Sepsis and Organ Dysfunction Bad and Good News on

Prevention and Management 1st Edition

Visit the link below to download the full version of this book:

https://medidownload.com/product/sepsis-and-organ-dysfunction-bad-and-good-news-
on-prevention-and-management-1st-edition/

Click Download Now

A.E. BAUE, M.D.
Department of Surgery, Saint Louis University, Health Sciences Center, St. Louis - USA
G. BERLOf, M.D.
Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic,
Trieste University Medical School, Trieste - Italy
A. GULLO, M.D.
Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic,
Trieste University Medical School, Trieste - Italy
J.-L. VINCENT, M.D.
Department of Intensive Care, Erasme University Hospital
Free University of Brussels - Belgium

O.EA. - ORGAN FAILURE ACADEMY, VIA BATIISTI, 1 - 34125 TRIESTE (ITALY)

Steering Committee
A.E. Baue, M.D., Department of Surgery, Saint Louis University
Health Sciences Center, St. Louis - USA
G. Berlot, M.D., Department of Clinical Sciences, Section of Anaesthesia, Intensive Care
and Pain Clinic, Trieste University Medical School, Trieste - Italy
A. Gullo, M.D., Department of Clinical Sciences, Section of Anaesthesia, Intensive Care
and Pain Clinic, Trieste University Medical School, Trieste - Italy
L. Silvestri, M.D., Department of Anaesthesia and Intensive Care, Gorizia Hospital, Gorizia - Italy
G. Sganga, M.D., Department of Surgery, and C.N.R. Shock Centre
Catholic University of Sacro Cuore, Rome - Italy
ISBN-13: 978-88-470-0137-4 e-ISBN-13: 978-88-470-2229-4
001: 10.1007/978-88-470-2229-4
© Springer-Verlag Italia, Milano 2001

Springer-Verlag Italia
A member of BertelsmannSpringer Science+Business Media GmbH

Library of Congress Cataloging-in-Publication Data: Applied for

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcast-
ing, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publica-
tion or parts thereof is only permitted under the provisions of the Italian Copyright Law in its current
version and permission for use must always be obtained from Springer-Verlag. Violations are liable for
prosecution under the Italian Copyright Law.
The use of general descriptive names, registered names, trademarks, etc., in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protec-
tive laws and regulations and therefore free for general use.
Product liability: the publishers cannot guarantee the accuracy of any information about dosage and
application contained in this book. In every individual case the user must check such information by
consulting the relevant literature.

Cover design: Simona Colombo, Milan, Italy

SPIN: 10789915
Table of Contents

Epidemiology of Infections in ICUs: Where Are We?

M. VIVIANI, G. BERLOT, AND A. GULLO ...................................................................................... II

Prothrombin Fragment 1+2 Levels Are Associated with Pulmonary and Renal
Responses to Cardiopulmonary Bypass
B. DIXON, J.D. SANTAMARIA, AND D.l CAMPBELL ..................................................................... 23

Oxidative Stress and Apoptosis in Sepsis and the Adult Respiratory Distress Syndrome
T. WHITEHEAD, AND H. ZHANG .................................................................................................... 33

Neutrophil Defensins in Lung Inflammation

H. ZHANG, AND T. WHITEHEAD.................................................................................................... 39

Nitric Oxide: Lessons Learned and Areas of Success

W.M. ZAPOL, AND R. JENNEY ...................................................................................................... 47

Protecting Renal Blood Flow in the Intensive Care Unit

J.A. KELLUM ................................................................................................................................ 53

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis

R.P. DELLINGER............................................................................................................................ 63

Light and Shadow: Perspectives on Host-Microbial Interactions in the Pathogenesis

of Intensive Care Unit-Acquired Infection
J.C. MARSHALL............................................................................................................................ 75

Identification and Characterization of Protein Tyrosine Phosphatases Expressed

in Human Neutrophils
J. KRUGER, T. FUKUSHIMA, AND G.P. DOWNEy............................................................................ 85

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption

with Immobilised Human Serum Albumin
K. REINHART, AND M. ZIMMERMANN ........................................................................................... 103

Hemofiltration in Intensive Care

G. BERLOT, AND M. VIVIANI ........................................................................................................ III

