Early peak temperature and mortality in critically ill patients with or without infection

Abstract

Purpose
To determine whether fever is associated with an increased or decreased risk of death in patients admitted to an intensive care unit (ICU) with infection.
Methods
We evaluated the independent association between peak temperature in the first 24 h after ICU admission and in-hospital mortality according to whether there was an admission diagnosis of infection using a database of admissions to 129 ICUs in Australia and New Zealand (ANZ) (n = 269,078). Subsequently, we sought to confirm or refute the ANZ database findings using a validation cohort of admissions to 201 ICUs in the UK (n = 366,973).
Results
A total of 29,083/269,078 (10.8%) ANZ patients and 103,191/366,973 (28.1%) of UK patients were categorised as having an infection. In the ANZ cohort, adjusted in-hospital mortality risk progressively decreased with increasing peak temperature in patients with infection. Relative to the risk at 36.5–36.9°C, the lowest risk was at 39–39.4°C (adjusted OR 0.56; 95% CI 0.48–0.66). In patients without infection, the adjusted mortality risk progressively increased above 39.0°C (adjusted OR 2.07 at 40.0°C or above; 95% CI 1.68–2.55). In the UK cohort, findings were similar with adjusted odds ratios at corresponding temperatures of 0.77 (95% CI 0.71–0.85) and 1.94 (95% CI 1.60–2.34) for infection and non-infection groups, respectively.
Conclusions
Elevated peak temperature in the first 24 h in ICU is associated with decreased in-hospital mortality in critically ill patients with an infection; randomised trials are needed to determine whether controlling fever increases mortality in such patients.

Full text

P1
Thrombin-activatable  brinolysis inhibitor and organ dysfunction
in disseminated intravascular coagulation associated with sepsis
MHayakawa1, ASawamura1, MSugano1, SUegaki1, NKubota1, SGando1,
SJesmin2
1Department of Anesthesiology and Critical Care Medicine, Hokkaido
University Graduate School of Medicine, Sapporo, Japan; 2Division of Gene
Therapeutics, National Center for Global Health and Medicine, Tokyo, Japan
Critical Care 2011, 15(Suppl 3):P1 (doi: 10.1186/cc10370)
Introduction Fibrinolytic shutdown plays a pivotalrole in the
pathogenesis of multiple organ dysfunction syndrome (MODS) in
disseminated intravascular coagulation (DIC). We tested the hypothesis
that the levels of thrombin activatable  brinolysis inhibitor (TAFI) are
not su cient to overcome  brinolytic shutdown, thus contributing to
MODS and the poor prognosis in sepsis-induced DIC.
Methods Fifty patients with sepsis, severe sepsis, or septic shock were
enrolled in the study. The DIC was diagnosed based on the Japanese
Association for Acute Medicine (JAAM) DIC criteria. The overt DIC scores
based on the International Society on Thrombosis and Haemostasis
(ISTH) were also calculated. On the day of sepsis diagnosis (day 1), and
days 3 and 5, we measured TAFI, soluble  brin, and global coagulation
and  brinolysis markers.
Results The JAAM DIC scores on day 1 and maximum JAAM DIC scores
were independent predictors of patient death and MODS, respectively.
The JAAM DIC patients, especially those who simultaneously met the
ISTH overt DIC criteria, showed lower TAFI antigen levels and activity,
and higher levels of soluble  brin in comparison with non-DIC patients.
There were di erences in the levels of soluble  brin and TAFI activity
between the patients with and without MODS. The  ndings of stepwise
logistic regression and multiple regression analyses suggested that low
TAFI activity is an independent predictor of patient death and MODS. A
multiple regression analysis also indicated that soluble  brin negatively
correlated with the TAFI activity in DIC patients.
Conclusion Thrombin activation results in the consumption of TAFI.
Low TAFI activity is involved in the pathogenesis of DIC-induced MODS
and poor prognosis.
P2
Anti-endotoxin immunity in abdominal sepsis patients
OButyrsky, VStarosek
Department of Surgery, Crimean Medical University, Simferopol, Ukraine
Critical Care 2011, 15(Suppl 3):P2 (doi: 10.1186/cc10371)
Introduction Anti-endotoxin immunity (AEI) has many biological e ects
but the problem of conjugation and elimination of lipopolysaccharides
(LPS) in peritonitis patients is not discussed. We investigated the role
of IgA, IgM and IgG in peritonitis and their association with humoral
immunity (HI).
Methods We investigated 33 patients (male:female = 25:8) with ab-
domi nal sepsis (total peritonitis in appendicitis, perforated duodenal
ulcer, pancreonecrosis). Anti-endotoxin (AE) antibodies (anti-LPS-IgA,
anti-LPS-IgM, anti-LPS-IgG) were determined by original modi cation
of hard-phase immunoenzyme analysis. Escherichia coli K30 LPS
was used as antigen for AE antibody detection. The level of general
immunoglobulin was determined by the microturbidimetric method
with human monospeci c sera to IgG, IgA and IgM. All data were
compared with healthy donors (99 patients).
Results A high level of AEI and HI was determined in 24% of patients
who recovered rapidly without complication after surgery, discharged
in 9 to 10 days. This was con rmed by clinical data (normalization of
body temperature, peristalsis, spontaneous stool) by 4 to 5 days. A low
level of AEI and HI was found in 42% of patients who recovered slowly; in
a favorable course of peritonitis, the increase of parameters was marked
by 8 to 10 days; in several with suppuration of wounds, discharge was
in 14 to 16 days. A few patients with a low level of immunity against the
background of abdominal sepsis required therapy with sandoglobulin
H that was accompanied with a sharp positive change of a postoperative
course of peritonitis and an increase of immunity indices. See Table1.
An evident decrease of AE antibodies may be a background for
translocation of endotoxin from the intestine to the portal and systemic
circulation. Disorder of AE mechanisms of endotoxin conjugation may
activate other mechanisms of neutralization (endotoxin-conjugating
protein) that stimulate CD14-receptor structures and mechanisms of
active production of proin ammatory cytokines and starting systemic
in ammatory response syndrome.
Conclusion Abdominal sepsis patients are determined dysfunction of
AEI (decrease of AE IgM and IgG). Successful treatment of peritonitis
is accompanied with normalization of the IgM and IgG concentration
and an increase of IgA above standard. Dynamics of AE antibodies may
be a marker of the clinical course and forecast of abdominal sepsis.
Comparative analysis of HI and AEI demonstrates parallelism of the
dynamic concentration of immunoglobulins during treatment.
© 2010 BioMed Central Ltd
Sepsis 2011
Beijing, China. 26–28 October 2011
Published: 27 October 2011
POSTER PRESENTATIONS
Table 1 (abstract P2)
AEI HI
IgA IgG IgM IgA IgG IgM
Before After Before After Before After Before After Before After Before After
surgery surgery surgery surgery surgery surgery surgery surgery surgery surgery surgery surgery
Peritonitis 0.28 ± 0.01, 0.45 ± 0.02 0.12 ± 0.01, 0.13 ± 0.02 0.21 ± 0.03, 0.29 ± 0.04 2.26 ± 0.16, 2.77 ± 0.18 10.1 ± 0.47, 10.98 ± 0.5 1.39 ± 0.11, 1.56 ± 0.12
patients P <0.05 P <0/01 P <0/05 P >0.5 P >0.5 P <0.05
Donors 0.35 ± 0.05 0.16 ± 0.01 0.33 ± 0.05 2.21 ± 0.08 10.54 ± 0.242 1.66 ± 0.06
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
© 2011 BioMed Central Ltd

P3
Presepsin (sCD14-ST) as a new diagnostic biomarker of sepsis:
development of diagnostic tools using the whole blood
TShozushima
Department of Critical Care Medicine, Iwate Medical University, School of
Medicine, Morika, Japan
Critical Care 2011, 15(Suppl 3):P3 (doi: 10.1186/cc10372)
Introduction CD14 is present in macrophage, monocyte, and
granulocyte cells and their cell membranes, and its soluble fraction
is present in blood and is thought to be produced in association with
infections. It is called the soluble CD14 subtype, and its generic name
is presepsin. Presepsin is a novel marker for the diagnosis of sepsis, and
the results of previous study in which an ELISA kit was used showed
a speci c increase in sepsis in the early stage that also correlated
well with severity. In the present study we developed a new rapid
measurement method for whole blood that use a chemiluminescence
enzyme immunoassay. We assessed the usefulness of presepsin values
in sepsis.
Methods The period of the study was the 10 months from August 2009
to June 2010. The subjects were 41 in-patients, age 62 ± 19 years old,
who had been brought to the Critical Care and Emergency Center of
Iwate Medical University and who ful lled at least two of the diagnostic
criteria for systemic in ammatory response syndrome (SIRS) on
arrival. Blood specimens were collected a total of six times; that is, on
admission, and 12 and 24 hours and 3, 5, and 7 days later. Presepsin
values were measured by immunoassay analyzer (PATHFAST; Mitsubishi
Chemical Medience Corporation, Japan). The sepsis marker PCT, IL-6,
and CRP were also measured for comparison. In addition, 128 healthy
subjects were assessed as controls.
Results The mean presepsin level in the 128 healthy subjects in the
control group was 190 pg/ml. The corresponding presepsin levels were
normal (non-infection), 294.2 ± 121.4 pg/ml; local infection, 721.0 ±
611.3 pg/ml; SIRS, 333.5 ± 130.6 pg/ml; sepsis, 817.9 ± 572.7 pg/ml; and
severe sepsis, 1,992.9 ± 1,509.2 pg/ml; the patients with local infection
or sepsis had signi cantly higher presepsin levels than the patients
who did not have infection as a complication. In addition, the presepsin
levels in SIRS that was not complicated by infection were signi cantly
lower than in sepsis. When we divided the patients into an infection
group and a no infection group and plotted the ROC curves of each
of the markers to compare presepsin with other markers, the results
showed that presepsin was the best.
Conclusion We were able to obtain results similar to those obtained
with the conventional ELISA method, and it was possible to diagnose
sepsis more rapidly and conveniently using the immunoassay analyzer.
P4
Investigation into problems associated with the endotoxin activity
assay
NMatsumoto, GTakahashi, MKojika, YIshibe, STatsuyori, SYasushi,
SKInada, SEndo
Department of Critical Care Medicine, School of Medicine, Iwate Medical
University, Morioka, Japan
Critical Care 2011, 15(Suppl 3):P4 (doi: 10.1186/cc10373)
Introduction Endotoxin activity assay (EAA) levels were compared with
endotoxin levels determined by the turbidimetric kinetic method.
Methods A speci c method for the measurement of endotoxin, in
the blood of patients under various conditions, and the in uence
of steroids on EAA levels and contamination of tubes used for the
measurements were investigated.
Results EAA levels increased in patients with injuries and acute
pancreatitis. EAA levels did not increase in patients infected with Gram-
positive bacteria. Endotoxin levels determined by the turbidimetric
kinetic method did not increase in patients with injuries and acute
pancreatitis and Gram-positive bacteria. When patients with long-
term steroid use developed shock due to infection with Gram-negative
bacteria, EAA levels did not increase but endotoxin levels determined
by the turbidimetric kinetic method increased. EAA levels, but not
endotoxin levels determined by the turbidimetric kinetic method,
were suppressed by giving steroids in vitro. Endotoxin was detected
in the tubes used for the measurements. This was suppressed by the
addition of polymyxin B and anti-factor C antibody. EAA levels tended
to increase immediately after direct hemoperfusion using a polymyxin-
B-immobilized  ber column (PMX-DHP).
Conclusion Our  ndings suggest that the EAA had limitations as a
method to measure endotoxin.
P5
Yale insulin protocol infusion in sepsis patients
SOliveira
Albert Schweitzer and Hospital Bangu Hospital, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 3):P5 (doi: 10.1186/cc10374)
Introduction Tight glycemic control is a major concern in critical care.
The objective of this study in the ICU was to evaluate the e ectiveness
and safety of the Yale insulin infusion protocol in sepsis patients.
Methods A retrospective, before–after cohort study. Selected end-
points were mean blood glucose levels, time to reach the target range
of 100 to 150 mg/dl, percentage of blood glucose in the target range,
and hypoglycemia incidence.
Results Were studied 78 patients: 42 in the control group (CG) and 36
in the protocol group (PG). Bedside blood glucose was measured 3,755
times for a mean value of 134.1 ± 15.4 mg/dl in the PG versus 1,730
times for a mean value of 172.7 ± 33.6 mg/dl in the CG. Blood glucose
values were in the target range 63% and 37% of the times, respectively,
for the PG and the CG (P <0.001). The median time to reach the glucose
target range was 8 hours (range 5 to 17 hours) for the PG and 53 hours
(range 23 to 218 hours) for the CG (P <0.001). The incidence of severe
hypoglycemia reached a statistically signi cant di erence: one patient
in the PG versus four patients in the CG (P <0.01). All patients reached
the target in 72 hours of insulin infusion in the PG while only 29 patients
in the CG reached this target.
Conclusion The Yale insulin infusion protocol was e ective and safe in
sepsis patients admitted to the ICU.
P6
Early vasopressin application in shock
SOliveira
Albert Schweitzer and Hospital Bangu Hospital, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 3):P6 (doi: 10.1186/cc10375)
Introduction Vasopressin is frequently used to maintain blood
pressure in refractory septic shock. We hypothesized that early infusion
of vasopressin compared with norepinephrine would decrease the
mortality rate and severity of septic status.
Methods In this randomized, double-blind study, we assigned patients
who need vasopressors and randomized to receive norepinephrine
(0.05 to 2.0 g/kg/minute) or vasopressin (0.01 to 0.03 U/minute)
with norepinephrine. Both groups had the vasoactive drug infusions
titrated and tapered to maintain a mean blood pressure between 65
and 75mmHg.
Results A total of 387 patients underwent randomization with 191
patients receiving vasopressin and 196 receiving norepinephrine.
There was no signi cant heterogeneity between these two study
groups. There was a signi cant di erence between the vasopressin and
norepinephrine groups in the mortality rate of 14 days (29.3% vs. 36.7%,
respectively, P = 0.05) and 28 days (34% and 42.3%, respectively, P =
0.03); however, in 7-day mortality there were no signi cant di erences
in the overall rates (21.2% vs. 23.9%, respectively; P = 1.1). Also note
a reduction in the incidence of single organ dysfunction (37.7% vs.
49.2%, respectively, P = 0.02) and multiple organ dysfunction using
vasopressin and norepinephrine (17.8% vs. 26%, P = 0.05; P = 0.03). The
length of stay in the ICU was 14 and 17 days (P = 0.29) and the time
of hospitalization was 23 and 28 days (P = 0.11), respectively, in the
vasopressin and norepinephrine groups.
Conclusion Early application of vasopressin reduced mortality rates
in 14 and 28 days as compared with norepinephrine alone, and also a
di erence in incidence of organ dysfunction. This observed di erence
can be attributed to early restoration of tissue perfusion and vascular
smooth muscle responsiveness that directly in uenced patient survival.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S2

