups. The preoperative supplementation period lasted
57 days and delivered 0.51 l of IMPACT specialized nutritional support per day, on average. In 10 of the
17 trials, IMPACT was compared with an isocaloric,
isonitrogenous control product; therefore, the observed
differences in clinical outcome cannot be attributed to the
calorie or nitrogen content of the experimental formula. Our
analyses revealed a significant relative reduction in the risk
of all described postoperative infections with the supplementation of IMPACT. Sensitivity analyses demonstrated
that, under supplementation with IMPACT, the results

regarding the significant reduction in infectious complications per se, in LOS in hospital by more than 2 days, and in
abdominal abscesses, wound infections, and pneumonia
were very robust.
The analyses provide the first comparison of the pre-,
peri-, and postoperative application of IMPACT specialized nutritional support. Building and expanding upon
previous meta-analyses25, which demonstrated significant immune-related treatment effects, our meta-analysis
detected significant reductions in specific types of postoperative infections. Pneumonia and UTIs were reduced
significantly (50%) in all studies, whereas wound
infections were down 35% and 40% in all and GIspecific studies respectively. The length of hospital stay
after surgery was reduced in 15 of the 17 studies analyzed, contributing to an overall reduction in LOS in
hospital of more than 2 days. Such findings parallel a
recent report that describes the health-economic advantage of using specialized nutrition for 5 days prior to
surgery in GI patients.40
The beneficial effects of IMPACT on clinical outcomes
in elective surgery patients were associated with the initiation of supplementation. Significant reductions in
infectious complications and LOS in hospital were observed in all groups independent of whether IMPACT
specialized nutrition support was given pre-, peri-, or
postoperatively. Our analysis of postoperative outcomes
in surgical patients suggests that the ingredient formulation of IMPACT appears to provide therapeutic advantages beyond the standard nutritional support. Best
outcomes were observed when IMPACT was supplemented during the preoperative period, generally 0.5
1.0 l/day for 57 days before surgery. Supplementation
significantly lowered complication rates for wound infections, pneumonia, UTIs, and abdominal abscesses in
elective surgical patients, as well anastomotic leaks in GI
surgical patients. Since a similar number of studies were
evaluated for all three modes of IMPACT supplementation, our results accurately reflect the reduced efficacy of
postoperative specialized nutritional support alone, however, better than standard methods of support.
Of the 17 studies analyzed, 14 involved patients undergoing upper or lower GI surgery, offering a unique opportunity to characterize a clinically homogeneous patient
population. Here, we saw the most substantial benefit of
IMPACT supplementation, an approximately 60% reduction in the frequency of abdominal abscesses following GI
surgery (Table 6). The incidence of anastomotic leaks, a
major complication of GI surgery with considerable rates of
morbidity and mortality (up to 30% mortality), and cost41,42,
was significantly reduced (50%) with perioperative IM-

Waitzberg et al.: Nutrition for Elective Surgery

PACT supplementation (Table 7). Our analyses show that

preoperative IMPACT specialized nutritional support is of
substantial benefit, while early postoperative supplementation does not increase the risk of anastomotic leak as
confirmed by subgroup analysis (Table 7). Improved oxygenation and perfusion of gut vasculature with IMPACT
supplementation11,28 may be responsible for this clinical
benefit. In fact, a significant correlation between decreased
colon microperfusion and increased rate of dehiscence in
rectal surgery has been reported.43 Our findings indicate
that IMPACT provides therapeutic benefits beyond standard nutrition. Furthermore, our findings indicate that IMPACT promotes the healing of intestinal anastomoses
performed during GI surgery, as we saw a consistent
reduction in the incidence of anastomotic leaks with preoperative supplementation. Collectively, IMPACT supplementation appears warranted in surgical patients as part of
the proactive approach to infection control. For the maximum benefit, patients should receive 0.51 l of IMPACT
for 57 days before surgery. In cases in which full preoperative treatment is not possible, early postoperative support with IMPACT can improve postoperative outcomes
beyond standard nutrition.

