Journal of Critical Care (2008) 23, 455460

A Focus on Sepsis

Implementation of the Surviving Sepsis Campaign

guidelines for severe sepsis and septic shock:
We could go faster
Massimo Zambon MD a , Marcello Ceola MD b , Roberto Almeida-de-Castro MD a ,
Antonino Gullo MD a , Jean-Louis Vincent MD, PhD b,
a

Department of Anesthesiology and Intensive Care, Cattinara Hospital, University of Trieste, Italy 447-34149
Department of Intensive Care, Erasme Hospital, Free University of Brussels, 1070 Brussels, Belgium

Keywords:
Sepsis bundle;
Intensive care unit;
Septic shock

Abstract
Purpose: The aim of this study is to evaluate the feasibility of applying sepsis bundles in the intensive
care unit (ICU) and their effect on outcomes.
Methods: In this prospective, observational study in a 31-bed capacity department of intensive care, we
measured the time taken to perform sepsis bundle interventions in 69 consecutive patients with severe
sepsis or septic shock.
Results: Compliance with the 6-hour bundle was obtained in 44 (72%) of 61 patients; these patients had
a lower mortality rate (16% vs 41%, P = .04) and shorter ICU stay (median [range], 5 [3-10] vs 9 [6-19]
days, P = .01) than other patients. Compliance with the 24-hour bundle was obtained in 30 (67%) of 44
eligible patients. The mortality rate and duration of ICU stay were not significantly lower in the 24-hour
compliant as compared with the noncompliant group (23% vs 33% and 6 [4-11] vs 9 [6-25] days,
respectively; P value is not significant). Patients who complied with the 24-hour sepsis bundle after
only 12 hours had a lower mortality rate (10% vs 39%, P = .036) and shorter stay (6 [4-10] vs 9 [6-25]
days, P = .055) than those who were compliant after 24 hours.
Conclusions: Correct application of the sepsis bundles was associated with reduced mortality and
length of ICU stay. Earlier implementation of the 24-hour management bundle could result in
better outcomes.
2008 Elsevier Inc. All rights reserved.

1. Introduction
Severe sepsis and septic shock are major problems in
critically ill patients. In the recent European Sepsis Occur Corresponding author. Tel.: +32 2 555 3380; fax: +32 2 555 4555.
E-mail address: jlvincen@ulb.ac.be (J.-L. Vincent).
0883-9441/$ see front matter 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcrc.2007.08.003

rence in Acutely ill Patients study [1], which collected data

on septic adult patients in 198 intensive care units (ICUs)
from 24 European countries, ICU mortality was 32% for
patients with severe sepsis and 54% when septic shock was
present. Recent guidelines provided by the Surviving Sepsis
Campaign (SSC) [2] provide an important tool for the
management of patients with severe sepsis and septic shock.
The next phase of the SSC involves the practical application

456
of the guidelines and assessment of their effectiveness in
improving outcomes, with the final aim being able to
decrease the relative mortality from sepsis by 25% over 5
years. For this purpose, and to simplify the complex
application of the guidelines, the concept of the sepsis
bundle has been developed [3], a bundle being a group of
interventions related to a disease process that, when executed
together, result in better outcomes than when implemented
individually. Two bundles have currently been proposed: a
6-hour resuscitation bundle and a 24-hour management
bundle (http://www.ihi.org/IHI/Topics/CriticalCare/Sepsis)
(Table 1).
A recent study by Gao et al [4] showed that
compliance with the 6-hour sepsis bundle was associated
with a more than 2-fold decrease in hospital mortality
(49% vs 23%), and compliance with the 24-hour bundle
showed a trend to reduced mortality; however, compliance
with both bundles was not very high (52% for the 6-hour
bundle and 30% for the 24-hour bundle). These results
raise several important questions. First, what compliance
rates can be achieved for each bundle? Second, is it
possible to reduce the time delays for the various bundle
components, and if so, does it make a difference? Third,
do some interventions affect the feasibility of the bundles
more than others do?

Table 1 Sepsis bundles http://www.ihi.org/IHI/Topics/

CriticalCare/Sepsis
Sepsis resuscitation bundle (6-h bundle)
1. Serum lactate measured
2. Blood cultures obtained before antibiotic administration
3. From the time of presentation, broad-spectrum antibiotics
administered within 3 h for ED admissions and 1 h for nonED ICU admissions
4. In the event of hypotension and/or lactate N4 mmol/L
(36 mg/dL)
Deliver an initial minimum of 20 mL/kg of crystalloid
or colloid equivalent)
Apply vasopressors for hypotension not responding to
initial fluid resuscitation to maintain mean arterial pressure
N65 mm Hg
5. In the event of persistent hypotension despite fluid
resuscitation (septic shock) and/or lactate N4 mmol/L
(36 mg/dL)
Achieve central venous pressure of N8 mm Hg
Achieve central venous oxygen saturation (ScVO2) of N70%
Sepsis management bundle (24-h bundle)
1. Low-dose steroids administered for septic shock in
accordance with a standardized ICU policy
2. Drotrecogin alfa (activated) administered in accordance with
a standardized ICU policy
3. Glucose control maintained greater than lower limit of
normal but b150 mg/dL (8.3 mmol/L)
4. Inspiratory plateau pressures maintained b30 cm H2O for
mechanically ventilated patients

M. Zambon et al.
The aim of this study was to evaluate whether the bundle
approach to sepsis management is reliable and easy to
implement in a large department of intensive care, and if its
implementation can have an impact on outcome. We also
investigated the feasibility of individual interventions and
whether earlier application of some interventions may be
associated with improved outcomes.