P7
Endotoxin removal by hemoperfusion in septic shock
SZielinski
Department of Anesthesia and Intensive Care, Wroclaw University of
Medicine, Wroclaw, Poland
Critical Care 2011, 15(Suppl 3):P7 (doi: 10.1186/cc10376)
Introduction Many symptoms of septic shock are due to the presence
of endotoxin in the bloodstream. The biological activity of endotoxins
is associated with lipopolysaccharide (LPS). LPS induces systemic
in ammatory response and a high level of endotoxin in blood is
associated with worse clinical outcome. Reduction of the level of
circulating endotoxins with hemoperfusion through the  lter with
high a nity for LPS could potentially interrupt the biological cascade
of sepsis. The aim of the study was to evaluate the e ciency of
extracorporeal endotoxin elimination in patients with Gram-negative
septic shock.
Methods The study was conducted at the Department of Anesthes-
iology and Intensive Therapy, Wroclaw Medical University, Poland.
Patients with septic shock, documented or suspected Gram-negative
infection, and with high endotoxin activity (EA >0.6 units) were
eligible for the study. The endotoxin activity in blood was measured
with chemiluminescent activity assay. Based on the enrolment criteria
and EA level, patients were assigned to the conventional treatment
group (Group 1) or the conventional plus hemoperfusion therapy
with LPS adsorber (Alteco Medical AB, Lund, Sweden) group (Group 2).
Hemoperfusion was performed for 2 hours with blood  ow maintained
at 150 ml/minute.
Results Seventeen patients with low EA (0.42 ± 0.14, Group 1) and 12
patients with high EA (0.76 ± 0.13, Group 2) (P <0.05) were included.
There were no signi cant di erences between Group 1 and 2 regarding
age (63 ± 2 and 61 ± 21), APACHE II score (22.7 ± 8.6 and 24.5 ± 7.2),
SOFA score (9.8 ± 3.0 and 11.3 ± 4.1), mean arterial pressure (MAP, 66.2±
8.1 mmHg and 71.5 ± 7.3 mmHg), and PaO2/FiO2 (255 ± 59 and 216 ±
105) at entry to the study. In the hemoperfusion group, nine patients
had Gram-negative and three had Gram-positive infection; seven
patients survived to the 28-day follow-up. High endotoxin activity at
baseline decreased signi cantly 24 hours after hemoperfusion to 0.5±
0.1 (P <0.01) in those who survived, but remained high (0.7 ± 0.1) in
nonsurvivors. At 24 hours after hemoperfusion, MAP signi cantly
increased (78.8 ± 20.8 to 89.2 ± 19.8 mm Hg, P <0.05) and vasopressor
requirements decreased in survivors but not in those who died (MAP,
64.2 ± 9.6 to 71.8 ± 15.3 mm Hg, P = nonsigni cant).
Conclusion Hemoperfusion with LPS adsorber added to standard
treatment improved the hemodynamic status of patients with septic
shock. The chemiluminescence assay for measurement of LPS activity
was a valuable diagnostic tool for rapid detection of endotoxemia.
Acknowledgements The authors declare no con ict of interest
related to this work. The study was supported by the Wroclaw Medical
University. LPS adsorbers were kindly provided by Alteco Medical AB,
Lund, Sweden.
P8
Urinary hepcidin is potentially a marker of systemic infection rather
than in ammation, in the setting of preserved renal function
NJGlassford1, AGSchneider1, GEastwood1, LPeck1, HYoung1,
MWesterman2, VOstland2, RBellomo1
1Department of Intensive Care, Austin Hospital, VIC, Australia;
2IntrinsicLifeSciences LLC, La Jolla, CA, USA
Critical Care 2011, 15(Suppl 3):P8 (doi: 10.1186/cc10377)
Introduction Urinary proteomics have recently identi ed hepcidin, a
key regulator of iron homeostasis, as a potential marker of tubular stress
[1]. It appears to be released in response to situations that predispose
to acute kidney injury (AKI), and greater concentrations of hepcidin
in the blood and in the urine have been associated with reduced risk
of AKI [2]. Catalytic iron is a biologically plausible mechanism for the
development of AKI as a consequence of tubular oxidative stress [3].
The relationship between serum creatinine, urinary hepcidin and CRP
may help de ne whether urinary hepcidin is more likely to re ect
systemic in ammation or renal events. The relationship in septic
patients has not yet been described. Patients with SIRS, oliguria and
a 25 µmol/l increase from baseline creatinine are known to be at an
increased risk of AKI [4]. We sought to determine if hepcidin correlated
more strongly with CRP or creatinine in these patients with a diagnosis
of sepsis and those without.
Methods Patients meeting the inclusion criteria within 48 hours of
admission had their CRP, urinary hepcidin, and serum and urinary
creatinine measured. The strength of the relationship between serum
creatinine or CRP and urinary hepcidin corrected for urinary creatinine
was determined using Spearman’s rank correlation coe cient.
Results We enrolled 103 patients between 31 August 2010 and 17
November 2010; 22 of whom had an APACHE III diagnosis of sepsis.
Serum creatinine only correlated weakly with direct and inverse
urinary hepcidin measurements in septic and nonseptic patients alike.
However, there was a moderately strong correlation between CRP and
urinary hepcidin in septic patients, a relationship not demonstrated in
the nonseptic group (Table 1).
Table 1 (abstract P8). Relationships between hepcidin, creatinine and CRP
Correlation
Nonseptic (n = 81) Septic (n = 22)
Variable Serum Cr CRP Serum Cr CRP
Urinary hepcidin –0.272 0.204 –0.225 0.506
(P = 0.013) (P = 0.064) (P = 0.314) (P = 0.016)
1 / urinary hepcidin 0.287 –0.19 0.225 –0.506
(P = 0.009) (P = 0.087) (P = 0.314) (P = 0.016)
Urinary hepcidin –0.146 0.241 –0.276 0.418
corrected for (P = 0.191) (P = 0.029) (P = 0.227) (P = 0.06)
urinary creatinine
1 / urinary hepcidin 0.158 –0.228 0.276 –0.418
corrected for (P = 0.159) (P = 0.041) (P = 0.227) (P = 0.06)
urinary creatinine
Conclusion Hepcidin is only weakly inversely correlated with serum
creatinine. A stronger relationship exists between hepcidin and CRP in
septic patients, suggesting that hepcidin may primarily be a marker of
infection that is  ltered in the urine when the glomerular  ltration rate
(GFR) is preserved and  ltered in lower amounts when the GFR is lost.
That this relationship is not replicated in nonseptic patients with clinical
evidence of SIRS suggests that the underlying pathophysiological
processes are di erent. Further investigation of the natural history of
AKI and biomarker release is warranted.
References
1. Ho J, Lucy M, Krokhin O, Hayglass K, Pascoe E, Darroch G, Rush D, Nickerson P,
Rigatto C, Reslerova M: Mass spectrometry-based proteomic analysis of
urine in acute kidney injury following cardiopulmonary bypass: a nested
case–control study. Am J Kidney Dis 2009, 53:584-595.
2. Haase M, Bellomo R, Haase-Fielitz A: Novel biomarkers, oxidative stress, and
the role of labile iron toxicity in cardiopulmonary bypass-associated acute
kidney injury. J Am Coll Cardiol 2010, 55:2024-2033.
3. Prowle JR, Westerman M, Bellomo R: Urinary hepcidin: an inverse biomarker
of acute kidney injury after cardiopulmonary bypass? Curr Opin Crit Care
2010, 16:540-544.
4. Ricci Z, Ronco C: Pathogenesis of acute kidney injury during sepsis. Curr
Drug Targets 2009, 10:1179-1183.
P9
Neutrophil gelatinase-associated lipocalin has a stronger
association with serum creatinine than C-reactive protein in
patients without sepsis; this relationship is lost in septic patients
NJGlassford, AGSchneider, GEastwood, LPeck, HYoung, RBellomo
Department of Intensive Care, Austin Hospital, Heidelberg, VIC, Australia
Critical Care 2011, 15(Suppl 3):P9 (doi: 10.1186/cc10378)
Introduction Neutrophil gelatinase-associated lipocalin (NGAL) predicts
the development of acute kidney injury (AKI) amongst critically ill
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S3

patients [1]. Serum and urinary NGAL have been shown to be elevated
in patients with SIRS, sepsis and septic shock [2], and the predictive
ability of NGAL in these patients is not so certain [3]. It is unclear,
however, whether this predictive relationship is due to the fact that
NGAL is produced by neutrophils and is, therefore, a biomarker of
in ammation and infection, or whether NGAL in blood and/or urine
mostly re ects tubular release. It is also unclear if the type of AKI that
develops in SIRS is di erent from that developing in septic patients.
Methods To test these hypotheses, we studied ICU patients with SIRS
and oliguria or a 25 µmol/l increase in serum creatinine. We sought
to determine whether blood and urine NGAL correlated more closely
with CRP or creatinine at the time of enrolment. The strength of the
relationship between serum creatinine or CRP and urine and serum
NGAL, as well as urinary NGAL corrected for urinary creatinine, was
determined using Spearman’s rank correlation coe cient.
Results We recruited 105 patients between 31 August 2010 and 17
November 2010; 22 of these had an APACHE III diagnosis of sepsis.
In nonseptic patients NGAL in blood or urine correlated only weakly
with CRP, but a stronger and statistically signi cant relationship was
observed between serum and/or urine NGAL and serum creatinine. A
similar strength of relationship was observed between creatinine and
NGAL and CRP and NGAL in septic patients, although it failed to reach
signi cance. See Table 1.
Table 1 (abstract P9). Relationships between NGAL, creatinine and CRP in
patients with and without sepsis
Correlation
Sepsis (n = 22) No sepsis (n = 83)
Variable Serum Cr CRP Serum Cr CRP
Urinary NGAL 0.345 0.302 0.391 0.057
(P = 0.116) (P = 0.173) (P <0.001) (P = 0.61)
Urinary NGAL 0.311 0.235 0.514 0.070
corrected for (P = 0.171) (P = 0.305) (P <0.001) (P = 0.532)
urinary creatinine
Serum NGAL 0.244 0.411 0.661 –0.034
(P = 0.274) (P = 0.057) (P <0.001) (P = 0.763)
Conclusion In patients without a diagnosis of sepsis, NGAL is only
weakly correlated with CRP and a stronger relationship is observed
between NGAL and serum creatinine. This suggests that NGAL is
more likely a biomarker of tubular injury or stress than systemic
in ammation in these patients. Similar relationships of moderate
strength are observed between NGAL in blood/urine and both
serum creatinine and CRP in patients with a diagnosis of sepsis. This
suggests that di erent pathophysiological processes may exist in the
genesis of septic AKI when compared with in ammatory AKI. Further
investigation regarding the natural history of AKI and the clinical and
biochemical association of renal biomarkers is warranted.
References
1. Haase M, Bellomo R, Devarajan P, Schlattmann P, Haase-Fielitz A, NGAL Meta-
analysis Investigator Group: Accuracy of neutrophil gelatinase-associated
lipocalin (NGAL) in diagnosis and prognosis in acute kidney injury: a
systematic review and meta-analysis. Am J Kidney Dis 2009, 54:1012-1024.
2. Bagshaw SM, Bennett M, Haase M, Haase-Fielitz A, Egi M, Morimatsu H,
D’amico G, Goldsmith D, Devarajan P, Bellomo R: Plasma and urine
neutrophil gelatinase-associated lipocalin in septic versus non-septic
acute kidney injury in critical illness. Intensive Care Med 2010, 36:452-461.
3. Mårtensson J, Bell M, Oldner A, Xu S, Venge P, Martling CR: Neutrophil
gelatinase-associated lipocalin in adult septic patients with and without
acute kidney injury. Intensive Care Med 2010, 36:1333-1340.
P10
Neutrophil gelatinase-associated lipocalin as a marker of tubular
damage appears to be unrelated to fractional excretion of sodium as
a marker of tubular function in septic patients, with or without AKI
NJGlassford, AGSchneider, GEastwood, LPeck, HYoung, RBellomo
Department of Intensive Care, Austin Hospital, Heidelberg, VIC, Australia
Critical Care 2011, 15(Suppl 3):P10 (doi: 10.1186/cc10379)
Introduction The utility of urinary biochemistry has recently been
challenged [1], while there is emerging evidence that renal biomarkers
may accurately quantify the risk of development of acute kidney injury
(AKI) [2]. Neutrophil gelatinase-associated lipocalin (NGAL) is a marker
of renal tubular damage [3]. Fractional excretion of sodium (FENa)
is a marker of renal tubular function, and is a signi cantly cheaper
investigation [4]. Insults damaging the tubules and resulting in AKI
should both stimulate NGAL production and prevent resorption of
sodium. Given the di erent pathological mechanisms underlying
septic and nonseptic AKI, it is plausible that the relationship between
these variables could be di erent in these two groups of patients [5].
Methods To test this hypothesis, we studied ICU patients developing
SIRS and oliguria or a 25 µmol/l increase in serum creatinine within
48 hours of ICU admission. We sought to determine if a relationship
existed between FENa and NGAL in patients, with and without sepsis,
developing AKI. We measured the serum and urinary NGAL, creatinine
and sodium of patients with SIRS and either oliguria or an increase in
creatinine within 48 hours of admission to a tertiary referral ICU. Point-
of-care creatinine measurements were used to identify the maximum
RIFLE category of AKI developed within the  rst 5 days of admission.
The strength of the relationship between variables was determined
using Spearman’s rank correlation coe cient.
Results We enrolled 93 patients between 31 August 2010 and 17
November 2010; 17 had an APACHE III diagnosis of sepsis. Serum NGAL
and urinary NGAL when corrected for urinary creatinine were found
to correlate moderately well with FENa in patients without sepsis, a
relationship that weakens with the progression of AKI in this group. No
other correlation showed a signi cant relationship (Table 1).
Conclusion The lack of a strong correlation FENa and NGAL in patients
developing RIFLE I and F AKI suggests that changes in NGAL and
changes in sodium resorption occur as a consequence of di erent
stimuli in the pathogenesis of the syndrome. The absence of any
observed relationship between NGAL and FENa in septic patients
suggests a pathological process di erent from that underlying
nonseptic AKI. The small sample size may be a confounding factor.
References
1. Bagshaw SM, Langenberg C, Wan L, May CN, Bellomo R: A systematic review
of urinary  ndings in experimental septic acute renal failure. Crit Care Med
Table 1 (abstract P10). Relationships between NGAL, FENa and AKI
Urinary NGAL
corrected for
Urinary NGAL urinary creatinine Serum NGAL Urine:serum NGAL ratio
FENa n Spearman P value Spearman P value Spearman P value Spearman P value
Nonseptic No AKI 43 0.153 0.328 0.587 <0.0001 0.450 0.002 –0.009 0.953
AKI 33 –0.039 0.830 0.438 0.011 0.258 0.148 –0.235 0.188
RIFLE R–F 14 0.015 0.958 0.235 0.418 –0.077 0.793 –0.068 0.817
Septic No AKI 12 –0.168 0.602 –0.007 0.983 0.232 0.467 –0.427 0.167
RIFLE R–F 5 –0.300 0.624 0.500 0.391 0.112 0.858 –0.400 0.505
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S4

2007, 35:1592-1598.
2. Siew ED, Ware LB, Ikizler TA: Biological markers of acute kidney injury. J Am
Soc Nephrol 2011, 22:810-820.
3. Cruz DN, de Cal M, Garzotto F, Perazella MA, Lentini P, Corradi V, Piccinni P,
Ronco C: Plasma neutrophil gelatinase-associated lipocalin is an early
biomarker for acute kidney injury in an adult ICU population. Intensive Care
Med 2010, 36:444-451.
4. Kanbay M, Kasapoglu B, Perazella MA: Acute tubular necrosis and pre-renal
acute kidney injury: utility of urine microscopy in their evaluation – a
systematic review. Int Urol Nephrol 2010, 42:425-433.
5. Ishikawa K, May CN, Gobe G, Langenberg C, Bellomo R: Pathophysiology of
septic acute kidney injury: a di erent view of tubular injury. Contrib
Nephrol 2010, 165:18-27.
P11
Renal biomarkers are less useful at predicting acute kidney injury in
patients with sepsis than those without
NGlassford1, ASchneider1, GEastwood1, LPeck1, HYoung1,
MWesterman2, VOstland2, RBellomo1
1Department of Intensive Care, Austin Hospital, Heidelberg, VIC, Australia;
2Intrinsic LifeSciences LLC, La Jolla, CA, USA
Critical Care 2011, 15(Suppl 3):P11 (doi: 10.1186/cc10380)
Introduction Multiple biomarkers have been proposed for identifying
patients at risk of developing the syndrome of acute kidney injury
(AKI) [1]. These biomarkers include urine and serum NGAL, and urinary
hepcidin. The pathophysiology of AKI in sepsis appears to be primarily
mediated by immunological, toxic and in ammatory factors as
opposed to renal ischaemia [2]. Di erent aetiologies of AKI are likely to
lead to di erential release of serum and urinary biomarkers. We sought
to determine if the predictive ability of several renal biomarkers for
predicting AKI varied in the presence of sepsis in the context of routine
ICU practice.
Methods We measured serum and urinary NGAL and urinary hepcidin
in patients admitted to the ICU of a tertiary referral hospital with SIRS
and either oliguria or a 25 µmol/l serum creatinine increase within
48hours of admission. We used point-of-care creatinine measurements
to identify the maximum RIFLE category of AKI within the  rst 5 days of
enrolment. We corrected both urinary biomarkers for urinary creatinine.
We calculated the reciprocal of hepcidin measurement and noted if
serum NGAL was greater than the upper limit of normal (149 ng/ml).
We derived the area under the curve (AUC) for the receiver operating
characteristic curve (ROC) for all biomarkers.
Results Between 31 August 2010 and 17 November 2010, we enrolled
92 patients; 17 of these patients had APACHE II diagnoses of sepsis. In
patients with a diagnosis of sepsis, the predictive ability of all of the
biomarkers measured was worse than in those without (Table 1).
Conclusion Although the sample size is limited, there is a marked
di erence in the predictive ability of the measured biomarkers to predict
AKI between septic and nonseptic patients. All patients admitted met
the criteria for a diagnosis of SIRS, suggesting that in ammation and
sepsis contribute to the development of AKI via di erent pathways. The
ability of these biomarkers to predict AKI in patients with a diagnosis
of sepsis in our cohort is limited. Further investigation is needed into
whether the combination of speci c biomarker patterns and clinical
features can better identify patients at risk, particularly in the setting
of sepsis. In addition, further work examining the relationship between
the various biomarkers and the aetiology and natural history of AKI is
required.
References
1. Ho E, Fard A, Maisel A: Evolving use of biomarkers for kidney injury in acute
care settings. Curr Opin Crit Care 2010, 16:399-407.
2. Ishikawa K, May CN, Gobe G, Langenberg C, Bellomo R: Pathophysiology of
septic acute kidney injury: a di erent view of tubular injury. Contrib
Nephrol 2010, 165:18-27.
P12
Cancer patients with sepsis admitted to a specialized onco-medical
ICU: incidence, ICU course and outcome
DJuneja, OSingh, YJaveri, PNasa, PKumar
Department of Critical Care Medicine, Max Super Speciality Hospital, Saket,
New Delhi, India
Critical Care 2011, 15(Suppl 3):P12 (doi: 10.1186/cc10381)
Introduction While cancer patients are known to be at higher risk for
infection and subsequent complications, there is a scarcity of data
regarding incidence and outcome of septic cancer patients admitted
to ICUs. Hence, we aimed to assess the incidence of cancer patients
admitted with sepsis to an onco-medical ICU and study their ICU
course and outcome.
Methods Data were collected prospectively from all cancer patients
admitted to a specialized onco-medical ICU of a tertiary care hospital,
over a period of 6 months. Sepsis was de ned as per the international
guidelines. Cancer patients were divided into two groups on the basis
of presence of sepsis on ICU admission and compared with regard to
their need for organ support, length of ICU stay and mortality. Severity
of illness was assessed by APACHE II score and organ failure by SOFA
score. Qualitative data were analyzed using the chi-squared test or
Fisher exact test as appropriate and quantitative data were analyzed
using Student’s t test. P <0.05 was considered signi cant.
Results Out of 104 cancer patients admitted during the study period, 43
(41.3%) patients were admitted with sepsis. Even though there was no
di erence in age (P = 0.13), sex (P = 0.382) and presence of metastasis
(P = 0.314) among the septic and nonseptic groups of patients, septic
patients had comparatively higher admission APACHE II (21.4 ± 7 vs.
18.21 ± 6.9; P = 0.023) and SOFA (6.4 ± 3.8 vs. 4.56 ± 3; P = 0.007) scores,
required invasive mechanical ventilation (65.1% vs. 19.7%; P = 0.000)
and vasopressor support (74.4% vs. 19.7%; P = 0.000) more often, had
a longer ICU stay (10.77 ± 8.4 vs. 7.44 ± 5.5; P = 0.017) and had a higher
ICU mortality (62.8% vs. 18%; P = 0.000). The odds ratio and relative risk
Table 1 (abstract P11). AUC ROC for the prediction of AKI
ROC AUC
RIFLE R, I or F RIFLE I or F
Septic Nonseptic Septic Nonseptic
Test result variable Area SE Area SE Area SE Area SE
Urinary NGAL 0.367 0.136 0.561 0.068 0.367 0.136 0.633 0.090
Urinary NGAL corrected for urinary creatinine 0.417 0.136 0.578 0.066 0.417 0.136 0.670 0.082
Serum NGAL 0.375 0.162 0.639 0.065 0.375 0.162 0.685 0.087
Serum NGAL positivity 0.492 0.158 0.611 0.066 0.492 0.158 0.674 0.082
Urine:serum NGAL ratio 0.483 0.140 0.498 0.068 0.483 0.140 0.543 0.081
1 / urinary hepcidin 0.508 0.153 0.624 0.066 0.508 0.153 0.611 0.080
1 / urinary hepcidin corrected for urinary creatinine 0.483 0.156 0.598 0.067 0.483 0.156 0.578 0.083
SE, standard error.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S5