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology
A.E. BAUE .................................................................................................................................... 123

Index ........................................................................................................................................... 133

Authors Index

BaueA.E.
Professor of Surgery Emeritus, Vice President for the Medical Center Emeritus, Department of Surgery,
Saint Louis University School of Medicine. Fishers Island, Ney York (U,S,A,)

Berlot G.
Department of Clinical Sciences. Section of Anaesthesia. Intensive Care and Pain Clinic, Trieste University
Medical School, Trieste (Italy)

Campbell D.J.
Intensive Care Centre. St Vincent's Hospital, Melbourne (Australia)

Dellinger R.P.
Department of Internal Medicine. Rush University, Chicago, Illinois (U,S,A,)

Dixon B.
Intensive Care Centre. St Vincent's Hospital, Melbourne (Australia)

Downey G.P.
Clinical Sciences Division. Medical Sciences Building, University of Toronto, Toronto, Ontario (Canada)

Fukushima T.
Department of Medicine, Division of Respirology, The University of Toronto, Toronto, Ontario (Canada),
and The Hospital for Sick Children Research Institute, Toronto. Ontario (Canada)

GulloA.
Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University
Medical School, Trieste (Italy)

Jenney R.
Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts (U,S.A)

Kellum J.A.
Departments of Anaesthesiology/CCM and Medicine, University of Pittsburgh Medical Centre, Pittsburgh,
Pennsylvania (U.S.A)

Kruger J.
Department of Medicine, Division of Respirology, The University of Toronto, Toronto, Ontario (Canada)

Marshall J.e.
Department of Surgery and Programme in Critical Care Medicine, Toronto General Hospital and the
University of Toronto, Toronto, Ontario (Canada)

Reinhart K.
Clinic for Anaesthesiology and Intensive Care, Friedrich-Schiller University Jena, Jena (Germany)

Santamaria J.D.
Intensive Care Centre, St Vincent's Hospital, Melbourne (Australia)

Viviani M.
Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University
Medical School, Trieste (Italy)

Whitehead T.
Divisions of Respiratory and Critical Care Medicine, Medical Sciences Building, University of Toronto,
Toronto, Ontario (Canada)
VIII

ZapoIW.M.
Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts (U.S.A.)

ZbangH.
Divisions of Respiratory and Critical Care Medicine, Medical Sciences Building, University of Toronto,
Toronto, Ontario (Canada)

Zimmermann M.
Fresenius HemoCare Adsorber Technology GmbH, St. Wendel (Germany)
Abbreviations

ALT, alanine aminotransferase MRSA, methicillin-resistant Staphylococcus

ARD, acute respiratory distress syndrome MuLV, murine leukemia virus
ARF, acute renal failure NO, nitric oxide
ATN, acute tubular necrosis PAIl, plasminogen activator inhibitor I
CAVH, continuous arteriovenous haemofil- PAl, plasminogen activator inhibitor
tration
PaOz/Fi0 2, arterial partial pressure of oxy-
CDC, Centers for Disease Control gen to inspired fraction of oxygen ratio
cGMP, cyclic guanosine monophosphate PCD, programmed cell death
CHF, continuous haemofiltration PCR, polymerase chain reaction
CI, cardiac index PCWP, pulmonary capillary wedge pressure
CNS, coagulase-negative Staphylococcus PNPP, p-nitrophenyl phosphate
CPB, cardiopulmonary bypass PPHN, persistent pulmonary hypertension of
the newborn
CVP, central venous pressure
PPM, potentially pathogenic microorganisms
CVVH, continuous venovenous haemofilta-
tion PTF 1+2, prothrombin fragment 1+2
DFP, diisopropylfluorophosphate PTK, protein tyrosine kinases
DTT, dithiothreitol PTP, protein tyrosine phosphatases
ECL, enhanced chemiluminescence PVRI, pulmonary vascular resistance index
ECMO, extracorporeal membrane oxygena- RBF, renal blood flow
tion
ROS, reactive oxygen species
EDRF, endothelium-derived relaxing factor
RT-PCR, reverse transcription polymerase
GFR, glomerular filtration rate chain reaction
Hb, haemoglobin SDD, selective decontamination of the diges-
tive tract
HNP, human neutrophil peptides
SDS-PAGE, sodium dodecyl sulfate-poly-
HRP, horse radish peroxidase
acrylamide gel electrophoresis
iHSA, immobilized human serum albumin
sGC, soluble guanylate cyclase
LAL, Iimulus amoebocyte lysate
SIRS, systemic inflammatory response syn-
LOCM, low osmolality radiocontrast media drome
LOS, length of stay SOD, superoxide dismutase
LV, left ventricle SVRI, systemic vascular resistance index
MAP, mean arterial pressure YAP, ventilator-associated pneumonia
MBP, mannose-binding protein VRE, vancomycin-resistant Enterococcus
MODS, multiple organ dysfunction syndrome WCC, white cell count
MPAP, mean pulmonary artery pressure XOD, xanthine oxidase
Epidemiology of Infections in ICUs: Where Are We?