ACKNOWLEDGEMENTS
The hypothesis for this investigation was conceived at a
workshop attended by leading practitioners from Asian
Pacific and Latin American regions and co-chaired by
Drs. Waitzberg and Saito. The authors gratefully
acknowledge Novartis Medical Nutrition, Nyon Switzerland, for sponsoring the workshop, and Dr. Heinz
Schneider of HealthEcon, Health Service Consultants,
Basel, Switzerland, for organizing the event. We also
thank Dr. T. Grunberger, Dr. R. Tepaske, and Dr. L.
Krahenbuhl for supplying outcome data prior to their
publication; A. Schmid for data accrual; and Dr. C. Galani
for statistical support.

REFERENCES
1. Tepaske R, te Velthuis H, Oudemans-van Straaten HM,
et al. Effect of preoperative oral immune-enhancing
nutritional supplement on patients at high risk of infection
after cardiac surgery: a randomised placebo-controlled
trial. Lancet 2001;358:696701.
2. Heys SD, Walker LG, Smith I, et al. Enteral nutrition supplementation with key nutrients in patients with critical illness and cancer. Ann Surg 1999;229:476477.

1603
3. Beale RJ, Bryg DJ, Bihari DJ. Immunonutrition in the critically ill: a systematic review of clinical outcome. Crit Care
Med 1999;27:27992805.
4. Heyland DK, Novak F, Drover JW, et al. Should immunonutrition become routine in critically ill patients? A systematic review of the evidence. JAMA 2001;286:944953.
5. Montejo JC, Zarazaga A, Lopez-Martinez J, et al. Immunonutrition in the intensive care unit. A systematic review
and consensus statement. Clin Nutr 2003;22:221233.
6. Wyncoll D, Beale R. Immunologically enhanced enteral
nutrition: current status. Curr Opin Crit Care 2001;7:128
132.
7. Daly JM, Lieberman MD, Goldfine J, et al. Enteral nutrition
with supplemental arginine, RNA, and omega-3 fatty acids
in patients after operation: immunologic, metabolic, and
clinical outcome. Surgery 1992;112:327338.
8. Gianotti L, Braga M, Frei A, et al. Health care resources
consumed to treat postoperative infections: cost savings
by perioperative immunonutrition. Shock 2000;14:325
330.
9. Kemen M, Senkal M, Homann HH, et al. Early postoperative
enteral nutrition with arginine, omega-3 fatty acids, and
ribonucleic acid-supplemented diet versus placebo in cancer patients: an immunologic evaluation of IMPACT. Crit
Care Med 1995;23:652659.
10. Senkal M, Kemen M, Homann HH, et al. Modulation of
postoperative immune response by enteral nutrition with a
diet enriched with arginine, RNA, and omega-3 fatty acids in
patients with upper gastrointestinal cancer. Eur J Surg
1995;161:115122.
11. Braga M, Gianotti L, Gestari A, et al. Gut function, immune
and inflammatory responses in patients fed perioperatively
with supplemented formulas. Arch Surg 1996;131:1257
1265.
12. Wachtler P, Hilger RA, Ko nig W, et al. Influence of a preoperative enteral supplement on functional activities of
peripheral leukocytes from patients with major surgery. Clin
Nutr 1995;14:275282.
13. Kulkarni AD, Rudolph FB, Van Buren CT. The role of dietary
sources of nucleotides in immune function: a review. J Nutr
1994;124:1442S1446S.
14. Kirk SJ, Hurson M, Regan MC, et al. Arginine stimulates
wound healing and immune function in elderly human
beings. Surgery 1993;114:155160.
15. Sato H, Zhao ZQ, McGee DS, et al. Supplemental L-arginine during cardioplegic arrest and reperfusion avoids regional postischemic injury. J Thorac Cardiovasc Surg 1995;
110:302314.
16. Jones SM, Thurman RG. L-arginine minimizes reperfusion
injury in a low-flow, reflow model of liver perfusion. Hepatology 1996;24:163168.
17. Nonami Y. The role of nitric oxide in cardiac surgery. Surg
Today 1997;27:583592.
18. OLeary MJ, Coakley JH. Nutrition and immunonutrition. Br
J Anaesth 1996;77:118127.