of death for septic cancer patients were 7.67 (95% CI = 3.121 to 18.85)
and 3.482 (95% CI = 1.945 to 6.234).
Conclusion A signi cant proportion of cancer patients are admitted
with evidence of sepsis. These patients were generally sicker, required
more intensive organ support, and had a longer ICU stay and a higher
ICU mortality than those cancer patients admitted to ICU with other
acute problems.
P13
Use of plasma protein fraction in patients with septic shock
admitted to the ICU
DJuneja, OSingh, PNasa, YJaveri, RDang
Department of Critical Care Medicine, Max Super Speciality Hospital, Saket,
New Delhi, India
Critical Care 2011, 15(Suppl 3):P13 (doi: 10.1186/cc10382)
Introduction Certain colloids like albumin and plasma protein fraction
(PPF) have been derived from human plasma and they are used as plasma
expanders to treat patients with shock. PPF, which more closely resembles
plasma in its constituents, contains albumin plus α and β globulins. We
conducted this study to assess the e ect of PPF on need for vasopressors,
organ support and ICU mortality in patients with septic shock.
Methods A retrospective study was conducted and data were collected
from the records of patients admitted to a 16-bed neuro and medical
ICU over a 1.5-year period. All adult patients admitted with septic
shock and requiring vasopressor support (for more than 6 hours) in
spite of aggressive  uid resuscitation were enrolled. Patients who were
transferred from some other ICU or ward and those who developed
shock during their ICU course were excluded from the analysis. Patients
were divided into two groups: patients in whom PPF was used along
with resuscitative  uids comprised the study group, whereas others
formed the control group. Patients in these groups were compared
according to need for organ support, ICU mortality and time taken
to stop vasopressor agents. PPF (Plasmanate®) was administered
in a protocolized way at the dosage of 10 to 20 ml/hour for the  rst
48hours. Development of any complication like allergy or hypotension
associated with PPF was also noted.
Results There was no signi cant di erence in the baseline characteristics of
patients in both groups in terms of age (P = 0.154), sex (P = 0.479), severity
of illness (APACHE II score, P = 0.356), and presence of organ failure (SOFA
score, P = 0.105). Among the outcome parameters there was no signi cant
di erence in terms of need for renal support (P = 0.814), mechanical
ventilation (P = 0.776), ICU stay (P = 0.122), hospital stay (P = 0.054) and
ICU mortality (P = 0.091). However, there was a signi cant di erence in
time taken to stop the vasopressors (P = 0.030) (Table 1). There were no
incidences of any complications or side e ects in any group.
Conclusion PPF may be used safely and e ectively for initial resuscitation
of patients with septic shock requiring vasopressor support. It may
lead to early termination of vasopressor support; however, it did not
translate to lesser need for organ support or reduced ICU mortality in
our patient cohort. To demonstrate such bene ts, larger multicenter
trials are warranted.
P14
ICU scoring systems: which one to use in patients with sepsis?
DJuneja, OSingh, PNasa, YJaveri, MMathur, RDang
Department of Critical Care Medicine, Max Super Speciality Hospital, Saket,
New Delhi, India
Critical Care 2011, 15(Suppl 3):P14 (doi: 10.1186/cc10383)
Introduction Disease-severity scoring systems have been developed
for strati cation of ICU patients. These systems have been tested
and validated in various general medical and surgical ICU patients.
However, the validity and e cacy of these systems, especially the
newer generation, has not been assessed in patients with sepsis, which
is the commonest indication for admission to a medical ICU. Hence, we
conducted this study to assess the performance of various ICU scoring
systems – Acute Physiology and Chronic Health Evaluation (APACHE) II,
III, IV; Simpli ed Acute Physiology Score (SAPS) II, III; Mortality Prediction
Model (MPM) II0, III0; and Sequential Organ Failure Assessment (SOFA)
scores – in septic patients admitted to a medical ICU.
Methods A prospective, observational study was conducted in a tertiary
care medical ICU and consecutive patients ful lling the diagnostic
criteria for sepsis during the  rst 24 hours of ICU admission were
included over a 2-year period. Data related to patient demographics
and that required to compute various scores were recorded. Predicted
mortality was calculated using original regression formulas. The
standardized mortality ratio (SMR) was computed for mortality
prediction. Calibration was assessed by calculating the Lemeshow–
Hosmer goodness-of- t C-statistic. Discrimination was assessed
by calculating the area under the receiver operating characteristic
(AUROC) curves. ICU mortality was the primary outcome measure.
Table 1 (abstract P13). Comparison between patient characteristics and ICU
course among patients in control and PPF groups
Parameter Control group PPF group
of interest (n = 87) (n = 99) P value
Age (years) 64.11 ± 15.4 67.28 ± 14.8 0.154
Sex, male 50 (57.5%) 63 (63.6%) 0.479
APACHE II score 20.69 ± 6.2 21.64 ± 7.6 0.356
PDR 39.51 ± 19.2 42.47 ± 22.5 0.341
SOFA score 9.53 ± 3.5 10.38 ± 3.7 0.105
RBC transfusions 32 (36.8%) 33 (33.3%) 0.735
Renal support 31 (35.6%) 38 (38.4%) 0.814
Mechanical ventilation 50 (57.5%) 60 (60.1%) 0.776
Time taken to stop 70.69 ± 55.2 51.29 ± 64.5 0.030*
vasopressors (hours)
ICU stay (days) 10.38 ± 11.4 13.42 ± 14.8 0.122
Hospital stay (days) 12.91 ± 11.8 16.83 ± 15.3 0.054
ICU mortality 34 (39.1%) 52 (52.5%) 0.091
*P <0.05. Bold text indicates statistically signi cant.
Table 1 (abstract P14). Comparison of the actual and predicted mortality
rates for the various scoring systems
Variable Actual mortality Predicted mortality SMR
APACHE II 0.244 0.296 0.824
APACHE IV 0.244 0.206 1.18
SAPS II 0.244 0.297 0.822
SAPS III 0.244 0.249 0.98
MPM II
0
0.244 0.314 0.777
MPM III
0
0.244
0.216 1.13
Table 2 (abstract P14). Area under the curve for predicting ICU mortality for
various scoring systems
Scoring Sensitivity Speci city
system AUC 95% CI Cut-o (%) (%)
APACHE II 0.880 0.845 to 0.914 >18.5 86.9 75.8
APACHE III 0.880 0.847 to 0.914 >63.5 82.2 70.4
APACHE IV 0.882 0.848 to 0.916 >17.7 83.2 74
SAPS II 0.849 0.808 to 0.890 >41.5 82.2 70.4
SAPS III 0.873 0.838 to 0.907 >53.5 85 72.2
MPM II
0
0.849 0.807 to 0.891 >30.1 77.6 76.1
MPM III
0
0.872 0.835 to 0.909 >18.7 81.3 77.9
SOFA 0.889 0.857 to 0.922 >5.5 86.9 79.2
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S6

Results Data were analyzed for 438 septic patients. The mean age of
patients was 64.5 ± 16.3 years and 301 (68.7%) were male. The mean
ICU and hospital length of stay was 6.39 ± 9.7 and 9.99 ± 10.5 days,
respectively. The observed ICU mortality was 107/438 (24.4%). Mortality
predicted by SAPS III score was closest to that of actual mortality with
a SMR of 0.98 followed by that of MPM III0 (SMR – 1.13) and APACHE IV
(SMR – 1.18) scores (Table 1). APACHE IV (χ2 = 4.416; P = 0.818) had the
best calibration followed by SAPS II (χ2 = 6.073; P = 0.639) and SAPS
III scores (χ2 = 6.538; P = 0.587). There was no statistically signi cant
di erence between the AUROCs of these scores; SOFA (AUROC = 0.
0.889) performed the best followed closely by APACHE IV (AUROC =
0.882) and APACHE III (AUROC = 0.880) scores (Table 2).
Conclusion Overall, the newer generation of scoring systems performed
better than their older counterparts and was more accurate. Older
scoring systems had a tendency to overpredict mortality. However, all
the scores tested had good e cacy and the di erence in e cacy was
not statistically signi cant.
P15
Comparison of the value of plasma and urine cystatin-C and
neutrophil gelatinase-associated lipocalin levels in prediction of
acute kidney injury in sepsis
GGursel
Department of Pulmonary Critical Care Medicine, Gazi University School of
Medicine, Ankara, Turkey
Critical Care 2011, 15(Suppl 3):P15 (doi: 10.1186/cc10384)
Introduction The aim was to study the impact of in ammation/
sepsis on the concentrations of cystatin-C and neutrophil gelatinase-
associated lipocalin (NGAL) in plasma and urine in adult ICU patients
and to estimate the predictive properties of cystatin-C and NGAL in
plasma and urine for early detection of acute kidney injury (AKI) in
patients with sepsis.
Methods The RIFLE class for AKI was calculated daily, while plasma and
urinary Cys-C and NGAL were determined on days 0 and alternate days
until ICU discharge. Test characteristics were calculated to assess the
diagnostic performance of urinary and plasma Cys-C and NGAL. The
diagnostic and predictive performances of the markers were assessed
from the area under the receiver operator characteristic curve (AUC).
Results One hundred and twenty-eight patients were studied, and
three groups were de ned: normal (n = 41); sepsis (n = 45); and sepsis
and AKI (n = 42). AUCs for diagnosis of AKI using plasma and uCys-C
were as follows: 0.89 (P <0.0001) and 0.91 (P <0.0001) Cut-o points for
AKI for plasma and uCys-C were 1.7 mg/l (sensitivity: 83%, speci city:
77%)and 0.11 mg/l (sensitivity = 92%, speci city = 80%), respectively.
Urinary NGAL showed fair discrimination for AKI diagnosis (AUC = 0.85).
Although plasma NGAL performed less well (AUC = 0.58).
Conclusion Plasma and urinary Cys-C are useful markers in predicting
AKI in sepsis. pNGAL is raised in patients with sepsis, and should be
used with caution as a marker of AKI in ICU patients with sepsis. uNGAL
is more useful in predicting AKI as the levels are not elevated in septic
patients without AKI.
P16
Clinical and biological e ects of high-dose sodium selenite,
continuously administered in septic shock
XForceville1, DVitoux2, WWasowicz3, MDehoux4, PVanAntwerpeen5,
DAnnane6, EPlouvier7, ABoutten4, JGromadzinska3, BLaviolle8,
ACombes1, EBellissant8
1Réanimation Polyvalente, CH de Meaux, France; 2Biochimie A, CHUStLouis,
Paris, France; 3Toxicology and Carcinogenesis, Nofer Institute, Lodz,
Poland; 4Biochimie métabolique et cellulaire, CHU Bichat, Paris, France;
5Chimie pharmaceutique organique, Université Libre de Bruxelles, Belgium;
6Réanimation, CHU Poincaré, Garches, France; 7Biochimie CH de Meaux,
France; 8Recherche – Santé Publique, CIC Inserm 0203, CHU Pontchaillou,
Rennes, France
Critical Care 2011, 15(Suppl 3):P16 (doi: 10.1186/cc10385)
Introduction Sodium selenite (Na2SeO3) has been proposed as an
early treatment of septic shock with discrepant results [1-3]. Bene cial
action is mainly believed through improvement of major antioxidant
selenoenzymes, but could on the contrary be related to a therapeutic
oxidant action reducing activity of hyperactivated circulating
phagocytic cells [4]. It has been suggested that the absence of bene cial
e ect of high-dose Na2SeO3 continuously administered [2] might be
related to toxicity, especially on the lung, of too much selenium (Se) as
mentioned in recent parenteral nutrition guidelines in intensive care
[5]. On additional clinical and biological data, our purpose was to assess
if there was argument for Na2SeO3 toxicity, especially on the lung, under
continuous administration of high-dose Na2SeO3 in the SERENITE study.
Methods In a randomized, double-blind multicenter study performed
in 60 septic shock patients [2], the e cacy and tolerance of Na2SeO3
(4 mg Se on day 1 (D1), then 1 mg/day during 9 days or placebo)
were evaluated on all components of the SOFA score measured daily,
infection rate, and plasma Se, selenoprotein-P (Sel-P), glutathione
peroxidase (GPx), lipid peroxidation, cytokines, and procalcitonin
measured at D0, D4, D7, D10 and D14.
Results No deleterious e ect of Na2SeO3 especially on the lung was
observed for any clinical or biological variables. PaO2/FiO2 was strictly
identical between groups (Table 1). As compared with placebo, mean
time occurrences of infections were delayed in the treated group (18±
24 days vs. 34 ± 28 days, respectively; P <0.0001). Plasma Se, Sel-P and
GPx concentrations were increased at D4 in the treated group, achieving
the high reference value for the plasma Se concentration (Figure 1).
Conclusion Continuous administration of high doses of Na2SeO3 (4 mg
Se D1) did not induce any deleterious e ect in septic shock patients.
We did not observe a bene cial e ect, contrasting with a comparable
study administering Na2SeO3 in bolus, potentially more toxic [1]. In
agreement with results obtained on a peritonitis sheep model [6], our
data support a therapeutic oxidant action of Na2SeO3, opening a new
 eld in septic shock treatment based on oxidant selenocompounds.
Table 1 (abstract P16). PaO
2
/FiO
2
according to group and time
Baseline
(D0) D2 D3 D4 D7 D10 D14
Na
2
SeO
3
20 ± 15 23 ± 12 24 ± 10 25 ± 14 26 ± 11 30 ± 13 28 ± 15
Placebo 22 ± 13
21 ± 11 25 ± 13 26 ± 11 28 ± 14 33 ± 17 36 ± 17
P value 0.60 0.37 0.92 0.81 0.65 0.52 0.13
Data presented as mean ± SD (KPa).
Figure 1 (abstract P16). Plasma Se concentration according to groups and time. Data presented as mean ± SD (μmol/l). Reference value for plasma Se
concentration: 1 ± 0.15 μmol/l. Treated group indicated in blue and nontreated in green.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S7

Acknowledgements The authors thank all the investigators,
biochemists, pharmacists and clinical research team involved in
the SERENITE Study, the Minister of Health for  nancing, and Meaux
Hospital as promotor. XF is the co-inventor with DV of patent FR 98
10889, PCT N°FR 99/ 02.66 (delivered: US 6,844,012 B1, Au 760 534;
EP 1107767), and has ownership of the corresponding patent. XF is
the main shareholder of a small start-up named SÉRÉNITÉ-Forceville
devoted to early diagnosis and treatment of septic shock especially by
selenocompounds.
References
1. Angstwurm MW, Engelmann L, Zimmermann T, Lehmann C, Spes CH, Abel P,
Strauss R, Meier-Hellmann A, Insel R, Radke J, et al.: Selenium in Intensive
Care (SIC): results of a prospective randomized, placebo-controlled,
multiple-center study in patients with severe systemic in ammatory
response syndrome, sepsis, and septic shock. Crit Care Med 2007,
35:118-126.
2. Forceville X, Laviolle B, Annane D, Vitoux D, Bleichner G, Korach JM, Cantais E,
Georges H, Soubirou JL, Combes A, et al.: E ects of high doses of selenium,
as sodium selenite, in septic shock: a placebo-controlled, randomized,
double-blind, phase II study. Crit Care 2007, 11:R73.
3. Vincent JL, Forceville X: Critically elucidating the role of selenium. Curr Opin
Anesthesiol 2008, 21:148-154.
4. Forceville X: Seleno-enzymes and seleno-compounds: the two faces of
selenium. Crit Care 2006, 10:180.
5. Singer P, Berger MM, Van den Berghe G, Biolo G, Calder P, Forbes A, Gri ths R,
Kreyman G, Leverve X, Pichard C, et al.: ESPEN Guidelines on parenteral
nutrition: intensive care. Clin Nutr 2009, 28:387-400.
6. Wang Z, Forceville X, Van Antwerpen P, Piagnerelli M, Ahishakiye D, Macours P,
De Backer D, Neve J, Vincent JL: A large-bolus injection, but not continuous
infusion of sodium selenite improves outcome in peritonitis. Shock 2009,
32:140-146.
P17
Temperature management of patients with sepsis and in ammation
in Australian and New Zealand ICUs: a point prevalence study
NEHammond1,2, MKSaxena1,2,3, CTaylor1,4, ISeppelt4,5, PGlass1,
JMyburgh1,2,3
1The George Institute for Global Health, Sydney, NSW, Australia; 3St George
Hospital, Sydney, NSW, Australia; 2St George Clinical School, University of NSW,
Sydney, NSW, Australia; 4Sydney Medical School, University of Sydney, NSW,
Australia
Critical Care 2011, 15(Suppl 3):P17 (doi: 10.1186/cc10386)
Introduction The use of pharmacological and physical antipyretic
therapies to reduce fever in febrile patients is common in hospital
settings. Actual evidence on the frequency of antipyretic use is
limited, however, both in general hospital populations and, more
speci cally, in adult intensive care [1-3]. We undertook a prospective
point prevalence study with the aim of identifying the prevalence of
physical and pharmacological antipyretic therapies in intensive care
patients with sepsis and in ammation. We also recorded the indication
for antipyretic therapies, temperature measurement site, and mean
temperatures on the study day.
Methods We conducted a single-day observational point prevalence
study in 38 ICUs in Australia and New Zealand. All patients in participating
ICUs at a 10:00 am census point were studied. Data were collected for the
24-hour study day that included the 10:00 am time point.
Results We studied 506 patients, with a mean age 59 years (SD =
17 years); 65% male; APACHE II score 17 (SD = 7), 28-day mortality
14%. Eighty percent of the ICU admissions were unplanned. Of the
506 patients, 311 patients had sepsis and in ammation with mean
peak temperature of 37.3°C (SD = 0.8°C). Of these, 35% (n = 100/311)
had a mean peak temperature above 38°C. In the 24-hour period,
paracetamol was used 50% (n = 152/311) of the time, nonsteroidal
anti-in ammatory drugs (NSAIDs) 0.6% (n = 2/311) and physical
cooling 1% (n = 3/311) (Figure 1). Of patients that had an indication for
paracetamol recorded, 64% was for pain (n = 92/152), 18% for both pain
and fever (n = 26/152); and 10% for fever alone (n = 14/152) (Figure 2).
Sixty-four percent (n = 92/152) of the patients who had paracetamol
were prescribed regular paracetamol and 36% (n = 51/143) had a PRN
order. Of the 40 patients who received paracetamol for an indication of
fever, the mean peak temperature was 38.3°C (SD = 0.8°C; range 36.1
to 40.2°C). Of the three patients who received physical cooling, the
mean peak temperature was 39.2°C (SD = 0.9°C; range 38.5 to 40.2°C).
Temperature measurement sites were mainly noncore (n = 251/311)
with axillary (37%; n = 116/311) and tympanic (35%; n = 110/311) most
common (Figure 3).
Conclusion This point prevalence study of intensive care patients with
sepsis and in ammation identi ed pharmacological antipyretics are
used regularly for pain management rather than fever management,
with paracetamol the most common therapy. The use of physical
cooling was rare, and noncore temperature measurements were
common. These results are important in understanding current
temperature management practice in intensive care and will aid in
designing future clinical trials on the subject.
Acknowledgements This study was undertaken as part of the
Australian and New Zealand Intensive Care Society (ANZICS) Clinical
Trials Group (CTG) Point Prevalence Program. The authors would like to
thank all participating sites.
References
1. Lamb SA, Henry RL: Paracetamol pro re nata orders: an audi t. J Paediatr
Child Health 2004, 40:213-216.
2. Isaacs S, Axelrod P, Lorber B: Antipyretic orders in a uni versity hospital. Am J
Med 1990, 88:31-35.
3. Albertson T, Walker V, Stebbins M, Ashton E, Owen K, Sutte r M: A population
study of the frequency of high-dose acetaminophen prescribing and
dispensing. Ann Pharmacother 2010, 44:1191-1195.
Figure 1 (abstract P17). Type of antipyretic and physical cooling used on the study day (n = 311).
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S8