M. VIVIANI, G. BERLOT, A. GULLO

Patients admitted to intensive care units (lCUs) represent 8 to 15% of overall

hospital population [1]. A frequent problem related to lCU stay is the relative
high incidence of infections, rates of which are higher than 40% as reported in
some prevalence studies [2, 3]. This percentage is 5-10 times higher when com-
pared with the infections in patients admitted in the normal wards; furthermore,
in some hospitals of United States, recurrence of infections in lCUs represents
more than 20% of overall nosocomial infections [4].
lCU-acquired infections represent a major concern reflecting 80% of total
episodes one week after admission in intensive care [2]. Localisation and epi-
demiology of infections are well reported in the current literature. Pulmonary,
urinary tract and bloodstream infections are mainly observed [5], but etiology is
changing in the last years. After the predominant presence of Gram-negative
bacteria, generally sensitive to common antimicrobial agents, a progressive in-
crease of Gram-positive infections is documented during the 1980s [6] and
1990s. Particularly, resistant strains (coagulase-negative Staphylococcus,
Staphylococcus aureus and Enterococci) are often involved in nosocomial and
lCU epidemiology. This change in etiology is mainly associated to aggressive
antibiotic therapy, type of admission and environmental factors. Hence preven-
tion of infection and a tailored policy of antimicrobial administration are recom-
mended [7], but, actually, the proposed strategies have obtained conflicting re-
sults. The selective decontamination of the digestive tract (SDD) has been re-
cently re-evaluated and considered a good strategy for the prevention of respira-
tory tract infection [8]. Unfortunately a consensus is not yet established in the
clinical practice because some controversies, such as emergence of resistant mi-
croorganisms and the impact on mortality, are attributed to SDD application.

Epidemiology of infection in intensive care units

Many publications [1-3] report a lot of data about infections and stress the im-
portance of epidemiology monitoring as a fundamental step in diagnosis, treat-
ment and prevention strategies. This is a relevant matter because lCUs receive
5-10% of hospital patients accounting for more than 20% of overall nosocomial
12 M. Viviani, G. Beriot, A. Gullo

infections [9]. This high incidence has been correlated with the length of stay
(LOS) in intensive care [2, 5] which, in tum, depends on other factors such as
the severity of illness and the use of invasive procedures. Furthermore, over-
crowding and animate reservoirs (colonized subjects), promoting cross trans-
mission, represent other major risk factors. A large European multicenter trial,
published in 1995, evaluated the prevalence of infections and related risk factors
in intensive care [2]. The authors observed that half of the subjects admitted to
ICUs were infected (50% of them had acquired the infection in ICU); further-
more increased risk of death was associated to LOS, pneumonia and sepsis. A
recent similar multicenter study [3] and some incidence studies reported similar
results [10] (Table 1).