1604
19. Meydani SN, Dinarello CA. Influence of dietary fatty acids
on cytokine production and its clinical implications. Nutr Clin
Pract 1993;8:6572.
20. Calder PC. Immunoregulatory and anti-inflammatory effects
of n-3-polyunsaturated fatty acids. Braz J Med Biol Res
1998;31:467490.
21. Luiking YC, Poeze M, Dehong CH, et al. Sepsis: an arginine
deficiency state? Crit Care Med 2004;32:(10) 21352145.
22. Burdett S, Stewart LA, Tierny JF. Publication bias and
meta-analysis. Int J Technol Assess Health Care 2003;
19:129134.
23. Heyland DK, MacDonald S, Keefe L, et al. Total parenteral
nutrition in the critically ill patient: a meta-analysis. JAMA
1998;280:20132019.
24. Jadad A. (1998) Random controlled trials BMJ, London.
25. Rothman KJ, Boice JD. (1979) Epidemiological analysis
with a programmable calculator National Institutes of
Health, Washington, DC.
26. Fleiss JL, Gross AJ. Meta-analysis in epidemiology, with
special reference to studies of the association between
exposure to environmental tobacco smoke and lung cancer.
J Clin Epidemiol 1991;44:127139.
27. Hedges LV. Estimation of effect size from a series of
independent experiments. Psychol Bull 1982;92:490499.
28. Braga M, Gianotti L, Vignali A, et al. Preoperative oral
arginine and 3 fatty acid supplementation improves the
immunometabolic host response and outcome after colorectal resection for cancer. Surgery 2002;132:805814.
29. Gianotti L, Braga M, Nespoli L, et al. A randomized controlled trial of preoperative oral supplementation with a
specialized diet in patients with gastrointestinal cancer.
Gastroenterology 2002;122:17831770.
30. Braga M, Gianotti L, Nespoli L, et al. Nutritional approach in
malnourished surgical patients: a prospective randomized
study. Arch Surg 2002;137:174180.
31. Daly JM, Weintraub FN, Shou J, et al. Enteral nutrition
during multimodality therapy in upper gastrointestinal cancer patients. Ann Surg 1995;221:327338.
32. Schilling J, Vranjes N, Fierz W, et al. Clinical outcome and
immunology of postoperative arginine, omega-3 fatty acids,
and nucleotide-enriched enteral feeding: a randomized
prospective comparison with standard enteral and low calorie/low fat i.v. solution. Nutrition 1996;12:423429.

Waitzberg et al.: Nutrition for Elective Surgery

33. Senkal M, Mumme A, Eickhoff U, et al. Early postoperative
enteral immunonutrition: clinical outcome and cost-comparison analysis in surgical patients. Crit Care Med
1997;25:14891496.
34. Heslin MJ, Latkany L, Leung D, et al. A prospective, randomized trial of early enteral feeding after resection of upper gastrointestinal malignancy. Ann Surg 1997;226:567
577.
35. Gianotti L, Braga M, Vignali A, et al. Effect of route of
delivery and formulation of postoperative nutritional support
in patients undergoing major operations for malignant
neoplasms. Arch Surg 1997;132:12221230.
36. Jiang ZM. The role of immune enhanced enteral nutrition on
plasma amino acid, gut permeability and clinical outcome in
post-surgery patients. Acta Chin Acad Med Sci 2001;
23:515518.
37. Senkal M, Zumtobel V, Bauer KH, et al. Outcome and costeffectiveness of perioperative enteral immunonutrition in
patients undergoing elective upper gastrointestinal tract
surgery: a prospective randomized study. Arch Surg
1999;134:13091316.
38. Snyderman CH, Kachman K, Molseed L, et al. Reduced
postoperative infections with an immune-enhancing nutritional supplement. Laryngoscope 1999;109:915921.
39. Braga M, Gianotti L, Radaelli G, et al. Perioperative immunonutrition in patients undergoing cancer surgery: results of a randomized double-blind phase 3 trial. Arch Surg
1999;134:428433.
40. Braga M, Bianotti L. Preoperative immunonutrition: costbenefit analysis. J Parenter Enteral Nutr 2005;29:(1) S57
S61.
41. Wheeler JM, Gilbert JM. Controlled intraoperative water
testing of left-sided colorectal anastomoses: are ileostomies
avoidable?. Ann R Coll Surg Engl 1999;81:105108.
42. Strickland A, Brogan A, Krauss J, et al. Is the use of specialized nutritional formulations a cost-effective strategy? A
national database evaluation. J Parenter Enteral Nutr 2005;
29:(1) S81S91.
43. Vignali A, Gianotti L, Braga M, et al. Altered microperfusion
at the rectal stump is predictive for rectal anastomotic leak.
Dis Colon Rectum 2000;43:7682.

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.