P18
A survey of fever management in febrile intensive care patients
without neurological injury
MKSaxena1,2,3, NEHammond1, CTaylor1,4, PYoung5, MCReade6,
RBellomo6, JMyburgh1,2,3
1The George Institute for Global Health, Sydney, NSW, Australia; 2St George
Hospital, Sydney, NSW, Australia; 3St George Clinical School, University of
New South Wales, Sydney, NSW, Australia; 4Sydney Medical School, University
of Sydney, NSW, Australia; 5Medical Research Institute of New Zealand,
Wellington, New Zealand; 6Austin Hospital & University of Melbourne, VIC,
Australia
Critical Care 2011, 15(Suppl 3):P18 (doi: 10.1186/cc10387)
Introduction Fever is a common observation during critical illness
[1,2] and may be due to many possible causes such as infection,
sterile in ammation and neurological injury. Clinical trials of fever
management lack su cient methodological quality to answer the
question of whether attempts at reduction in temperature improves
patient-centred outcomes in patients with sepsis, in ammation or
neurological injury [3-7]. We undertook a survey to describe the
attitudes of critical care clinicians in Australia and New Zealand towards
fever management in critically ill patients without neurological injury
or hyperthermic syndromes.
Methods An online scenario-based questionnaire survey was
distributed to medical and nursing members of the Australian and New
Zealand Intensive Care Society Clinical Trials Group (ANZICS-CTG) and
their intensive care colleagues. Main outcome measures: the choice
of drug and preferred threshold temperature for intervention with
antipyretics in clinical practice and in a clinical trial.
Results There were 588 email invitations distributed through the
ANZICS-CTG and Research Coordinator mailing list. Four hundred and
forty-seven responses were received from 308 nurses (69%), 137 doctors
(31%), and two others (0.5%). The majority of respondents having
more than 8 years of experience (62%) worked in mixed medical and
surgical units (84%) in a metropolitan or tertiary hospital setting (77%).
The primary  ndings of our survey suggest that fever management is
highly variable. Most clinicians administer an intervention to reduce
temperature at or below 39°C (Figure 1); and initially use a combination
of both pharmacological and physical interventions, with an increase
in intensity of physical interventions for persistent fever (Figure2).
There were di erences between the professions, with doctors
choosing higher temperature thresholds for intervention and nurses
generally using more physical cooling (Figure 1 and Table 1); fourthly,
temperature thresholds for a clinical trial were 39.0°C (SD = 0.7°C) for a
permissive strategy and 38.0°C (SD = 0.75°C) for an intensive strategy;
 nally, there was broad support for a clinical trial of fever management.
Conclusion This survey suggests there is considerable clinical
variability in fever management in patients with sepsis and without
neurological injury or hyperthermic syndromes. At present, no
particular management strategy is known to be superior to any
other and it remains possible that current practice may be harming
substantial numbers of patients. A temperature threshold of up to 40°C
may be acceptable to clinicians for the design of a future randomized
controlled trial. Further observational data may be informative for the
design of such clinical trials.
Figure 2 (abstract P17). Indication for paracetamol administration (n = 152).
Figure 3 (abstract P17). Temperature measurement site for patients with sepsis and in ammation (n = 311). PAC, pulmonary artery catheter.
Table 1 (abstract P18). Preference of  rst-line and second-line interventional
category of antipyretic by profession
First line (n = 418) Second line (n = 409)
Nurse Doctor P Nurse Doctor P
(%) (%) value (%) (%) value
Pharmacological only 23 40 0.0002 5 5 0.87
Physical only 13 5 0.0087 58 62 0.36
Pharmacological and physical 64 55 0.077 38 33 0.39
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S9

References
1. Laupland KB, Shahpori R, Kirkpatrick AW, Ross T, Gregson DB, Stelfox HT:
Occurrence and outcome of fever in critically ill adults. Crit Care Med 2008,
36:1531-1535.
2. Ryan M, Levy M. Clinical review: fever in intensive care unit patients. Crit
Care 2003, 7:221-225.
3. Bernard GR, Wheeler AP, Russell JA, et al.: The e ects of ibuprofen on the
physiology and survival of patients with sepsis. The Ibuprofen in Sepsis
Study Group. N Engl J Med 1997, 336:912-918.
4. Hammond NE, Boyle M: Pharmacological versus non-pharmacological
antipyretic treatments in febrile critically ill patients: a systematic review
and meta-analysis. Aust Crit Care 2011, 24:4-17.
5. Schulman CI, Namias N, Doherty J, et al.: The e ect of antipyre tic therapy
upon outcomes in critically ill patients: a randomized, prospective study.
Surg Infect 2005, 6:369-375.
6. Saxena M, Andrews P, Cheng A: Modest cooling therapies (35°C to 37.5°C)
for traumatic brain injury. Cochrane Database Syst Rev 2008, 3:CD006811.
7. Je eries S, Weatherall M, Young P, et al.: The e ect of antipyre tic medications
on mortality in critically ill patients with infection: a systematic review and
meta-analysis. Crit Care Resusc 2011, 13:125-131.
P19
Assessment of the usefulness of presepsin (soluble CD14 subtype)
in s eptic patients
TNishida1, HIshikura1, AMurai1, YIrie1, TUme mura1, TKamitani1, SEndo2
1Depar tment of Emer gency and Critical Care Medicin e, Fukuoka Un iversi ty
Hospital, Fukuoka, Japan; 2Department of Critical Care Medicine, School of
Medicine, Iwate Medi cal University, Iwate, Japan
Critical Care 2011, 15(Suppl 3):P19 (doi: 10.1186/cc10388)
Introduction Sepsis is a life-threatening condition characterized by a
whole-body in ammatory state. The early diagnosis and treatments
of sepsis will improve the outcome of patients. The aim of this study
was to compare blood levels of presepsin (renamed from soluble CD14
subtype), procalcitonin (PCT), IL-6 and C-reactive protein (CRP) and to
investigate the most useful biomarker for early diagnosis of sepsis.
Methods A single-center, prospective, observational study. Patients
who had one or more systemic in ammatory response syndrome
criteria were included in this study. The blood samples for measuring
the biomarkers were collected and the severity of sepsis was evaluated
at the time of admission and every other day for a week. Fort y-two
patients were enrolled for the prospective study from June 2010 to
December 2010.
Results Twenty-three patients were diagnosed with sepsis and 19
patients were without sepsis. In the receiver operating characteristics
(ROC) curve analysis, the area under the curve (AUC) to distinguish
sepsis was the largest for presepsin (0.930) followed by IL-6 (0.896), PC T
(0.854) and CRP (0.840). Presepsin may be able to discriminate between
patient groups with or without sepsis. From the ROC curve analysis, a
cut-o value of presepsin was 929 pg/ml with sensi tivity and speci city
of 76% and 81%, respectively, with odds ratios and 95% CIs of 0.996
(0.992 to 0.998) and 3.376 (1.497 to 6.094). And the presepsin values
were signi cantly higher in the patients with the more severe septic
condition (for example, sepsis, severe sepsis, septic shock). In addition,
a signi cant correlation was found between the Sepsis-related Organ
Failure Assessment scores and the presepsin values (r2 = 0.320; P =
0.0003). But there was a no signi cant correlation between APACHE II
scores and the presepsin values.
Conclusion In this study, presepsin is the most valuable predictor about
sepsis compared with PCT, IL-6 and CRP. Moreover, the results suggest
that presepsin values can serve as a parameter that closely re ects the
pathology. So we strongly suggest that the presepsin will be not only a
very useful new biomarker of the diagnosis of sepsis, but also useful for
monitoring the severity of the disease in the near future.
P20
An audit of awareness about maternal sepsis in a UK district general
hospital
CDonohue, ETeh, MKitching
Department of Anaesthesia, Lister Hospital, Stevenage, UK
Critical Care 2011, 15(Suppl 3):P20 (doi: 10.1186/cc10389)
Introduction The UK national body, which reviews maternal mortality
(Centre for Maternal and Child Enquiries (CMACE)), has recently
published their 2006 to 2008 report. This highlighted an increase in
maternal sepsis, making it the leading cause of direct death amongst
peripartum women in the UK (26 out of a total 107 direct deaths)
[1]. The Surviving Sepsis Campaign (SSC) published updated sepsis
resuscitation and management bundles in 2008 [2]. We decided to
audit awareness about sepsis amongst sta caring for perpartum
women.
Methods A questionnaire was devised and distributed to midwives,
obstetricians and anaesthetists. This asked the criteria for the systemic
in ammatory response syndrome (SIRS), common sites of maternal
sepsis, the initial duties of care (‘sepsis six’ resuscitation bundle:
delivery of oxygen, intravenous  uids, intravenous antibiotics, taking
of blood cultures, measurement of plasma haemoglobin, lactate and
urine output) and recognition and management of severe sepsis.
Results There was a 98% response rate with 41 completed question-
naires returned, 15 from midwives, 13 from obstetricians and 13 from
anaesthetists. We found that awareness was suboptimal within all
groups. Of the six criteria for SIRS, suggested by the SSC, two criteria
(altered consciousness and hyperglycaemia) were poorly identi ed
and few responders were aware that two or more criteria indicated SIRS
(Figure 1). Most healthcare professionals correctly identi ed genital
tract infection as the leading source of maternal sepsis. The majority
of responders had not heard of the ‘sepsis six’ (Figure 2). Out of the
initial duties of care, delivery of oxygen and monitoring urine output
were poorly identi ed. Respondents were not con dent in identifying
features of severe sepsis and with the exception of hypotension despite
 uids, other markers of end organ dysfunction were underreported.
Conclusion I t is all of our responsibilities to focus e orts on the emerging
threat of maternal sepsis as highlighted by CMACE. Historically, we
have seen a signi cant improvement in maternal mortality rates when
speci c interventions have targeted those issues raised in previous
Figure 1 (abstract P18). Thresholds for initiation of antipyretic
interventions between professions. Nurse/other (n = 289); doctor
(n=134) (P <0.0001).
Figure 2 (abstract P18). First-line and second-line treatment preferences
for fever management (n = 458).
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S10

CMACE reports (for example, venous thromboembolism). We therefore
propose to develop local clinical guidelines, posters and factsheets with
formal teaching sessions and multidisciplinary simulator workshops
to raise awareness, optimise care and minimise preventable deaths
from maternal sepsis. We will re-audit awareness in 6 months time to
complete the audit cycle.
References
1. Cantwell R, Clutton-Brock T, Cooper G, et al.: Saving mothers lives: reviewing
maternal deaths to make motherhood safer: 2006–2008. The Eighth
Report of the Con dential Enquiries into Maternal Deaths in the United
Kingdom. BJOG 2011, 118(Suppl 1):1-203.
2. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, Reinhart K,
Angus DC, Brun-Buisson C, Beale R, et al.: Surviving sepsis campaign:
international guidelines for management of severe sepsis and septic
shock: 2008. Crit Care Med 2008, 36:296-327.
P21
Septic shock and vasopressor requirement is associated with lower
vitamin D levels in critically ill children
KMadden1,2, HAFeldman3,4,5, ESmith1, CMGordon3,5,6, SKeisling1,
RSullivan1, BWHollis7, AAAgan1, AGRandolph1,2
1Department of Anesthesia, Children’s Hospital Boston, MA, USA; 2Department
of Anaesthesia, Harvard Medical School, Boston, MA, USA; 3Division of
Endocrinology, Children’s Hospital Boston, MA, USA; 4The Clinical Research
Program, Children’s Hospital Boston, MA, USA; 5Department of Pediatrics,
Harvard Medical School, Boston, MA, USA; 6Division of Adolescent and
Young Adult Medicine, Children’s Hospital Boston, MA, USA; 7Department of
Pediatrics, Medical University of South Carolina, Charleston, SC, USA
Critical Care 2011, 15(Suppl 3):P21 (doi: 10.1186/cc10390)
Introduction Vitamin D plays an important role in immune and
cardiovascular function. There is evidence that low 25-hydroxyvitamin
D (25(OH)D) levels are associated with an increased risk of life-
threatening infections [1,2]. Our objective was to determine the
prevalence of 25(OH)D de ciency (<20 ng/ml) in critically ill children
and to identify any association with illness severity and infection.
Methods From November 2009 to November 2010, we collected blood
samples and clinical data on children (<21 years old) near the time of
admission to the pediatric ICU, excluding those admitted for short-
term monitoring. We measured plasma 25(OH)D concentrations in
plasma using Diasorin radioimmunoassay on all subjects. Vasopressor
requirement was measured using the cardiovascular component of the
Sequential Organ Failure Assessment (CV-SOFA) score.
Results Among 511/818 (62.5%) eligible children, 40.1% were 25(OH)D
de cient (median level 22.5 ng/ml (IQR = 16.4, 31.3)). Children with
a con rmed (n = 144, 28.2%) or suspected (n = 94, 18.1%) diagnosis
of infection on admission did not have lower 25(OH)D levels overall,
except for those presenting in severe septic shock (n = 51, median =
19.2 ng/ml, IQR = 12.6, 24.8; P = 0.0008). In the multivariate analysis,
older age and nonwhite race were associated with vitamin D de ciency
while summer season, vitamin D supplementation and formula intake
were strongly protective. Patients with higher pediatric ICU admission
day illness severity by PRISM-III score quartiles had lower vitamin
D levels (OR = 1.19 per 5 ng/ml decrease in 25(OH)D, 95% CI = 1.10,
1.28, P <0.0001) after adjusting for risk factors. When septic shock was
added to this model, there was no e ect on the association between
25(OH)D level and PRISM-III quartile (OR = 1.18 (95% CI = 1.09, 1.27,
P <0.0001)). There was also an inverse association between 25(OH)D
level and maximal vasopressor use as measured by the CV-SOFA score
in a multinomial regression model (OR = 1.13, 95% CI = 1.01, 1.27, P =
0.03). Including septic shock in the multivariable model did not a ect
the e ect of vitamin D level (OR = 1.16, 95% CI = 1.02, 1.31, P = 0.02))
on CV-SOFA score.
Conclusion The overall prevalence of vitamin D de ciency in critically ill
children is high, and patients with severe septic shock had signi cantly
lower vitamin D levels than the general population. This association
between vitamin D and septic shock may be due to the cardiovascular
e ects of vitamin D or to increased severity of infection with diminished
25(OH)D levels. These results suggest a role for the vitamin D axis in
sepsis and hemodynamic instability that deserves further investigation.
References
1. Braun A, Chang D, Mahadevappa K, Gibbons FK, Liu Y, Giovannucci E,
Christopher KB: Association of low serum 25-hydroxyvitamin D levels and
mortality in the critically ill. Crit Care Med 2011, 39:671-677.
2. McNally JD, Leis K, Matheson LA, Karuananyake C, Sankaran K, Rosenberg AM:
Vitamin D de ciency in young children with severe acute lower
respiratory infection. Pediatr Pulmonol 2009, 44:981-988.
P22
The endothelial glycocalyx degenerates with increasing sepsis
severity
MKöhler1, IKaufmann1, JBriegel1, MJacob1, JGoeschl1, WRachinger2,
MThiel3, MRehm1
1Clinic of Anesthesiology, University of Munich, Germany; 2Neurosurgery,
University of Munich, Germany; 3Department of Anesthesiology and Critical
Care Medicine, University of Heidelberg, Mannheim, Germany
Critical Care 2011, 15(Suppl 3):P22 (doi: 10.1186/cc10391)
Introduction The endothelial glycocalyx is a recently discovered
structure at the luminal side of blood vessels consisting of proteo-
glycans and glycosaminoglycans, which play an important role in
vascular barrier function and cell adhesion. Due to its vulnerability, the
endothelial glycocalyx may easily be altered by hypoxia [1], TNFα [2],
oxidized lipoproteins [3] and other nonphysiological conditions. We
raised the question of whether the glycocalyx may be shed from the
endothelium in dependence of severity of sepsis.
Methods This clinical prospective study – approved by the local ethics
committee – was performed to assess plasma levels of the glycocalyx
components (hyaluronane, syndecan, heparan sulfate) by ELISA
technique and polymorphonuclear leukocyte (PMN) function by  ow
cytometry in eight healthy volunteers (HV) and 37 patients who were
prospectively enrolled within 24 hours of onset of signs of infection,
if they met the criteria for sepsis (n = 10), severe sepsis (n = 9) and
septic shock (n = 18) as de ned by the members of the ACCP/SCCM
Consensus Conference Committee (Table 1). Blood was drawn within
Figure 1 (abstract P20).
Figure 2 (abstract P20).
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S11

24 hours after onset of sepsis. Informed consent was obtained from all
patients or their legal representatives, respectively.
Results Plasma levels of the glycocalyx components were signi cantly
higher in septic patients than in healthy volunteers and even more
pronounced in patients with severe sepsis and septic shock (all P <0.05;
Figure 1). Hyaluronan and syndecan plasma levels correlated positively
with the APACHE II, SOFA and MOD scores (Figure 1 and Table 2).
Hyaluronan displayed a positive correlation with the C-reactive protein,
procalcitonin and IL-6 in plasma (Table 3). The PMN dysfunction –
characterized by an increase in cytotoxic capability and a decrease in
microbicidity – showed a parallel course to the heparan sulfate plasma
levels.
Conclusion Elevated plasma levels of hyaluronan, syndecan and
heparan sulfate are suggestive of a glycocalyx shedding from
endothelium with increasing sepsis severity. This process might
contribute to the vascular dysfunction and development of edema in
septic patients.
References
1. Rehm M, Bruegger D, Christ F, et al.: Shedding of the endothelial glycocalyx
in patients undergoing major vascular surgery with global and regional
ischemia. Circulation 2007, 116:1896-1906.
2. Henry CB, Duling BR: TNFα increases entry of macromolecules into luminal
endothelial cell glycocalyx. Am J Physiol Heart Circ Physiol 2000,
279:H2815-H2823.
3. Vink H, Constantinescu AA, Spaan JA: Oxidized lipoproteins degrade the
endothelial surface layer: implications for platelet-endothelial cell
adhesion. Circulation 2000, 101:1500-1502.
P23
Abstract withdrawn
Figure 1 (abstract P22). (a) to (c) Increase in the glycocalyx components in plasma of healthy volunteers (HV) and of patients with increasing sepsis severity.
(d) Correlation between APACHE II score of septic patients and hyaluronan plasma concentrations.
Table 1 (abstract P22). Demographic data
Healthy Severe Septic
volunteers Sepsis sepsis shock
(n = 8) (n = 10) (n = 9) (n = 18)
Age (years) 29.1 ± 2.9 51.6 ± 19.7 63.3 ± 23.5 63.3 ± 21.4
APACHE II n.b. 7.6 ± 3.9 17.8 ± 6.9 27.9 ± 5.3
MOD n.b. 2.1 ± 1.6 6.9 ± 3.2 9.4 ± 3.6
SOFA n.b. 4.1 ± 2.8 9.0 ± 3.0 13.3 ± 3.4
Table 2 (abstract P22). Correlation between the glycocalyx components
(hyaluronan, syndecan) and the APACHE II, SOFA and MOD score of septic
patients
APACHE II SOFA MOD
Hyaluronan r
2
= 0.583, r
2
= 0.529, r
2
= 0.435,
P = 0.000 P = 0.001 P = 0.008
Syndecan r
2
= 0.425, r
2
=0.476, r
2
= 0.529,
P = 0.010 P = 0.003 P = 0.001
Table 3 (abstract P22). Correlation between the glycocalyx components
(heparan sulfate, hyaluronan) and the C-reactive protein, procalcitonin and
IL-6 in plasma of septic patients
CRP PCT IL-6
Heparan sulfate r
2
= –0.63, r
2
= 0.20, r
2
= 0.505,
P = 0.714 P = 0.928 P = 0.012
Hyaluronan r
2
= 0.398, r
2
= 0.723, r
2
= 0.468,
P = 0.016 P = 0.000 P = 0.021
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S12