Table 1. Infections in ICU pati~nts. The percentage of community, hospital and ICU episodes are
referred to overall infected subjects enrolled in the studies

Vincent Leon-Rosales Legras

et al. [2] et al. [3] et al. [10]
Total infected patients 44.8% 58.2% 42.4%
Community - Acq. Infection 32.5% 41.1% 49.5%
Hospital - Acq. Infection 21.6% 19.0% 19.6%
ICU - Acq. Infection 45.9% 39.9% 30.9%

Low respiratory tract, urinary tract and bloodstream represent the main local-
izations of infections in ICU patients. Anyway the incidence varies among
countries, hospitals and even different ICUs in the same hospital. This particular
distribution is associated to widespread use of mechanical ventilation, bladder
catheterisation and intravascular catheters [11]; moreover considerable varia-
tions can be attributed to intrinsic factors (i.e. age, LOS, severity of illness, type
of admission) and extrinsic factors (i.e. differences in diagnostic methods and
prevention measures). Consequently even large prevalence multicenter trials
show some limits in the evaluation of this complex problem.
The most relevant point lies on the fact that cross-sectional studies provide
only a snapshot at the time of infective episodes, which implies the overestima-
tion of long-duration infections and the underestimation of short-duration
episodes. Then, longitudinal trials seem to be more useful to assess the inci-
dence rates of infectious diseases because the calculation can be corrected on
patients' ICU stay or patients' days devices [11, 12].
Interestingly Weber et al. [13], evaluating NNISS data, reported that the fre-
quency of different sites of infections and the attributed risk factors change ac-
cording to the different type of ICU. The rate of ventilator-associated pneumo-
nia, for instance, is higher in surgical, neurosurgical, bum and trauma units than
Epidemiology of Infections in ICDs: Where Are We? 13

in the medical and coronary ICUs, whilst paediatric and bum patients show a
greater incidence of central venous lines associated infections. Noteworthy,
Wallace observed that trauma patients admitted in intensive care were signifi-
cantly more infected than surgical subjects [14]. In this study the increased risk
was attributed to many emergency procedures (intubation and vascular cannula-
tion), massive transfusion and head injury found in the trauma group.
The characteristic of patients and the type of admission are not the only fac-
tors influencing the distribution of infections. In a comparative analysis involv-
ing five French ICUs, Legras et al. [10] showed a significant variation of ICU-
acquired infection among different units. Moreover, the incidence rates varied in
the same ICU when two distinct periods were considered. The authors conclud-
ed that such variations could be related to the difficulty of achieving standard-
ized definitions of infections (especially pneumonia). The impossibility of ob-
taining homogeneous diagnostic methods was a further confounding factor
showing the difficulty to achieve good comparative results even when standard-
ized protocols were adopted.
In conclusion, although large trials are essential for the epidemiologic as-
sessment, the knowledge of local micro-ecology may play an important role for
a correct approach to infection monitoring, therapy and prevention program.

Classification of infections: are we adopting the right approach?

The classification of infections is very important, particularly for the prevention
and management strategies adopted in the intensive care setting. In 1988 the
Centers of Disease Control and Prevention (CDC) [15] proposed the well-known
time related criteria to distinguishing infections in ICU. Unfortunately, CDC es-
tablished no specific "cut off" time to determine the difference among commu-
nity, hospital and ICU acquired infections. Arbitrary periods, varying from 48 to
120 hours were employed in epidemiological trials [16, 17]. Adopting this mod-
el and establishing the "cut off" at 48 hours, many studies showed that infections
were mainly ICU-acquired (Table 1) whilst lower rates of community-acquired
episodes were found. Only de Leon-Rosales [3] observed a high rate of commu-
nity infections in patients admitted to ICUs. The explanation consisted in the
common finding of untreated chronic infections outside the hospitals.
As stated before, the main criticism to the CDC model is that a change in the
"cut off" time implies a modification in the attributed classification confounding
the ICU acquired infections with the other two categories. In fact, some authors,
attempting to overcome this problem, introduced the concept of "early onset"
and "late onset" infections [18, 19]. Furthermore, short knowledge about the
ICU ecology is obtained adopting the time criterion.
A novel pathophysiological classification, based on patients' carrier state has
been proposed to distinguish the origin of infections and to organize a suitable
14 M. Viviani, G. Berlot, A. Gullo