P24
Early peak temperature and mortality in critically ill patients with or
without infection
MSaxena1,2, PYoung3,4, RBeasley3,4, MBailey5, RBellomo5, DPilcher6,
SFinfer1,7, DHarrison8, JMyburgh1,2, KRowan8
1George Institute for Global Health, Sydney, NSW, Australia; 2St George Clinical
School, University of New South Wales, Sydney, NSW, Australia; 3Intensive
Care Unit, Wellington Regional Hospital, Capital and Coast District Health
Board, Wellington, New Zealand; 4Medical Research Institute of New Zealand,
Wellington, New Zealand; 5Australian and New Zealand Intensive Care
Research Centre, School of Public Health and Preventive Medicine, Monash
University, Melbourne, VIC, Australia; 6Australian and New Zealand Intensive
Care Society Centre for Outcome and Resource Evaluation, Melbourne, VIC,
Australia; 7Sydney Medical School, University of Sydney, NSW, Australia;
8Intensive Care National Audit and Research Centre, London, UK
Critical Care 2011, 15(Suppl 3):P24 (doi: 10.1186/cc10393)
Introduction The febrile response in the context of infection may
be linked to a protective host response through enhanced immune
function at elevated body temperatures [1-9]. Alternatively the use
of antipyretics may reduce metabolic expense, patient discomfort, or
protect against neurological injury.
Objective To determine whether fever is associated with reduced risk
of death in patients admitted to an ICU with infection compared with
other patients.
Methods A retrospective cohort study using a database of Australian
and New Zealand (ANZ) ICU admissions as a development cohort and
a database of UK ICU admissions as a validation cohort. The sample
included 129 ICUs in ANZ and 201 ICUs in the UK. The ANZ development
cohort consisted of 269,078 patients and the UK validation cohort
consisted of 366,973 patients. All patients were admitted to an ICU
between 2005 and 2009. A total of 29,083/269,078 (10.8%) ANZ patients
and 103,191/366,973 (28.1%) UK patients were categorised as having
an infection at the time of ICU admission. The main outcome measures
were the association between peak temperature in the  rst 24 hours
after ICU admission and in-hospital mortality in patients admitted with
or without infection.
Results In the ANZ cohort, adjusted in-hospital mortality risk
progressively decreased with increasing peak temperature in patients
with infection. Relative to 36.5 to 36.9°C, the lowest risk was at 39 to
39.4°C (adjusted OR = 0.56; 95% CI = 0.48 to 0.66). In patients without
infection, the adjusted mortality risk progressively increased above
39.0°C (adjusted OR = 2.07 at ≥40.0°C; 95% CI = 1.68 to 2.55). In the UK
cohort,  ndings were similar with adjusted odds ratios at corresponding
temperatures of 0.77 (95% CI = 0.71 to 0.85) and 1.94 (95% CI = 1.60
to 2.34) for the infection and non-infection groups, respectively. See
Figures 1 and 2.
Conclusion Peak temperature in the  rst 24 hours in the ICU is
associated with decreased in-hospital mortality in critically ill patients
with an infection; randomised trials are needed to compare the e ect
on mortality of controlling fever against a permissive approach to fever
management in such patients.
References
1. Kluger MJ: Fever: Its Biology, Evolution and Function. Princeton, NJ: Princeton
University Press; 1979:224 pp.
2. Mackowiak PA: Fever: blessing or curse? A unifying hypothesis. Ann Int Med
1994, 120:1037-1040.
3. Hench PS, Slocumb CH, Popp WC: Fever therapy: results for gonorrheal
arthritis, chronic infectious (atrophic) arthritis, and other forms of
‘rheumatism’. J Am Med Assoc 1935, 104:1779-1790.
4 Wagner-Jauregg J: Uber die Einwirkung der Malaria auf die progressive
Paralyse. Psychiatrisch-neurologische Wochenschrift 1919, 20:251-255.
5. Eyers S, Weatherall M, Shirtcli e P, Perrin K, Beasley R: The e ect on mortality
of antipyretics in the treatment of in uenza infection: Systematic review
and meta-analysis. J R Soc Med 2010, 103:403-411.
6. Shann F: Antipyretics in severe sepsis. Lancet 1995, 345:338.
7. Doran T, de Angelis C, Baugardner R, Mellits E: Acetaminophen: more harm
than good for chickenpox? J Pediatr 1989, 114:1045-1048.
8. Brandts CH, Ndjave M, Graninger W, Kremsner PG: E ect of paracetamol on
parasite clearance time in plasmodium falciparum malaria. Lancet 1997,
350:704-709.
9. Stanley E, Jackson G, Panusarn C, Rubenis M, Dirda V: Increased virus
shedding with aspirin treatment of rhinovirus infection. JAMA 1975,
231:1248-1251.
Figure 1 (abstract P24). Adjusted* odds ratios for in-hospital mortality
versus peak temperature in the  rst 24 hours in the ICU for patients in
the infection group. *Odds ratios adjusted for illness severity using the
ICNARC (2009) model predicted log odds of acute hospital mortality with
the temperature component removed for the UK data and the APACHE
III predicted log odds risk of death with the temperature component
removed for the ANZ data.
Figure 2 (abstract P24). Adjusted* odds ratios for in-hospital mortality
versus peak temperature in the  rst 24 hours in the ICU for patients in the
non-infection group. *Odds ratios adjusted for illness severity using the
ICNARC (2009) model predicted log odds of acute hospital mortality with
the temperature component removed for the UK data and the APACHE
III predicted log odds risk of death with the temperature component
removed for the ANZ data.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S13

P25
Role of procalcitonin and high-sensitivity C-reactive protein in
sepsis: a prospective study
SDas1, DAnand1, SRay2, SBhargava1, AManocha1, MKankra1,
LMSrivastava1
1Department of Biochemistry, Sir Ganga Ram Hospital, New Delhi, India;
2Department of Critical Care Medicine, Sir Ganga Ram Hospital, New Delhi, India
Critical Care 2011, 15(Suppl 3):P25 (doi: 10.1186/cc10394)
Introduction Sepsis is the most common cause of morbidity and
mortality in ICU patients. The clinical signs of infection and routine
laboratory tests are not speci c and at times misleading. Even the
bacteriological evidence of infection is not sensitive enough. In view
of this diagnostic and therapeutic dilemma, an e ective and speci c
marker is needed that can support or exclude the diagnosis of infection.
Hence we evaluated the usefulness of biochemical markers such as
procalcitonin (PCT) and high-sensitivity C reactive protein (hsCRP) in
monitoring the therapeutic response to treatment in patients with
sepsis.
Methods Fifty-seven patients admitted to the ICU of Sir Ganga Ram
Hospital, New Delhi, India from July 2010 to April 2011 with a fresh
episode of sepsis were included in the study. PCT and hsCRP were
analyzed in serum samples at 0 hours, 24 hours and 72 hours. Blood
cultures were performed on the day of admission (0 hours) and the
patients were categorized as culture positive or culture negative.
Patients were followed up for 28 days and were then grouped as
survivors and nonsurvivors.
Results During the observation period of 28 days, 44 patients survived
and 13 expired. Over a period of 0 to 72 hours the PCT level decreased in
86% survivors as compared with 46% nonsurvivors (P =0.001), whereas
it increased in 13.6% survivors as compared with 53.8% nonsurvivors
(P = 0.001). The change in the levels of hsCRP in both surviving and
nonsurviving patients was not signi cant. The 0-hour blood culture
was positive in 24 out of 57 patients. The PCT levels at 0 hours was
signi cantly high in culture-positive patients (P = 0.002) as compared
with culture-negative patients, whereas there was no signi cant
di erence in the hsCRP levels in the above-mentioned groups. The
area under the curve of PCT and hsCRP in culture-positive and culture-
negative patients was 0.743 (95% CI = 0.608 to 0.878, P = 0.002) and
0.564 (95% CI = 0.414 to 0.715, P = 0.405), respectively.
Conclusion These observations indicate that PCT is a better prognostic
indicator than hsCRP and is also a better predictor of bacteremia in the
newly admitted critically ill patients of sepsis. Hence, routine use of PCT
as a monitoring tool may aid in appropriate therapeutic intervention in
patients with sepsis.
Acknowledgements The authors thank the Indian Council of Medical
Research, New Delhi, India for  nancial support.
P26
Prehospital identi cation of sepsis patients and alerting of
receiving hospitals: impact on early goal-directed therapy
KHalimi1, JFreeman-Garrick2, CAgcaoili1, KChoy1, FClaridge3, MJacobs3,
MTaigman4
1Department of Emergency Medicine and Intensive Care, Washington
Hospital, Fremont, CA, USA; 2Highland Hospital, Oakland, CA, USA; 3Alameda
County Emergency Medical Services, San Leandro, CA, USA; 4American
Medical Response, Alameda County, CA, USA
Critical Care 2011, 15(Suppl 3):P26 (doi: 10.1186/cc10395)
Introduction Over the past several years, the early identi cation and
aggressive treatment of sepsis patients has become a standard of care
in the hospital setting. A relatively small number of emergency medical
service (EMS) systems have started programs to screen for sepsis; an
even smaller number provide treatment based on that screening
process in the prehospital setting.
Objective The purpose of this study is twofold. First, the study aims
to determine how e ectively paramedics working in the county EMS
system can use a screening tool to identify potential sepsis patients
and provide an alert to the receiving hospital. Second, the study will
examine whether or not an early identi cation process in the  eld leads
to improved treatment of sepsis. The end goal is to reduce morbidity
and mortality of sepsis patients in the hospital setting.
Methods This is a multi-site prospective observational study with
comparison to retrospective cohort. Patient data will be collected to
determine whether or not the alert process leads to early obtaining of
a serum lactate measurement and early goal-directed therapy.
Results Data points being analyzed from prehospital care reports:
criteria from the sepsis screening tool include evidence of infection,
temperature, heart rate, respiratory rate; and EMS  eld clinical
impression (for comparison with emergency department (ED)
admitting diagnosis). Data points being analyzed in the hospitals
include the following: ED admitting diagnosis; serum lactate values,
blood culture, timestamps; evidence of early goal-directed therapy–
timestamps/values for  uid and antibiotic administration; hospital
admitting diagnosis; and discharge diagnosis. The results of the study
will be available by 2012/2013.
Conclusion Preliminary anecdotal and early data analysis reports from
EMS and hospital sta suggest that patients are being identi ed as
septic prior to ED arrival and have lower lactate levels. Patients are also
treated more timely on arrival to the ED for sepsis. Final data analysis
will shed more light on our hypothesis. Early identi cation of septic
patients has implications for further research both in the  eld and
hospital settings.
P27
Abstract withdrawn
P28
Thalidomide in combination with augmentin (amoxicillin with
clavulanic acid) protects BALB/c mice su ering from Klebsiella
pneumoniae B5055-induced sepsis
VKumar, SChhibber
Department of Microbiology, Panjab Univeristy, Chandigarh, India
Critical Care 2011, 15(Suppl 3):P28 (doi: 10.1186/cc10397)
Introduction Despite extensive research in the  eld of sepsis
pathogenesis and its management, mortality associated with sepsis in
hospitals remains very high. For example, more than 18 million people
are a ected by sepsis worldwide and have an expected 1% increase
annually in ICUs. Sepsis is the outcome of a deregulated immune
system occurring during systemic bacterial (that is, Gram-negative or
Gram-positive) infection. So modulating the immune system by an
immunomodulatory approach may prove bene cial to sepsis patients.
In the present study, we evaluated the protective immunomodulatory
e ect of thalidomide alone or with augmentin in Klebsiella pneumoniae
B5055-induced sepsis in BALB/c mice.
Methods The mouse model of sepsis was developed by placing
K.  pneumoniae B5055 entrapped in  brin and thrombin clots in
the peritoneal cavity of mice. The septic mice were treated with
thalidomide alone (30 mg/kg/day p.o.), with augmentin alone
(20 g/ml i.p.) and with their combination. The bacterial load in blood
was estimated by blood culture on MacConkey’s agar plates along with
measuring the other systemic in ammatory parameters. For example,
lipid peroxidation was measured in terms of malondialdehyde (MDA)
and nitric oxide (NO) levels in serum by biochemical methods. Levels
of proin ammatory cytokines (that is, TNFα and IL-1α) and anti-
in ammatory cytokine (that is, IL-10) levels in serum were measured
by ELISA.
Results The thalidomide-alone-treated mice showed 75% survival
whereas 60% of the augmentin-alone-treated group survived.
However, their combination (thalidomide + augmentin) treatment
provided 100% survival. Treatment with thalidomide alone signi cantly
(P <0.05) decreased TNFα, IL-1α, NO and MDA levels in the serum
without signi cantly (P <0.05) decreasing the bacterial count in
blood. However, levels of IL-10 in serum were found to be signi cantly
(P<0.05) elevated upon thalidomide treatment. Augmentin alone only
decreased the bacterial load in blood signi cantly (P <0.05), while no
signi cant decrease was observed on in ammatory mediators studied.
However, a combination thalidomide with augmentin signi cantly
(P <0.05) decreased both the bacterial count as well as in ammatory
mediators (that is, TNFα, IL-1α, NO and MDA) and provided 100%
protection to animals.
Conclusion Thalidomide can be used as an immunomodulatory agent
along with antibiotics for sepsis management.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S14

P29
A novel DDAH-1 inhibitor improved sepsis-induced impairment in
vasoreactivity to noradrenaline in a rat endotoxaemia model
ZWang, VTaylor, RStidwill, JLeiper, MSinger
Bloomsbury Institute of Intensive Care Medicine, Department of Medicine,
University College London, UK
Critical Care 2011, 15(Suppl 3):P29 (doi: 10.1186/cc10398)
Introduction In septic shock, iNOS activation and nitric oxide (NO)
overproduction contribute to vascular hyporeactivity to adrenergic
vasopressors. The consequent hypotension often necessitates high
doses of catecholamine administration. However, this may lead to
detrimental e ects on tissue perfusion, immune function and myo-
cardial function. Asymmetric dimethlyarginine (ADMA), an endo ge nous
inhibitor of NO synthase, is extensively metabolised by dimethylarginine
dimethylaminohydrolase (DDAH). Competitive inhi bi tion of the DDAH-1
isoform should thus reverse hypotension but, as this isoform is absent
in immune cells, it should not compromise the immune e ects of NO.
Hence, we investigated whether L257, a novel DDAH-1 inhibitor, could
spare norepinephrine dosing in a rat endotoxic shock model.
Methods Anaesthetised, spontaneously breathing male Wistar rats
(body weight 270 to 330 g) had their left carotid artery and right
internal jugular vein cannulated for arterial pressure monitoring and
 uid infusion, respectively. Then 40 mg/kg Klebsiella pneumoniae
lipopolysaccharide was administered intravenously over 30 minutes
followed by  uid resuscitation at a rate of 10 ml/kg/hour thereafter.
When the mean arterial pressure fell over 20% below baseline, they
received norepinephrine titrated to maintain arterial pressure at ±10%
baseline. Thirty minutes post commencement of norepinephrine,
animals were randomized to receive either L-257 (3 mg/kg bolus
then infusion of 125 µg/hour) or, in controls, an equivalent volume
of saline. Experiments were terminated 3 hours post commencement
of norepinephrine titration, before which echocardiography was
performed and serum samples were collected for biochemistry.
Results L-257-treated animals (n = 8) required a signi cantly lower total
dose of noradrenaline over 3 hours compared with the eight control
animals (38 ± 9 vs. 48 ± 4 µg, P <0.05). The heart rate was signi cantly
lower in the treatment group (P <0.05), which associated with a trend
of increased stroke volume and cardiac output. Serum BUN and urea
were also signi cantly lower in the treatment group (P <0.05, Table 1).
Conclusion In this acute endotoxic rat model, we demonstrate that
DDAH-1 inhibition by L-257 could reduce norepinephrine dosage and
ameliorate its harmful e ects. This agent warrants further study as a
putative therapy for septic shock.
Acknowledgements This study was funded by Wellcome Trust in the UK.
P30
A novel DDAH-1 inhibitor improved cardiovascular function in a
short-term anesthetized rat model of sepsis
ZWang, VTaylor, RStidwill, JLeiper, MSinger
Bloomsbury Institute of Intensive Care Medicine, Department of Medicine,
University College London, UK
Critical Care 2011, 15(Suppl 3):P30 (doi: 10.1186/cc10399)
Introduction Excessive NOS activity and NO overproduction are
believed to play an important role in sepsis-induced macrocirculatory
and microcirculatory dysfunction. Asymmetric dimethylarginine
(ADMA), an endogenous inhibitor of NO synthesis, is extensively
metabolised by dimethylarginine dimethylaminohydrolase (DDAH).
Table 1 (abstract P29).
Variable NE NE + L-257 P value
SV (ml/minute) 0.18 ± 0.04 0.23 ± 0.04 0.07
HR (beats/minute) 500 ± 15 449 ± 37 0.03
CO (ml) 90 ± 18 102 ± 19 0.31
U (mM) 16.6 ± 1.2 12.8 ± 1.7 0.01
BUN (mg/dl) 44.8 ± 3.7 35.9 ± 4.8 0.03
Cr (μM) 51.3 ± 15.2 31.8 ± 4.3 0.08
ALT (IU/l) 98 ± 79.8 71.3 ± 48.3 0.6
Figure 1 (abstract P30). (a) The e ect of bolus doses of L-257 at 0 mg/kg, 3 mg/kg, 30 mg/kg and 300 mg/kg on the mean arterial pressure in the short-
term organ function study on anaesthetized Wistar rats (n = 6 in the each group). The bolus L-257 was given at 0 minutes for 10-minute infusion. Animals
treated with 300 mg/kg L-257, ×; animals treated with 30 mg/kg L-257, open triangle; animals treated with 3 mg/kg, open square; control animals, solid circle.
(b) The initial change of mean arterial pressure 30 minutes after bolus injections among the four groups. *P <0.05, 30 mg/kg vs. control; #P <0.05, 30 mg/kg
vs. 3 mg/kg; †P <0.05, 300 mg/kg vs. control; ‡P <0.05, 300 mg/kg vs. 3 m/kg. (c), (d) The e ects of bolus doses of L-257 at 0 mg/kg, 3 mg/kg and 30 mg/kg on
perfused capillary density and microcirculatory index (n = 4 in each group). *P <0.05, 30 mg/kg vs. control; #P <0.05, 30 mg/kg vs. 3 m/kg.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S15

The DDAH-1 isoform is present in vascular smooth muscle so its
inhibition should theoretically reverse sepsis-induced hypotension. We
thus investigated the dose-dependent cardiovascular e ec ts of a novel
DDAH-1 competitive inhibitor, L-257, in experimental sepsis.
Methods Anaesthetised, spontaneously breathing male Wistar rats
(body weight 270 to 330 g) had their left carotid artery and right
internal jugular vein cannulated for arterial pressure monitoring and
 uid infusion, respectively. Then 40 mg/kg Klebsiella pneumoniae
lipopolysaccharide was administered intravenously over 30 minutes
followed by  uid resuscitation at a rate of 10 ml/kg/hour thereafter.
When the mean arterial pressure fell over 20% below baseline, groups
(n = 6) were randomized to receive a bolus dose of L-257 of 0 (control),
3, 30 or 300 mg/kg. Animals were sacri ced 2 hours later with prior
measurement of gastrocnemius muscle microcirculatory perfusion and
with collection of plasma samples for biochemistry, arginine, ADMA
and nitrate/nitrite measurements.
Results The bolus doses of L-257 were given after approximately 60 to
90 minutes post endotoxin when the mean BP fell over 20%. Arterial
pressure, perfused capillary density and microcirculatory  ow index were
better maintained than in controls, especially at higher doses (Figure 1,
P <0.05). Signi cantly higher plasma ADMA concentrations and ADMA/
arginine ratios were seen in the 30 mg/kg bolus group (Figure 2, P <0.05).
Plasma nitrate/nitrite levels in the treated animals were signi cantly
lower compared with those in controls (Figure 2, P<0.05).
Conclusion In this short-term rat model of endotoxaemia, we
demonstrated protective dose-dependent e ects of a novel DDAH-1
inhibitor, L-257, on cardiovascular function. This was associated with an
elevation of plasma ADMA level and a resultant reduction of plasma
nitrate/nitrite level.
Acknowledgements This study was funded by Wellcome Trust in the UK.
P31
Early detection of serum enteric bacterial DNA with real-time PCR in
patients with SIRS
JM-CYang
Department of Surgery, Kaohsiung Veterans General Hospital, Taiwan, China
Critical Care 2011, 15(Suppl 3):P31 (doi: 10.1186/cc10400)
Introduction Sepsis remains a major and increasing healthcare problem
with a mortality exceeding 25%. The early detection of infection is
important in treating sepsis. Nucleic acid amplication methods have
the potential to improve the timeliness, sensitivity, and accuracy of the
tests used to detect respiratory pathogens. We used a quantitative real-
time PCR (rt-PCR) to detect the enteric bacterial counts in blood from
patients in the emergency room.
Methods EDTA samples were collected from patients with systemic
in ammatory response syndrome (SIRS) presenting to the emergency
room after obtaining informed consent. Enteric bacterial loads in blood
samples were assayed by rt-PCR to quantitate the bacterial 23S rDNA
and EB rDNA loads. Descriptive and clinical data were collected from
the medical records and compared with 23S and EB rDNA results.
Results From January 2011 to April 2011, 39 patients (mean age
71.15± 17.12, range 22 to 93) were enrolled in the study. There was no
correlation between serum lactate and enteric bacterial load in patients
with SIRS. However, in a subgroup comprising patients presenting with
respiratory distress and abnormal blood white cell count, the enteric
bacterial rDNA load was higher and showed correlation with serum
lactate level. The serum enteric bacterial rDNA loads were signi cantly
higher in patients with positive cultures and in patients presenting with
higher serum lactate. Correlations between serum lactate and enteric
bacterial rDNA load were also signi cant in the patients with positive
culture results.
Conclusion The quantitative assay for enteric bacterial rDNA could be
a useful tool to detect early enteric bacterial translocation in patients
presenting to the emergency room with elevated serum lactate level or
with respiratory distress and abnormal white blood cell counts.
P32
Direct e ects of esmolol, ultra-short-acting β-blockers, on cardiac
function, ion channels, and coronary arteries in guinea pigs
SShibata
Akita University Graduate School of Medicine, Department of Cell Physiology,
Akita, Japan
Critical Care 2011, 15(Suppl 3):P32 (doi: 10.1186/cc10401)
Introduction β1-adrenergic antagonists have been recently used in
septic patients to improve sepsis-induced immune, cardiovascular and
coagulation dysfunction. But it is di cult and one is hesitant to use
these drugs in septic shock patients who have already had hypotension
Figure 2 (abstract P30). The e ect of bolus doses of L-257 at 0 mg/kg, 3 mg/kg and 30 mg/kg on (a) plasma arginine and (b) ADMA concentrations, (c) the
ADMA/arginine ratio and (d) plasma nitrate/nitrite concentrations in the short-term organ function study on anaesthetized Wistar rats (n = 6 in each group).
*P <0.05, 30 mg/kg vs. control; #P <0.05, 30 mg/kg vs. 3 m/kg; †P <0.05, 3 mg/kg vs. control.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S16

because these drugs sometimes trigger excessive hypotension due to
direct e ects on heart function in addition to their β1 blocking e ects.
Since little is known about their acute direct e ects on mammalian
heart, we therefore evaluated the direct e ects of esmolol, ultra-
short-acting β-blockers, on cardiac performance and single cell-
electrophysiology in guinea pig hearts, and compared these e ects
with those of landiolol.
Methods All animal experiments were approved by the University
Animal Ethics Committee. Under deep anesthesia with pentobarbital,
the heart was excised and mounted on a Langendor apparatus
to measure the coronary perfusion pressure (CPP). The saline- lled
balloon was inserted into the left ventricle to measure the heart rate
(HR) and systolic left ventricular pressure (sLVP). The coronary  ow
was maintained at a constant value during the experiments. Single
ventricular cells were enzymatically isolated from hearts and cardiac
ion currents were investigated by the patch clamp methods. Group
comparisons were conducted by one-way repeated-measures analysis
of variance with Dunnett’s or Turkey’s multiple comparison test.
Di erences at P <0.05 were considered to denote signi cance.
Results Esmolol increased CPP in a concentration-dependent manner,
and decreased both the sLVP and HR signi cantly at concentrations
>10 M. Esmolol also shortened the action potential duration (APD) in a
concentration-dependent manner, and inhibited the inward recti er K+
current (IK1), while the L-type Ca2+ current (ICaL) and outward current (IKs
and IKr) and ATP-sensitive K+ current were hardly a ected. Furthermore,
with the application of BAPTA from patch pipettes, the chelation
of intracellular calcium ion did not antagonize APD shortening by
esmolol. On the other hand, landiolol had minimal e ects on cardiac
coronary perfusion, cardiac contractility, action potential, and cardiac
ionic currents. In the Kyoto Model computer simulation, sole inhibition
of IK1 or ICaL failed to simulate APD shortening induced by esmolol.
Conclusion The present  ndings demonstrated that esmolol has
more direct e ects on the heart than landiolol; that is, the elevation
of coronary perfusion pressure and negative inotropic action. The
negative inotropic action is accompanied with the APD shortening in
single cardiomyocytes. Inhibition of IK1 and ICaL, and inhibition of ionic
current systems other than those we identi ed may be involved in the
APD shortening caused by esmolol.
P33
Abstract withdrawn
P34
Sepsis-induced lung  brosis in baboons is reduced by the treatment
with a complement inhibitor
FLupu1,2, HZhu1, RSilasi-Mansat1, CGeorgescu3, NPopescu1, GPeer4,
CLupu1, FTaylor1,2, GKinasewitz4, JLambris5
1Cardiovascular Biology Research Program, Oklahoma Medical Research
Foundation, Oklahoma City, OK, USA; 2Department of Pathology, Oklahoma
University Health Sciences Center, Oklahoma City, OK, USA; 3Neuroscience
Research Department, Mayo Clinic Jacksonville, FL, USA; 4Department
of Medicine, Pulmonary and Critical Care Division, Oklahoma University
Health Sciences Center, Oklahoma City, OK, USA; 5Department of Pathology
and Laboratory Medicine, School of Medicine, University of Pennsylvania,
Philadelphia, PA, USA
Critical Care 2011, 15(Suppl 3):P34 (doi: 10.1186/cc10403)
Introduction Pulmonary  brosis is a major and common medical
condition, characterized by progressive scaring and decline in lung
function. Persistent in ammation and acute lung injury in response
to sepsis are potential triggers of the  brotic response. Recently,
we have reported that Escherichia coli sepsis in baboons strongly
induces procoagulant responses and a ects the integrity of the lung.
These e ects are diminished by the treatment with compstatin, a C3
convertase complement inhibitor [1].
Methods Here we used the baboon model described [1] in conjunction
with detailed gene expression analysis, as well as biochemical and
histological assays to determine if E. coli sepsis triggered metabolic
and signaling pathways related to lung remodeling and  brosis, and
whether complement inhibition could attenuate these pathways.
Results Microarray gene expression analysis shows that sepsis
augments several  brotic gene clusters in the lung as early as 24 hours
post E. coli challenge. Immunochemical and biochemical analysis
reveals enhanced collagen synthesis, induction of pro brotic factors
and increased cell recruitment and proliferation. Compstatin treatment
decreases sepsis-induced expression of extracellular matrix genes,
including eight collagen genes. Sirius Red and immuno uorescence
staining for procollagens 1 and 3 con rms the collagen deposition in
the lung. Ingenuity® pathway analysis of transcriptomics data shows
that compstatin treatment reduces sepsis-induced expression of
genes involved in  broblast transformation and connective tissue
production, cell chemotaxis, migration and proliferation (see Table1).
Immunocytochemistry and pathway-oriented transcriptomics and
phospho-proteomics analysis reveal changes of multiple processes
mediated by transforming growth factor beta (TGF-β), connective tissue
growth factor and other TGF-β controlled proteins. Immunostaining
for cell proliferation markers demonstrates that compstatin treatment
strongly reduces cell proliferation in  broblastic foci. Moreover,
biochemical analysis shows decreased production in the compstatin-
treated group of two chemokines responsible for  brocyte recruitment
(CCL2 and CXCL12) and of the type 1 tissue inhibitor of metalloproteases
that controls extracellular matrix remodeling.
Conclusion Our data demonstrate that bacterial sepsis initiates pulmonary
collagen deposition, and complement inhibition e ectively attenuates the
 brotic response. This suggests that complement inhibitors could be used
for prevention of sepsis-induced pulmonary  brosis.
Acknowledgements The authors thank Dr Bart Frank (OMRF) for help
with protein array scanning and quantitation. This work was supported
by grants from the National Institutes of Health (GM097747-01 to FL
and JL; 2P20RR018758-06A2 and 1RC1GM09739-02 to FL; AI068730
and GM062134 to JL).
Reference
1. Silasi-Mansat R, Zhu H, Popescu NI, Peer G, Sfyroera G, Magotti P, Ivanciu L,
Lupu C, Mollnes TE, Taylor FB, et al.: Complement inhibition decreases the
procoagulant response and confers organ protection in a baboon model
of Escherichia coli sepsis. Blood 2010, 116:1002-1010.
Table 1 (abstract P34).
E. coli E. coli + CS T+5
Modi ed genes Modi ed genes
Up Down Total P value Up Down Total P value
Fibroblast transformation 32 10 42 1.41 x 10
–6
12 69 81 1.09 x 10
–7
Connective tissue disorder 133 59 192 2.62 x 10
–5
92 341 434 3.1 x 10
–7
Chemotaxis 39 13 52 5.0 x 10
–3
31 92 123 1.98 x 10
–5
Cell migration 95 40 135 2.11 x 10
–5
80 250 330 9.36 x 10
–8
Cell proliferation 187 70 257 2.67 x 10
–8
118 466 584 5.83 x 10
–13
Fibroblast proliferation 23 7 30 4.80 x 10
–3
10 62 72 5.56 x 10
–6
CS T+5, compstatin-treated animals at T + 5 hours.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S17

P35
AB103, a CD28 antagonist peptide: a new therapeutic agent in a
model of severe sepsis
DTeleman1, C-SChung2, AAyala2, SOpal3, AChahin3, AShirvan1
1Atox Bio, Ness Ziona, Israel; 2Alpert Medical School at Brown University,
Rhode Island Hospital, Providence, RI, USA; 3Memorial Hospital of Rhode
Island, Pawtucket, RI, USA
Critical Care 2011, 15(Suppl 3):P35 (doi: 10.1186/cc10404)
Introduction AB103 is a novel CD28 antagonist peptide currently in
clinical development that modulates CD28 signaling in T cells, without
a ecting the normal humoral immune response. In experimental
models of Gram-positive, Gram-negative and polymicrobial sepsis,
AB103 demonstrated signi cant activity, increasing overall survival.
Methods The AB103 activity and mode of action (MOA) were evaluated
in a murine model of cecal ligation and puncture (CLP). AB103 (5 mg/kg)
was administered to mice (Balb/c) at various times points following CLP
(2 to 24 hours), together with moxi oxacin.
Results A single dose of AB103, given at 12 or 24 hours post CLP, rescued
100% and 62.2% of the animals (respectively) from sepsis-induced
mortality, whereas moxi oxacin alone (LD90, given at 12 hours) rescued
only 25% (P <0.05) of the animals. In a separate set of experiments
investigating the MOA, AB103 administration (5 mg/kg, given without
antibiotics 2 hours post CLP) was associated with: a reduction in Th-1
cytokine levels in peritoneum (TNFα, IL-3, IL-17 and Rantes) and plasma
(IL-3 and IL-6); a reduction in splenocyte proliferation, stimulated ex
vivo with anti-CD3 and anti-CD28 antibodies; a reduction in neutrophil
recruitment to the spleen, liver and kidney, as determined by MPO
activity; and a reduced bacterial load in peritoneum, blood and tissues
(kidney, liver, spleen).
Conclusion These data demonstrate that attenuation of CD28 signaling
is a viable therapeutic approach to the treatment of sepsis. Due to its
robust activity and good safety pro le in humans already established in
a phase 1 study, AB103 should be clinically evaluated in sepsis patients.
P36
Cardiovascular e ects of β-blockade in a sheep model of severe
sepsis
PCalzavacca1,2,3, RRamchandra1, LBooth1, RBellomo3, CNMay1
1Howard Florey Institute, University of Melbourne, Parkville, VIC, Australia;
2Department of Anaesthesia and Intensive Care, AO Melegnano, PO Uboldo,
Cernusco sul Naviglio, Italy; 3Department of Intensive Care and Department
of Medicine, Austin Health, Melbourne, VIC, Australia
Critical Care 2011, 15(Suppl 3):P36 (doi: 10.1186/cc10405)
Introduction In sepsis, sympathetic nerve activity is di erentially
increased in individual organs. The increased cardiac sympathetic
nerve activity is partly responsible for the increase in heart rate (HR)
and cardiac output (CO) opposing the development of hypotension [1].
Recently, in a rat septic model, β-blockade appeared safe and decreased
the in ammatory response and mortality [2]. Accordingly, we sought to
investigate the cardiovascular e ects of selective β1-receptor blockade
in a sheep model of sepsis.
Methods Eight merino ewes were studied in a university-a liated
research institute in Melbourne. The study design was a prospective
interventional crossover animal study. The animals had renal and
cardiac  ow probes implanted to continuously measure CO and renal
blood  ow (RBF). Every animal was randomly allocated to receive sepsis
and atenolol (atenolol group, AG) or sepsis alone (control group, CG) and
then crossed over. After 24 hours of baseline period, sepsis was induced
through a bolus of live Escherichia coli by a continuous infusion for a
total 24 hours of sepsis. After the  rst 8 hours of sepsis (development
sepsis period, DS), a bolus of atenolol (10 mg bolus) was given followed
by a continuous infusion of 0.125 mg/kg/hours for 16hours. Two-way
repeated-measure ANOVA was performed to compare the average of
periods and group interaction. P <0.05 was considered signi cant (not
signi cant (NS), P >0.05).
Results Animals in the AG and CG had similar baseline values and
developed a similar hyperdynamic state in the DS (Figure 1 and
Table 1). Atenolol reduced CO and HR without changes in stroke
volume. Hypotension was slightly greater in the AG than in the CG
(MAP: 81.5 vs. 86.1 mmHg) with a greater decrease in total peripheral
conductance (16.8 vs. 22.1 l/minute/mmHg). Changes in lactate level
were similar. Similar increases in RBF and in renal vascular conductance
(RVC) were observed in the AG and CG and after an initial increase in
diuresis in the DS, oliguria similarly subsequently developed in both
groups. Creatinine clearance decreased in a similar way in the AG and
CG from 59.2 (± 2.8) to 32 (± 5.7) ml/minute and from 65.2 (± 9.9) to
36(± 7) ml/minute, respectively (P = 0.381). One animal in the AG and
two in the CG died in the 24 hours after the end of sepsis.
Conclusion β-blockade in hyperdynamic sepsis appears safe. It results
in only limited decreases in mean arterial pressure, and does not
increase lactate levels or worsen renal function.
References
1. Ramchandra R, Wan L, Hood SG, Frithiof R, Bellomo R, May CN: Septic shock
induces distinct changes in sympathetic nerve activity to the heart and
kidney in conscious sheep. Am J Physiol Regul Integr Comp Physiol 2009,
297:R1247-R1253.
2. Ackland GL, Yao ST, Rudiger A, et al.: Cardioprotection, attenuated systemic
in ammation, and survival bene t of β1-adrenoceptor blockade in severe
sepsis in rats. Crit Care Med 2010, 38:388-394.
P37
A sepsis progression model in humans: characterization of
biomarkers descriptive of sepsis progression
BWFooter1,2, C-BHsiao1,3, DMParish2, PWickremasingha4, A-BHalim4,
GSenaldi4, RScheyer4, JJSchentag1,2
1State University of New York at Bu alo, Bu alo, NY, USA; 2CPL Associates,
LLC, Bu alo, NY, USA; 3Erie County Medical Center, Bu alo, NY, USA; 4Daiichi
Sankyo Pharma Development, Edison, NJ, USA
Critical Care 2011, 15(Suppl 3):P37 (doi: 10.1186/cc10406)
Introduction Previously our group has developed neural net
progression models to characterize the development of organ failure
in an ovine only as well as an integrated human/ovine model of acute
lung injury using early clinical information. The goal of this study was
to expand our model of disease progression using clinically available
Table 1 (abstract P36). Hemodynamic and renal  ndings during baseline,
development (DS) and intervention sepsis periods in the CG and AG groups
Development Sepsis
Baseline sepsis intervention
Group period period (DS) period P value
CO CG 3.68 (0.29) 3.68 (0.29) 4.12 (0.29) <0.001
AG 3.21 (0.22) 3.21 (0.22) 5.63 (0.53)
HR CG 66.0 (6.5) 66.0 (6.5) 108.8 (8.6) <0.001
AG 59.7 (6.4) 59.7 (6.4) 111.6 (8.4)
MAP CG 93.7 (5.1) 102.5 (3.6) 86.1 (4.1) 0.035
AG 96.3 (5.0) 102.5 (3.8) 81.5 (4.3)
RBF CG 217.3 (14.8) 217.3 (14.8) 324.8 (19.7) 0.194
AG 214.7 (19.9) 214.7 (19.9) 292.3 (27.4)
UO CG 31.1 (7.1) 84.1 (20.4) 22.1 (4.8) 0.097
AG 34.8 (7.0) 52.3 (17.4) 16.8 (11.0)
TPC CG 38.2 (3.4) 38.9 (3.4) 63.9 (6.8) 0.084
AG 34.3 (3.1) 37.3 (3.2) 51.4 (6.8) <0.001
SV CG 51.4 (3.7) 39.1 (3.6) 35.5 (8.2) 0.147
AG 51.2 (4.6) 36.7 (3.5) 39.6 (3.6)
RVC CG 2.39 (0.18) 3.26 (0.26) 3.74 (0.41) 0.55
AG 2.23 (0.16) 2.88 (0.26) 3.53 (0.44)
Values are the mean (± standard error). P value: two-way repeated-measures
ANOVA interaction between treatment group and time (see text for de nitions).
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S18