prevention program in the intensive care setting. Using this approach, infections
are split into primary endogenous, secondary endogenous and exogenous [20].
The first is caused by potentially pathogenic microorganisms [21] (PPM) car-
ried by the patient in throat and rectum on admission to ICUs. The second type
is invariably caused by hospital PPM, which are acquired in the oropharynx and
gut during the ICU stay but not present on admission. The causative bacteria of
exogenous infections are hospital PPM never carried by the patient during the
ICU stay.
A recent survey published by Silvestri and colI. [22] interestingly observed
that primary endogenous infections accounted for 60% of total diagnosed
episodes. These results were in stark contrast with the CDC time criterion
(adopting this method 80% of infections were ICU-acquired) and are in accor-
dance with data obtained in our experience (unpublished data) (Fig. 1). More-
over the authors found that the best time "cut off" between ICU-acquired infec-
tions and communitylhospital acquired infections was 9 days.
Hence the carrier-state criterion gives a new insight into the type of infection
showing that most episodes are not really ICU-acquired but imported into the
intensive care unit. As a consequence change in prevention and therapy strate-
gies is necessary. Finally this method, unlike the CDC classification, suggests to
the intensivists a novel pathophysiological approach to the infection problem
stressing the pivotal role of causative bacteria and ICU ecology.

Resistant bacteria: a major problem in leu

Many epidemiological studies show that Gram-negative bacteria are the main
etiologic agents isolated in ICU infections [2, 5]. Anyway increasing evidence
of Gram-positive infections was observed in the last years [3, 10]. Possibly, this
trend could be explained by the selective prevention of Gram-negative infec-
tions (antibiotic use), the widespread usage of indwelling devices and the grow-
ing importance of some Gram-positive bacteria (coagulase-negative Staphylo-
coccus or CNS) as pathogenic agents [6]. In contrast data about yeasts infec-
tions are controversial. Even if large epidemiological trials report a relative high
rate of Candida spp. infections accounting for 8-15% [2, 5] of cases, a lower in-
cidence « 2%) was found in others studies [23,24]. The substantial discrepan-
cy between these data could be explained by the high rate of Candida coloniza-
tion among critically ill patients [25] (especially in the lower airways and uri-
nary tract) and the difficult diagnostic approach to fungal infections [26] (i.e.
isolation of yeasts from specimens and the need of pathological findings). Simi-
larly to other bacteria, many mycetes isolated in intensive care units are resistant
to antimicrobials, especially to azoles [27]. This is an emerging crucial problem
because previous prolonged course of azoles can induce resistance among Can-
dida species [28].
Epidemiology of Infections in ICUs: Where Are We? 15

80 0 Silvestri L. (74 infectious episodes)

70
III Viviani M. (206 infectious episodes)
60
50
% 40·
30 i
20
10
o
u.i 0 & & &
x u u
~
cti UJ <: <:
P. E. ;:J
cj :I: S2
Fig. 1. Classification of infections in intensive care using pathophysiological (P.E. = primary en-
dogenous; S.E. = secondary endogenous; Exo. = exogenous) and time (c. Acq. = community-ac-
quired; H. Acq. = hospital-acquired; ICU Acq. = ICU-acquired) criteria (see full text). From Sil-
vestri [22] and Viviani (unpublished data)

Another important issue consists in the distribution of causative bacteria

among different intensive care units. It varies in accordance to the site of infec-
tion, type of ICU, type of admission, length of stay and countries. Actually,
many authors [5, 12-14] observed that Gram-negatives were mainly involved in
low respiratory and urinary tract infections, whilst bloodstream and cardiovas-
cular infections were caused principally by Gram-positive bacilli (Staphylococ-
cus aureus, CNS, Enterococci). Hence strong antibiotic therapy is often adopted
to control such infections promoting the risk of emerging resistant strains.
Current literature reports the crucial problem regarding the resistant micro-
organisms as cause of particularly nosocomial and ICU infections [13]. Al-
though Gram-negative resistant bacteria such as Pseudomonas and Acinetobac-
ter are well described in intensive care [ 13, 29], data obtained from epidemio-
logical trials show that infections due to Gram-positive resistant strains are
emerging (Fig. 2) in the last years [6, 13, 30, 31]. Methicillin-resistant Staphylo-
coccus (MRSA), Enterococcus resistant to vancomycin (VRE) and CNS are iso-
lated frequently in critically ill patients especially in the bloodstream, wounds
and low respiratory tract. Interestingly the distribution of MRSA infections
varies among different countries. A recent European study showed a wide diffu-
sion of MRSA in Italy, France and Spain (> 30%), whilst the presence of MRSA
in northern Europe is minimal. The basic mechanism related to resistance relies
upon some important considerations [30, 32]. First, over time bacteria accumu-
late mutations which inhibit the activity of entire classes of antimicrobials; sec-
ond, the severity of illness, the large use of indwelling devices, the length of
stay are factors favouring bacteria overgrowth especially in the gastrointestinal
16 M. Viviani, G. Berlot, A. Gullo