data as well as more exploratory biomarkers, such as the endotoxin
activity assay (EAA), cytokines, D-dimer, copeptin, and procalcitonin, in
an adult population with sepsis.
Methods Three North American study sites enrolled adult patients
within 24 hours of meeting at least two SIRS criteria with clinical
evidence of infection. Biomarker sampling occurred daily on days 1 to
7 and on days 14, 21, and 28. Clinical data from the 24 hours preceding
the  rst sampling point as well as the baseline biomarker values were
used as model inputs. Model outputs were serum creatinine (Scr)
and organ metric (OM) over the study duration. OM is a composite
parameter similar to the SOFA score with the CNS category removed
and a continuous rather than categorical value. A neural net was used
to perform a multiple parameter logistic regression while allowing for
non-linear (usually sigmoidal) dependence on input parameters. Input
parameters are  rst used individually to model the output and are then
ranked based on the minimum mean squared error (MMSE) in these
single-parameter models. The two parameters with the lowest MMSE
are used to create the  nal multi-parametric model, which yields a
lower modeling error than the original single-parameter models.
Results Thirty patients were enrolled with the two most common
infection types being pneumonia and bloodstream. Seventy per cent
of patients had at least one organ failure at enrollment. Diastolic blood
pressure (DBP), red blood cell count (RBC), and copeptin had the smallest
MMSE when individually predicting OM. Combining DBP and RBC yielded
good agreement between the modeled and actual OM value (r2 = 0.60).
Individually, the prothrombin time (PT), copeptin, and phosphorus had
the smallest MMSE when modeling Scr. The r2 value between the model
and actual Scr was 0.64 when combining PT and copeptin.
Conclusion When analyzed using a neural net model, changes in overall
organ dysfunction and serum creatinine were predicted from early
clinical data in a population of adult patients with sepsis. Identifying
predictive biomarker patterns and coupling this information with
known drug/intervention response could aid in optimizing treatment
timing for greatest clinical bene t.
P38
Abstract withdrawn
P39
Interplay between angiopoietin-2, vascular endothelial growth
factor and peroxynitrite is an important determinant of vascular
hyperpermeability during methicillin-resistant Staphylococcus
aureus sepsis
PEnkhbaatar, YZhu, LTraber, DTraber
Department of Anesthesiology, University of Texas Medical Branch, Galveston,
TX, USA
Critical Care 2011, 15(Suppl 3):P39 (doi: 10.1186/cc10408)
Introduction We have reported that nitric oxide (NO) production and
microvascular hyperpermeability were signi cantly higher in septic
sheep with methicillin-resistant Staphylococcus aureus (MRSA) than
with Pseudomonas aeruginosa. We hypothesize that peroxynitrite,
a byproduct of NO, causes vascular hyperpermeability in MRSA
sepsis via promoting vascular endothelial growth factor (VEGF) and
Figure 1 (abstract P36). Hemodynamic variables during baseline and sepsis in the atenolol (AG, white circles, dotted line) and control (CG, black triangles,
continuous line) groups.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S19

angiopoietin-2 (Ang-2). The hypothesis was tested, using both a
well-established ovine sepsis model and cultured human umbilical
endothelial cells (HUVECs).
Methods Female ewes were chronically instrumented with multiple
catheters and live MRSA (USA300, 1011 CFU) was instilled into the both
lungs by bronchoscope under deep iso urane anesthesia. The sheep
were then randomly allocated to control and treated (nonspeci c NOS
inhibitor L-NAME, 25 mg/kg, i.v., every 12 hours) groups and monitored
for 24 hours for cardiopulmonary hemodynamics. The cells were
challenged with 105 CFU of live MRSA or 50 M peroxynitrite and co-
incubated with or without L-NAME, peroxynitrite scavenger FeTMPyP,
Tie-2 and Ang-2 antibody, and VEGF and its antibody. At di erent times
after the treatment, the permeability was measured by quantifying
the amount of FITC-Dextran that passed through the con uent HUVEC
monolayer (n = 4). Ang-2 mRNA was determined by RT-PCR in those
cells with or without treatment as well (n = 4). Statistical analysis: one-
way ANOVA (Bonferroni).
Results In vivo, L-NAME signi cantly reduced MRSA-induced  uid
accumulation and requirement, as well as expression of VEGF. HUVEC
permeability was time-dependently increased following MRSA co-
incubation, reaching a plateau at 2 and 4 hours. These permeability
changes (73 ± 4 RFUs) were signi cantly (P <0.001) inhibited by 1 mM
L-NAME (28 ± 1), 5 µM FeTMPy (34 ± 2), 5 g/ml Tie-2 antibody (32 ± 2),
and 5 µg/ml Ang-2 antibody (30 ± 1). In HUVECs, the Ang-2 mRNA
was time-dependently (picks at 30 minutes) and dose-dependently
increased by peroxynitrite (highest at 50 M). Treatment of HUVECs
with 5 M VEGF augmented the MRSA-induced Ang-2 mRNA increases.
The latter was reversed with FeTMPyP and 5 M VEGF antibody.
Conclusion Ang-2 and VEGF, Tie-2 receptor, NO and its byproduct
peroxynitrite play an important role in MRSA-induced vascular hyper-
permeability. The results strongly suggest that peroxynitrite increases
vascular hyper-permeability by promoting Ang-2 release through
stimulating the VEGF expression during MRSA-induced Gram-positive
sepsis.
P40
Clinical characteristics, management, and outcomes of sepsis in
Lusaka, Zambia
SMChimese1, BAndrews1,2, SLakhi1,3
1Department of Internal Medicine, University of Zambia, Lusaka, Zambia;
2Institute for Global Health, Vanderbilt University, Nashville, TN, USA;
3Department of Internal Medicine, University Teaching Hospital, Lusaka,
Zambia
Critical Care 2011, 15(Suppl 3):P40 (doi: 10.1186/cc10409)
Introduction Although infectious diseases are the leading causes
of death in sub-Saharan Africa, there are few studies describing
sepsis in the region. Available data suggest that HIV prevalence is
disproportionately high among septic patients and that treatment,
particularly  uid administration, may be suboptimal [1]. Our study
evaluated the clinical characteristics, management, and hospital
outcomes of patients admitted with sepsis in Zambia. We hypothesized
that patients with bacteremia have higher in-hospital mortality than
those without.
Methods We conducted a prospective observational study of patients
admitted with sepsis to the Adult Filter Clinic (medical ER) of the
University Teaching Hospital (UTH) in Lusaka Zambia. Sepsis was
de ned as two or more SIRS criteria and clinically suspected infection.
Baseline characteristics and laboratory results were recorded, as
was the timing of antibiotics and  uid administration. Patients were
followed until discharge or death.
Results In 3 months, 161 septic patients were enrolled. One hundred
and ten (68%) patients were HIV positive; 23 (14%) had unknown HIV
status. Ninety-one (57%) had severe sepsis. Organ dysfunction included
altered mentation (31%), renal dysfunction (16%), severe respiratory
distress (respiratory rate ≥40) (11%), thrombocytopenia (11%), and
hepatic dysfunction (7%). Multiple organ dysfunction occurred in 26%.
After excluding contaminants, blood cultures were positive in 29 (18%)
patients. Staphyloccoccus aureus, salmonella species, Streptococcus
pneumoniae, and Klebsiella pneumoniae were the most common
pathogens. Only 29% of patients received intravenous  uids within
1hour of presentation. Eighty-four percent of patients received ≤1l
within the  rst 6 hours of presentation, and 55% received ≤1 l in the  rst
24hours. Overall in-hospital mortality was 40.4% (65/161). In-hospital
mortality for severe sepsis was 54.9% (50/91). Important predictors
for in-patient mortality (Table 1) were low Glasgow Coma Scale on
admission (adjusted odds ratio (AOR) 16.0 (2.9 to 87.1)), positive blood
culture (AOR 4.8 (1.5 to 15.0)), and positive and unknown HIV status
(AOR 4.20 (1.0 to 17.0) and AOR 7.7 (1.2 to 47.7), respectively).
Conclusion In-hospital mortality due to sepsis is higher in Zambia than
in most studies from the developed world. Low Glasgow Coma Scale
and positive blood cultures are associated with increased in-hospital
mortality. Insu cient i.v.  uid administration probably contributes to
the high overall mortality. Standardized management including early
 uids and antibiotics might improve outcomes of sepsis and severe
sepsis in sub-Saharan Africa.
Reference
1. Jacob S, et al.: Severe sepsis in two Ugandan hospitals: a prospective
observational study of management and outcomes in a predominantly
HIV-1 infected population. PLoS One 2009, 4:e7782.
P41
AZD9773, a novel anti-TNFα immune Fab in development for severe
sepsis and septic shock: demonstration of safety and e cacy in a
murine CLP sepsis model
PNewham1, PCeuppens2, SDas3, JWTYates4, RKnight1, JSMcKay1
1Global Safety Assessment, AstraZeneca, Maccles eld, Cheshire, UK; 2Discovery
Enabling Capabilities and Sciences, AstraZeneca, Maccles eld, Cheshire, UK;
3Clinical Pharmacology & DMPK, AstraZeneca, Maccles eld, Cheshire, UK;
4Oncology iMed DMPK, AstraZeneca, Maccles eld, Cheshire, UK
Critical Care 2011, 15(Suppl 3):P41 (doi: 10.1186/cc10410)
Introduction TNFα is thought to play a central role in the pathogenesis
of sepsis and septic shock. AZD9773 is an ovine polyclonal anti-human
TNFα immune Fab comprising TNFα-directed and nonspeci c Fab
populations. AZD9773 potency and pharmacokinetic attributes such as
a shorter half-life distinguish it from anti-TNF monoclonal antibodies,
which have been assessed previously in clinical sepsis models. Here
we explore the preclinical safety/e cacy of AZD9773 in a mouse
cecal ligation puncture (CLP) model. There are currently no reports
of anti-TNF agent e cacy in mouse CLP; rather, TNFα neutralization
Table 1 (abstract P40). Risk factors for in-hospital death
Variable Adjusted OR Crude OR
Anemia, Hb <9 g/dl 0.91 (0.74 to 4.93) 1.05 (0.56 to 1.80)
Blood culture positive 4.8 (1.50 to 15.0) 2.38 (1.14 to 4.95)
GCS
≥13 1 1
9 to 12 5.50 (1.90 to 16.20) 3.60 (1.60 to 8.11)
<9 16.00 (2.90 to 87.10) 11.2 (3.50 to 36.4)
HIV
Negative 1 1
Positive 4.20 (1.00 to 17.00) 2.35 (0.88 to 6.28)
Unknown 7.70 (1.20 to 47.70) 8.38 (2.36 to 29.7)
<1 hour to IVF 0.40 (0.10 to 1.10) 0.86 (0.43 to 1.72)
MAP <65 2.10 (0.70 to 6.80) 1.25 (0.56 to 2.81)
IVF in  rst 6 hours
0 1 1
1 0.80 (0.30 to 2.00) 0.81 (0.41 to 1.61)
2 or more 0.30 (0.10 to 1.10) 0.60 (0.23 to 1.58)
Statistically signi cant ORs in bold.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S20

is associated with minimal/no e ect or increased mouse mortality.
As AZD9773 is di erentiated from other anti-TNFα agents based on
neutralizing potency and pharmacokinetic attributes, we studied both
the safety and e cacy of this product in mouse CLP models.
Methods We studied AZD9773 (plus imipenem) e ects in two mouse
CLP models: a mild-grade model to explore the potential for AZD9773
to compromise mouse survival, and a severe-grade model to test
AZD9773 e cacy. CLP (mild-grade sepsis) comprised 100% cecal
ligation and single 20-gauge needle puncture, while CLP (severe-grade
sepsis) comprised 100% cecal ligation and single 18-gauge needle
puncture. Saline resuscitation and imipenem administration were
included in both models. Since AZD9773 does not bind or neutralize
murine TNFα, the CLP models were established in Tg1278/–/– (human
TNFα transgene/mouse TNFα null) mice. An equivalent protein dose of
DigiFab (plus imipenem) served as an irrelevant Fab control. Survival
was monitored for 5 days.
Results The control severe-grade model resulted in approximately
20% survival at 5 days. Therapeutic i.p. dosing of AZD9773 bid from
24 to 60 hours ( rst dose 4,000 units/kg, second to fourth doses 2,000
units/kg) resulted in statistically signi cant increases in survival (>70%)
compared with i.p. DigiFab control (n = 15 per group). The mild-grade
model resulted in 63% survival with imipenem alone, 65% survival with
AZD9773 and 69% survival with DigiFab at 5 days. Thus, therapeutic
dosing of AZD9773 bid from 24 to 60 hours (schedule/dose/route as
previously) did not result in signi cantly di erent survival outcomes
versus either DigiFab or imipenem alone (n = 60 per group).
Conclusion These data demonstrate for the  rst time that TNFα
neutralization in a murine CLP model improves survival in a severe
sepsis setting. Moreover, contrasting with previous reports, TNF
suppression in mild-grade CLP models is not associated with increased
mortality. These  ndings support the hypothesis that AZD9773
has potential to be di erentiated from other anti-TNF agents as a
therapeutic intervention in sepsis.
Con icts of interest All authors are employees of AstraZeneca.
P42
E cacy of endotoxin absorption therapy on sepsis by polymyxin
B-attached  bers
KAtagi
Department of Critical Care Medicine, Osaka City General Hospital, Osaka,
Japan
Critical Care 2011, 15(Suppl 3):P42 (doi: 10.1186/cc10411)
Introduction Endotoxin plays a role in the development of Gram-
negative bacterial sepsis. In Japan, polymyxin B-attached  bers
(PMX-B) are used clinically as an endotoxin absorption therapy to
neutralize the biological activity of lipid A, the immunomodulatory
center of lipopolysaccharide (LPS) endotoxin. Because hemodynamic
improvement is not seen in all cases, it cannot be assumed that this
therapy will be e ective against all cases of sepsis.
Hypothesis Endotoxin absorption therapy is e ective against
abdominal infection. Moreover, the mortality rate signi cantly
improved in endotoxin-positive cases of abdominal infection.
Methods Between 1997 and April 2008, endotoxin absorption therapy
was performed on 105 septic patients in the ICU of Hyogo College
of Medicine and the Osaka City General Hospital. The 105 cases were
divided into an abdominal infection group (n = 45) and a nonabdominal
infection group (n = 60). Before and after therapy, the endotoxin level
was measured in patients using the limulus amoebocyte lysate (LAL) and
endotoxin activity assay (EAA) methods. Moreover, we measured blood
pressure, cardiac index, and the administered dose of catecholamine.
Using a retrospective analysis, we compared Sequential Organ Failure
Assessment (SOFA) scores; the Risk, Injury, Failure, Loss, and End stage
(RIFLE) criteria; and the 28-day survival rate between the two groups.
Results After the endotoxin absorption therapy, mean blood pressure
increased signi cantly from 67.9 ± 11.4 to 86.4 ± 6.3 mmHg in the
abdominal infection group, whereas there was no change in the
nonabdominal infection group. After the therapy, the SOFA scores and
RIFLE criteria improved in both groups, but they improved signi cantly
in the abdominal infection group. Patients in the abdominal infection
group, especially the endotoxin-positive cases, recovered earlier from
shock and had a signi cantly higher rate of survival than the abdominal
infection group.
Conclusion In endotoxin-positive patients with an abdominal infection,
absorption therapy improved survival rate and cardiac and renal
dysfunction due to sepsis or septic shock. However, further studies are
required to verify the e ectiveness of endotoxin absorption therapy.
P43
Lactate clearance as a simple bedside instrument to predict
short-term mortality of severe septic patients
WHambali1, KChen1,2, DWidodo1,2, EDewiasty1, HTPohan1,2, SSuwarto1,2
1Internal Medicine Department, Faculty of Medicine, University of Indonesia,
Jakarta, Indonesia; 2The Indonesian Society for the Study of Tropical Medicine
and Infectious Diseases, Jakarta, Indonesia
Critical Care 2011, 15(Suppl 3):P43 (doi: 10.1186/cc10412)
Introduction Severe sepsis is major health problem with a high
mortality rate, and still its incidence continues to rise [1-5]. Lactate
clearance, measurement of the lactate level at two consecutive times,
is an inexpensive and simple clinical parameter that can be obtained
by a minimally invasive means [6-8]. This parameter represents kinetic
alteration of the anaerobic metabolism that makes it a potential
parameter to evaluate disease severity and intervention adequacy.
Lactate clearance early in the hospital course may indicate a resolution
of global tissue hypoxia and is associated with improved outcome
[7-9]. Nevertheless, the relationship between lactate clearance and
short-term mortality in severe septic patients is still poorly understood.
Understanding the presence of confounder factors is also important
to strengthen the role of lactate clearance in the treatment of severe
septic patients.
Objective To evaluate the clinical course between lactate clearance
groups, and determine the role of confounder variables that in uence
its relationship.
Methods This is a prospective cohort study conducted in Ciptomangun-
kusumo Hospital, from March to May 2011. Patients were categorized
into the high lactate clearance group if there were di erences in 6-hour
lactate levels ≥10%, and conversely were categorized into the low
lactate clearance group [7,8]. Deaths were observed within the  rst
10 days. After data collection, the statistical methods were analyzed
using survival analysis. Analysis of confounder variables was performed
by multivariate Cox regression test.
Results During the research period there were 60 patients recruited,
consisting of 30 patients grouped into high lactate clearance and the
remainder grouped into low lactate clearance. The survival rates in
high and low lactate clearance groups were 60.0% versus 26.7% (see
Figure 1). In the low lactate clearance group the median survival was
3 days, while the mortality rate did not reach 50% in the high lactate
clearance group. The  rst interquartile was 1 day and 4 days. The hazard
ratio between groups was 2.87 (95% CI = 1.41 to 5.83). Steps taken to
analyze the role of variables that potentially act as confounder factors
Figure 1 (abstract P43). Kaplan–Meier curves between lactate clearance
groups.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S21

were by using bivariate analysis, in which variables that in uenced the
occurrence of deaths (indicated by P <0.25) underwent multivariate
analysis subsequently. On multivariate analysis the presence of septic
shock, degree of organ dysfunction, vasoactive drug usage, blood
transfusion, and  uid resuscitation change the hazard ratio by no
more than 10% (Table 1). For that reason, these parameters were not
considered as confounders.
Conclusion Severe septic patients with high lactate clearance have a
better survival rate compared with the low lactate clearance group,
and its relationship is not in uenced by the presence of confounder
variables.
Acknowledgments The authors thank the nurses and administrative
sta in the Division of Tropical Medicine and Infectious Diseases,
Department of Internal Medicine, Faculty Medicine, University of
Indonesia for their assistance in this study.
References
1. Moss M: Epidemiology of sepsis: race, sex, and chronic alcohol abuse. CID
2005, 41(Suppl 7):S490-S497.
2. Guidet B, Aegerter P, Gauzit R, Meshaka P, Dreyfuss D: Incidence and impact
of organ dysfunctions associated with sepsis. Chest 2005, 127:942-951.
3. O’brien JM Jr, Ali NA, Abraham E: Year in review in critical care, 2004: sepsis
and multi-organ failure. Crit Care 2005, 9:409-413.
4. Ely EW, Kleinpell RM, Goyette RE: Advances in the understanding of clinical
manifestations and therapy of severe sepsis: an update for critical care
nurses. Am J Crit Care 2003, 12:120-133.
5. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR:
Epidemiology of severe sepsis in the united states: analysis of incidence,
outcome, and associated costs of care. Crit Care Med 2001, 29:1303-1310.
6. Nguyen HB, Loomba M, Yang JJ, Jacobsen G, Shah K, Otero RM, Suarez A,
Parekh H, Jaehne A, Rivers EP: Early lactate clearance is associated with
biomarkers of in ammation, coagulation, apoptosis, organ dysfunction
and mortality in severe sepsis and septic shock. J In amm 2010, 7:6-17.
7. Nguyen HB, Rivers EP, Knoblich BP, Jacobsen G, Muzzin A, Ressler JA,
Tomlanovich MC: Early lactate clearance is associated with improved
outcome in severe sepsis & septic shock. Crit Care Med 2004, 32:1637-1642.
8. Arnold RC, Shapiro NI, Jones AE, Schorr C, Pope J, Casner E, Parrillo JE,
Dellinger P, Trzeciak S: Multicenter study of early lactate clearance as a
determinant of survival in patients with presumed sepsis. Shock 2009,
32:35-39.
9. Jones AE. Shapiro NI, Trzeciak S, Arnold RC, Claremont HA, Kline JA: Lactate
clearance vs central venous oxygen saturation as goals of early sepsis
therapy. JAMA 2010, 303:739-746.
P44
E ect of low-dose steroid on NF-κB and caspase-3 intestinal
expression in a sepsis mouse model
HAGuntur1,2, HPDiding1,2, HTPohan3, DWidodo3
1Indonesian Society for the Study of Tropical Medicine and Infectious
Diseases, Surakarta, Indonesia; 2Faculty of Medicine, Sebelas Maret University,
Surakarta, Indonesia; 3Indonesian Society for the Study of Tropical Medicine
and Infectious Diseases, Jakarta, Indonesia
Critical Care 2011, 15(Suppl 3):P44 (doi: 10.1186/cc10413)
Introduction The use of low-dose corticosteroids in sepsis early
stages is still debated. The association of LPS–LBP complexes to CD14
receptors and will interact with TLR4 to induce NF-κB as a signal and
transcription of proin ammatory cytokines [1,2]. Excessive production
of in ammatory cytokines will cause activation of SIRS, especially in
gut-associated lymphoid tissues [3], which induces metabolic changes
leading to apoptosis network, MOF, septic shock and death [3-5].
Changes in apoptosis are mediated by caspases, including caspase-3
that acts as an e ector caspase [6,7]. Low-dose corticosteroids can
inhibit the production of proin ammatory cytokines, production
of in ammatory mediators, and lower adhesion of leukocytes to the
endothelium [8].
Objective The aim of this study was to analyse NF-κB and caspase-3
intestinal expression, and also survival from use of low-dose steroid in
the early stages of sepsis in the Balb/C mouse model of sepsis.
Methods Male Balb/C mice were inoculated with lipopolysaccharide
for the sepsis mouse model. Sepsis mouse model grouping was to a
sepsis group (Group I) and to sepsis with steroid (methylprednisolone
1 to 1.5 mg/kg BW/day) (Group II). Detection of intestinal NF-κB and
caspase-3 expression used the immunohistochemistry technique on
days 1, 3, 5 and 7. Survival was seen until the 7th day. The two-tailed
Fisher exact test for the analysis of mortality, independent-sample t test
for intestinal NF-κB and caspase-3 expression, and P <0.05 were used to
determine signi cant di erences.
Results Acute in ammatory response occurs in the early stages of
sepsis (the  rst 5 days of exposure) and the process of death occurs in
advanced stages of sepsis (after the  rst 5 days of exposure) [9]. This
study shows that the use of low-dose corticosteroids in sepsis early
stages ( rst 5 days) signi cantly inhibited the expression of NF-κB (see
Table 1), so cytokine production of proin ammatory cytokines was not
excessive. Reduced product proin ammatory cytokines would reduce
the expression of intestinal caspase-3 (see Table 2), which will reduce
Table 1 (abstract P43). Variables that potentially act as confounder factors
Variable Nonsurvivors Survivors P value Adjusted HR (95% CI) HR change (%)
Septic shock within 6 hours
With septic shock 11 4 0.081 3.083 (1.505 to 6.317) 7.4
Without septic shock 23 22
Initial SOFA score
>9 11 3 0.038 2.919 (1.388 to 6.138) 1.7
≤9 19 21
Vasoactive drugs within 6 hours
Without vasoactive drugs 22 23 0.013 2.988 (1.462 to 6.106) 4.1
With vasoactive drugs 12 3
Invasive ventilation within 6 hours
a
Without mechanical ventilation 31 23 0.777 – –
With mechanical ventilation 3 3
PRC transfusion within 6 hours
Without transfusion 29 25 0.069 3.077 (1.493 to 6.340) 7.2
With transfusion 5 1
Fluid resuscitation within 6 hours
<1,000 cm
3
15 18 0.166 2.942 (1.444 to 5.994) 2.5
≥1,000 cm
3
19 8
HR, hazard ratio; SOFA, Sequential Organ Failure Assessment; PRC, packed red cells. aInvasive ventilation parameter not included in multivariate analysis, because
P<0.25.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S22

the excessive apoptosis in the intestinal tract. Decreased expression of
NF-κB and caspase-3 in the intestinal tract would further reduce the
excessive mucosal cell death. This situation will block the destruction
and disruption of mucosal defense function of the digestive tract,
thereby increasing immune response. The end result will be seen that
low-dose corticosteroids can reduce mortality. This study found dead
animals for Group I were 70%, while Group II were 10% (P = 0.020).
Conclusion Low-dose steroids can reduce NF-κB and caspase-3
intestinal expression and also mortality in early sepsis.
Acknowledgements The authors thank the Faculty of Medicine,
University of Sebelas Maret for their  nancial support and assistance
given in this study.
References
1. Hongwei Q, Cynthia AW, Sun JL, Xueyan Z, Etty NB: LPS induces CD40 gene
expression through the activation of NF-κB and STAT-1α in macrophages
and microglia. Blood 2005, 106:3114-3122.
2. Kristine MJ, Sarah BL, Anncatrine LP, Jesper EO, Thomas B: Common TNF-α,
IL-1β, PAI-1, uPA, CD14 and TLR4 polymorphisms are not associated with
disease severity or outcome from gram negative sepsis. BMC Infect Dis
2007, 7:108.
3. Arul MC, Markus HL, Chandan KS, Terrence RB, Sunita SS, Vidya JS, Vaishalee
AP, Peter AW: Molecular signatures of sepsis multiorgan gene expression
pro les of systemic in ammation. Am J Pathol 2001, 159:1199-1209.
4. Elena GR, Alejo C, Gema R, Mario D: Cortistatin, a new antiin ammatory
peptide with therapeutic e ect on lethal endotoxemia. J Exp Med 2006,
203:563-571.
5. Javier C, José Y, David HE, Yolanda M, Ruben M, Isabel A, Antonia M, Pascual P,
Vicente V: Role of lipopolysaccharide and cecal ligation and puncture on
blood coagulation and in ammation in sensitive and resistant mice
models. Am J Pathol 2005, 166:1089-1098.
6. Chung CS, Chaudry IH, Ayala A: The apoptotic response of the lymphoid
immune system to trauma, shock and sepsis. In Yearbook of Intensive Care
and Emergency Medicine. Editted by Vincent JL. Berlin: Spinger-Verlag;
2000:27-40.
7. Chung CS, Song GY, Lomas J, Simms HH, Chaudry IH, Ayala A: Inhibition of
Fas/Fas ligand signaling improves septic survival: di erential e ects on
macrophage apoptotic and functional capacity. J Leukoc Biol 2003,
74:344-351.
8. Annane D, Caillon DM: Corticosteroid in sepsis: from bench to bedside.
Shock 2003, 20:197-207.
9. Xiao H, Siddiqui J, Remick DG: Mechanisms of mortality in early and late
sepsis. Infection Immunity 2006, 74:5227-5235.
P45
Etiological agents of bacterial sepsis in a newly constructed medical
center in Saint Petersburg, Russia
EBarantsevich, NBarantsevich, NRybkova, IChurkina, NPestova,
MKarpenko
Almazov Federal Center of Heart, Blood and Endocrinology, Saint Petersburg,
Russia
Critical Care 2011, 15(Suppl 3):P45 (doi: 10.1186/cc10414)
Introduction Sepsis is one of the factors of high mortality in ICUs in
critically ill patients. Annual mortality from this condition is estimated
at 30 to 50 deaths per 100,000 population [1,2,4,5]. The aim of the
present study was to reveal the spectrum and resistance to antibiotics
of microorganisms, causing sepsis in hospitalized patients of the
multidisciplinary medical center, that accumulates patients from all
regions of Russia, during the  rst year from its foundation, with low
possibility of local nosocomial strains formation.
Methods The diagnosis of sepsis was veri ed by isolation of bacteria
from blood (only two or more positive results were considered) and the
presence of two or more criteria of systemic in ammatory response
syndrome [3]. The cultures were isolated from blood with BactAlert
(BioMerieux, France). The identi cation was performed by routine
phenotypic methods and sequencing of the 16sRNA gene (ABI Prism
3130, MicroSeq ID v2.0 Software, MicroSeq ID 16s rDNA500 Library
v2.0). Resistance to routinely used antibiotics was investigated with the
disc di usion method and by dilution techniques for MIC determination
on Muller–Hinton agar (Oxoid, UK).
Results Sepsis, associated with bloodstream infections, was revealed
in 89 cases – Gram-positive cocci predominated. Staphylococcus spp.
were responsible for 35 (39.3%) cases: Staphylococcus aureus was the
causative agent in eight (8.9%), coagulase-negative staphylococci
(Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylo-
coccus hominis novobiosepticus) in 27 (30.3%) infections. Other Gram-
positive cocci were Enterococcus faecalis in eight (8.9%), Enterococcus
faecium in six (6.7%). Gram-negative microorganisms included Acineto-
bacter baumannii that was found in 11 (12.4%), Klebsiella spp. (Klebsiella
pneumonia, Klebsiella rhinoscleromatis, Klebsiella oxytoca) in  ve
(5.6%), Escherichia coli in eight (8.9%), Enterobacter spp. (Enterobacter
cloaceae, Enterobacter aerogenes, Enterobacter hormaechea) in three
(3.4%), Pseudomonas aeruginosa in two (2.2%) patients with sepsis.
Rarely isolated bacteria were Bacillus thuringiensis in one (1.1%),
Stenotrophomonas maltophylia in one (1.1%), Pantoeae agglomerans in
one (1.1%), Corynebacterium mucifaciensis in one (1.1%) case. Two and
more species were isolated from blood in seven (7.9%) patients. In total,
105 strains were isolated in sepsis cases. Resistance to antibiotics was
observed in 96 (91.4%) bacterial isolates; 56 (53.3%) were multidrug-
resistant strains. All E. faecium, and nine (81.8%) strains of A. baumannii
were resistant to seven or more antibiotics. All E. faecium strains were
susceptible to linezolide, A. baumannii to tigecycline. Methicillin
resistance was detected in two (15.4%) strains of S. aureus and 18
(60.0%) strains of coagulase-negative staphylococci; four (66.7%)
strains of E. faecium were vancomycin resistant.
Conclusion Gram-positive bacteria were the leading causative agents
of sepsis, associated with bloodstream infections in the newly-
constructed hospital in Saint Petersburg. The majority of strains (91.4%)
were resistant to antibiotics, and more than half of the isolates were
multidrug resistant. Methicillin resistance was observed predominantly
in coagulase-negative staphylococci, vancomycin resistance in
E. faecium. All polyresistant E. faecium strains were susceptible to
linezolide, A. baumannii to tigecycline.
References
1. Antonacci Carvalho PR, Trotta EA: Advances in sepsis diagnosis and
treatment. J Pediatria 2003, 79:S195-S204.
2. Braun L, Riedel A A, Cooper LM: Severe sepsis in managed care: analysis of
incidence, one-year mortality, and associated costs of care. J Managed Care
Pharmacy 2004, 10:521-530.
3. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, Schein RM,
Sibbald WJ: ACCP/SCCM consensus conference: de nitions for sepsis and
organ failure and guidelines for the use of innovative therapies in sepsis.
Chest 1992, 101:1644-1655.
4. Dellinger RP, Carlet JM, Masur H, Gerlac H, Calandra T, Cohen J, Gea-
Banacloche J, Keh D, Marchall JC, Parker MM, et al.: Surviving Sepsis
Table 1 (abstract P44). NF-κB intestinal expression
Amount of cell expression
Day Group I Group II P value
1 17.5 ± 4.7 10.8 ± 3.6 0.019
3 26.5 ± 4.4 15.8 ± 3.6 0.001
5 39.3 ± 4.1 31.2 ± 7.0 0.033
7 54.8 ± 9.6 50.5 ± 10.7 0.476
Data presented as mean ± standard deviation.
Table 2 (abstract P44). Caspase-3 intestinal expression
Amount of cell expression
Day Group I Group II P value
1 8.5 ± 2.9 4.3 ± 1.9 0.014
3 14.7 ± 3.1 9.8 ± 2.2 0.012
5 33.2 ± 8.3 12.8 ± 4.5 0.000
7 42.3 ± 3.2 37.7 ± 8.2 0.336
Data presented as mean ± standard deviation.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S23

Campaign guidelines for management of severe sepsis and septic shock.
Crit Care Med 2004, 32:858-873.
5. Lever A, Mackenzie I: Sepsis: de nition, epidemiology, and diagnosis. BMJ
2007, 335:879-883.
P46
E cacy and safety of gelatin for  uid therapy in hypovolemia:
asystematic review and meta-analysis
CSHartog1,2, VVlasakov1, DOThomas-Rueddel1,2, HRueddel1,2,
RHutagalung1, KReinhart1,2
1Department of Anesthesiology and Intensive Care Medicine, Jena University
Hospital, Jena, Germany; 2Center for Sepsis Control & Care, Jena University
Hospital, Jena, Germany
Critical Care 2011, 15(Suppl 3):P46 (doi: 10.1186/cc10415)
Introduction Gelatin is frequently used as volume expander. There are
growing concerns about safety.
Objective To systematically assess clinical evidence concerning
mortality, coagulation and renal function.
Methods Systematic review of randomised controlled trials (RCT)
on gelatin in hypovolemia in comparison to any other  uid with
a comprehensive search strategy (Ovid Medline (1948 to May
2011), EMBASE (1947 to May 2011), Cochrane Library). Data were
independently extracted and risk of bias assessed using the 2010
Cochrane tool. Primary outcome was overall mortality. Secondary
outcomes were the number of patients exposed to allogeneic
transfusion, frequency of renal replacement therapy (RRT) or acute
renal failure (ARF). Albumin and crystalloid solutions were de ned as
suitable, and other synthetic colloids as unsuitable control  uids since
they carry similar risk of side e ects. Relative risks (RR) and weighted
mean di erences with 95% CIs were calculated. Data were pooled
using a random-e ects model (RevMan 5.1, Cochrane Collaboration).
Results The search yielded 1,288 citations, 210 reports were read in
full. The  nal sample contained 72 RCT in English, German, French and
Italian, published between 1975 and 2010, with 5,915 patients overall,
2,523 of which received gelatin. The median sample size in the gelatin
groups was 20 patients (range 10 to 249). In 53 RCT (74%), the study
period was ≤24.0 hours. Total gelatin dose was 20 ml/kg (median,
range 6 to 62). Only 38 RCT (53%) used suitable control  uids. Forty-
nine RCT (68%) investigated elective surgical patients, mostly from
cardiac surgery (32 RCT, 44%). Nine RCT (13%) investigated critically ill
patients, six RCT (8%) were in emergency patients and seven RCT (10%)
were in children. The RR for mortality was 1.02 (CI 0.87 to 1.19, data
from 23 RCT with 2,694 patients which reported mortality). Numbers of
patients exposed to allogeneic transfusions were provided in 11 RCT,
n= 1,148 patients and the RR was 1.16 (0.94 to 1.44). When only studies
with suitable control  uids were included, the RR for mortality was 1.13
(0.88 to 1.46, 10 RCT, 1,392 patients) and risk for transfusion exposure
was 1.35 (0.88 to 2.08, seven RCT, n = 672), tending towards control.
Only six RCT (n = 662 patients) reported the occurrence of RRT or ARF,
 ve of them in comparison with HES solutions. Three RCT reported
anaphylactoid events.
Conclusion Most published studies on gelatin are small and short-
time, use unsuitable control  uids and report too few events to reliably
assess the safety of gelatin.
Cite abstracts in this supplement using the relevant abstract number, e.g.:
Hartog CS, et al.: E cacy and safety of gelatin for  uid therapy in
hypovolemia: a systematic review and meta-analysis [abstract]. Critical Care
2011, 15(S upp l 3):P46.
Critical Care 2011, Volume 15 Suppl 3
http://ccforum.com/supplements/15/S3
S24