70

60
rII
c::
.~ 50
rII

N
.!!!
40

30
e
rII

.....
0 20
-'-MRSA(2)
?f-
lO -+- Enterococcus

Fig. 2. Proportion of resistant strains over time. (eNS = Schaberg et aI. [6]; MRSA (1) = resistant
strains isolated in bloodstream, Livermore [30]; MRSA (2) = Huang et al. [31]; Enterococcus =
Weber et al. [13]

tract [33]; third, the excessive antibiotic usage is related to the development of
resistance; last but not least, the failing development of new antibiotics in the re-
cent years promotes the acquisition of resistance [32, 34].

Prevention of infections
As described herein, the infections in intensive care represent a very complex is-
sue for the intensivist. The different epidemiological observed data, the impor-
tance of classification and the evident spread of resistance are stimulating argu-
ments for a tailored policy approach to leu infections. Prevention plays a piv-
otal role in this context, but conventional measures [35] including isolation and
cohorting of colonized and/or infected patients, hand washing, aseptic proce-
dures, early removal of devices and antibiotic restriction have led to conflicting
results. Weinstein [1] argued that these manoeuvres are directed against exoge-
nous pathogens, whilst leu infections are due principally to endogenous flora.
Moreover the wide variation in the adoption of prevention practices may explain
the limited benefits of infection control protocols [36]. Some authors [37-39] ar-
gue that high standards of hygiene, isolation of carriers and halting transmission
of bacteria between patients and healthcare workers may reduce the morbidity
associated with multi-resistant pathogens in leus. Anyway, up to now, there are
no properly validating surveys showing a relationship between hand washing
and reduction of leu infections [40]. The use of gloves and gowns is not stan-
dardised in intensive care and it is unhelpful when adopted as single practice
Epidemiology of Infections in ICDs: Where Are We? 17

[41,42]. The environmental problems including inadequate location of intensive

care beds, ventilation systems, structures for storage favour the acquisition and
transmission of exogenous pathogens [43]. Furthermore, leu overcrowding and
shortage of nursing staff hamper the isolation of carriers and infected patients
with resistant organisms. Lastly, the implementation of antibiotic policies in in-
tensive care settings is very difficult because many factors are involved in the
decision-making process [44]. First, the bad clinical conditions of patients and
the high probability of resistant pathogen infections contribute to the prescrip-
tion of broad-spectrum antimicrobials. Second, microbiologic surveillance, an-
tibiotic therapy monitoring and strict cooperation with microbiologists and
pharmacologists are necessary to improve antibiotic control policy.
Many investigators have evaluated the relationship between leu infections,
intestinal flora and gut physiology [45-48]. Bacterial translocation is well docu-
mented in animals but it is debated in humans. Possibly the main reason consists
in experimental technical difficulties [49]. Anyway O'Boyle et al. [50] argues
that low levels of bacteria translocation associated to microorganism over-
growth, immunosuppression and increased intestinal permeability represent im-
portant factors promoting infections. This is an intriguing hypothesis knowing
that the widespread systemic antimicrobial usage adopted in leu settings pro-
mote multi-resistant PPM overgrowth in the gut overwhelming normal intestinal
flora [32, 51].
Selective decontamination of digestive tract has shown to reduce low respira-
tory tract infections and mortality in critically ill patients [8, 52]. The philoso-
phy of this novel approach consists in the eradication of PPM from the gut with
minimal impact on normal intestinal flora [53]. The full SDD protocol includes
the following: parenteral infusion of cephalosporins (i.e. cefotaxime) on pa-
tients' admission for the prevention of primary endogenous infections; topical
application of non-absorbable antibiotics for the prevention of secondary en-
dogenous infections; high standards of hygiene and surveillance cultures from
throat and rectum for the prevention of exogenous infections and for the evalua-
tion of resistant strains and efficacy of SDD manoeuvres.
The main criticisms to SDD application are [54]: the threat of emerging re-
sistant pathogens, high costs, the promotion of Gram-positive colonization and
infection. Interestingly a recent meta-analysis [8], evaluating more than 30 tri-
als, didn't report the emergence of resistant microorganisms during SDD treat-
ment. Differently from systemic administration of antimicrobials, topical appli-
cation of non-absorbable antibiotics eradicates PPM (Pseudomonas, Enterobac-
ter etc.) because high concentrations in saliva and faeces are achieved. Some in-
vestigators [55, 56] undertook studies evaluating total costs of stay in leu set-
tings. They observed that expenditure was lower in SDD treated patients com-
pared to control groups. Gram-positives, in particular MRSA, are intrinsically
resistant to SDD. Hence SDD application could be questioned because of pro-
moting MRSA gut overgrowth [56] and infections. Furthermore other factors,
such as parenteral administration of broad-spectrum antibiotics, underlying dis-
18 M. Viviani, G. Berlot, A. Gullo

eases, large use of invasive devices, breaches of hygiene, person-to-person

transmission contribute to the spread of MRSA infections in intensive care set-
tings [57]. Some investigators evaluated the effects of MRSA carriage eradica-
tion from nasal and oropharyngeal cavity on MRSA infection in ICUs [58,59].
Mupirocin nasal ointment application reduced significantly MRSA carriage and
infections. Unfortunately rapid emergence of resistant strains to mupirocin [60,
61] discouraged the use of this strategy. Topical bacitracin [59], chlorhexidine,
vancomycin [62] and povidone-iodine cream [63] has been evaluated as alterna-
tive to mupirocin. In our two-year experience topical application of vancomycin
in the oropharynx and through nasogastric tubes abolished MRSA gastrointesti-
nal overgrowth reducing secondary endogenous infections caused by MRSA in
ICU intubated patients [64].
Actually selective decontamination of the digestive tract has to be consid-
ered an evidence-based-medicine intervention; probably other difficulties must
be considered for its limited diffusion [65]. Mainly, proper application of SDD
strategy requires a strict cooperation between ICU staff and microbiologists and
pharmacologists; special efforts are requested for monitoring efficacy of SDD
policies, diagnosis of infections and emergence of resistance. Then a highly
professional ICU team is necessary considering that pharmaceutical companies
are not interested in supporting a much cheaper method for the prevention of
infections.

References
1. Weinstein RA (1991) Epidemiology and control of nosocomial infections in adult intensive
care units. AmJ Med 91[SuppI3B]:179S-184S
2. Vincent JL, Bihari DJ, Suter PM et al (1995) The prevalence of nosocomial infections in in-
tensive care units in Europe. JAMA 274:639-644
3. de Leon-Rosales SP, Molinar-Ramos F, Dominguez-Cherit G et al (2000) Prevalence of infec-
tions in intensive care units in Mexico: a multicenter study. Crit Care Med 28:1316-1321
4. Craven DE, Kunches LM, Lichtemberg DA et al (1988) Nosocomial infection and fatality in
medical and surgical intensive care unit patients. Arch Intern Med 148:1161-1168
5. NNlS System Report (1997) National Nosocomial Infections Surveillance (NNIS) report,
data summary from October 1986-April 1997, issued May 1997. Am J Infect Control 25:
477-487
6. Schaberg DR, Culver DH, Gaynes RP (1991) Major trends in the microbial etiology of noso-
comial infection. Am J Med 91[SuppI3B]:72S-75S
7. Bergogne-Berezin E (1999) Current guidelines for the treatment and prevention of nosocomi-
al infections. Drugs 58:51-67
8. D'Amico R, Pifferi S, Legnetti C et al (1998) Effectiveness of antibiotic prophylaxis in criti-
cally ill adult patients: systematic review of randomised controlled trials. Br Med Journal
316:1275-1285
9. Trilla A (1994) Epidemiology of nosocomial infections in adult intensive care units. Int Care
Med 20:S1-S4
10. Legras A, Malvy D, Quinioux AI et al (1998) Nosocomial infections: prospective survey of
incidence in five French intensive care units. Int Care Med 24:1040-1046