2. Methods
2.1. Patients
The study was conducted in the medical-surgical department of intensive care of Erasme University Hospital; this
department has 31 beds and a 4-bed stabilization room
(shock lab). All consecutive adult (18 years old) patients
admitted between May 1 and November 30, 2005, with a
diagnosis of severe sepsis or septic shock were enrolled.
Severe sepsis was defined as the presence of an infection
with at least one organ dysfunction, and septic shock as fluidrefractory hypotension (mean arterial pressure, b65 mm Hg)
requiring the administration of vasopressor agents [5].
Patients who developed sepsis after admission were not
included, and patients were excluded if there was a decision
to limit therapy (do not resuscitate order). The study was
approved by the ethics committee of the hospital, which
waived the need for informed consent, in view of the purely
observational nature of the study.

2.2. Data collection

All data were recorded by a research fellow not involved
in the care of the patient. For each patient, demographic
characteristics, such as age, sex, main diagnosis, comorbidities, and source of infection, were obtained, and the
sequential organ failure assessment (SOFA) score and
acute physiology and chronic health evaluation II score
were calculated in the first 24 hours after admission.
We considered as time 0 the time of admission to the
ICU for the hospitalized patients, or to the emergency
department (ED), and calculated the time delay of the
various interventions from that time. The time of insertion
of arterial and central venous catheters (CVCs) and the
time of blood sampling for lactate, blood cultures, and
arterial and venous blood gas analyses were recorded.
Collected data included mean arterial pressure, central
venous pressure, heart rate, respiratory rate, arterial blood
gas values, blood glucose and lactate concentrations,
central venous oxygen saturation (ScVO2), type and amount
of fluids administered, vasopressors and inotropes administered, and urine output. Ventilatory parameters were also
recorded in mechanically ventilated patients. Times to
administration of first antibiotic(s), steroids, and activated
protein C infusion were also recorded. The most likely

SSC guidelines for severe sepsis and septic shock

457

2.4. Statistical analysis

The data are presented as means SD or median (with
interquartile ranges), unless otherwise indicated. We
assessed compliance for each single intervention (as the
percentage of patients in whom the intervention was
performed within the right time frame) and the compliance
for the entire bundle (all or none). 2 tests were used to
compare differences in proportions between groups. Group
mean or median values were compared using Student t test or
Mann-Whitney U test. A P value of less than .05 was
considered statistically significant.
All statistical analyses were performed using the SPSS
software (version 13.0; SPSS Inc, Chicago, Ill).

Fig. 1

Study schema.

sources of infection and the type(s) of microorganism(s)

isolated were noted.
Data were recorded for the first 24 hours; for outcomes
and microbiology, patients were followed until discharge or
death in the ICU. Outcome measures were defined as ICU
mortality and length of stay.

3. Results
The study included 69 consecutive patients with severe
sepsis or septic shock, 8 of whom were excluded because of
missing data (Fig. 1). The baseline characteristics of the
remaining 61 patients are shown in Table 2. Thirty-six (59%)
patients had septic shock and needed vasopressors; 44 (72%)
patients were candidates for the management bundle. The
overall ICU mortality rate was 23% (14/61); the hospital
mortality rate was 30% (18/61).

2.3. Sepsis bundles and compliance

3.1. Resuscitation bundle
The sepsis bundles used, presented in Table 1, were those
that are currently recommended (http://www.ihi.org/IHI/
Topics/CriticalCare/Sepsis).
We considered as compliant those patients in whom
all the interventions described in the bundle were
obtained during the predetermined time interval (first 6
hours for the resuscitation bundle, first 24 hours for the
management bundle). For antibiotic administration,
patients were not divided according to ED or non-ED
admission, and antibiotic administration within 3 hours
was considered as compliant. When continuous ScVO2
monitoring was not available, we considered ScVO2
measurements to guide therapy as being compliant with
the bundle. If one or more interventions were performed
after the proposed time limit, the patient was included in
the noncompliant group.
Patients with persisting severe sepsis after 6 hours were
considered candidates for the 24-hour management
bundle. If in the first 24 hours a patient who was a candidate
to receive activated protein C did not receive it because of
contraindications, and this was clearly written in the patient
notes or discussed during rounds, we considered this
component as having been assessed, and the patient was
considered compliant for this intervention. Compliance with
the blood glucose control component was considered as
having been achieved if 2 consecutive values at 24 hours
were less than 150 mg/dL.

The 6-hour bundle was correctly applied in 44 (72.1%) of

61 patients. Of the 44 compliant patients, 7 died (mortality
rate 16%) vs 7 of 17 (mortality rate 41%) in the
noncompliant group (P = .047). The odds ratio for death
was 3.7 (confidence interval [CI], 1.05-13.03). The ICU
length of stay was shorter in the compliant than in the

Table 2 Demographic characteristics, disease severity, and

mortality rates for the 61 patients
Age (mean SD)
Sex, male/female
Medical/Surgical
Source of infection:

Shock at admission
SOFA score, mean SD (range)
APACHE II score, mean SD
ICU mortality
Hospital mortality

69 12 y
35:26
39:22
Lung (22, 36%)
Urinary (16, 26%)
Abdominal (15, 25%)
Others (5, 8%)
Unknown (3, 5%)
36 (59%)
7.5 2.9 (3-17)
20.6 5.4
14 (23%)
18 (30%)

APACHE II indicates acute physiology and chronic health evaluation II.

458
Table 3

M. Zambon et al.
Characteristics and time to achieve the individual interventions in the 6-hour bundle for complaint and noncompliant groups
All patients

6-h compliance

Patients, n (%)
61
Age (y)
69 12
Shock at admission, n (%)
36 (59)
SOFA score
7.5 2.9 (3-17)
ICU deaths, n (%)
14 (23)
Length of stay (d)
7 (4-12)
Timing of intervention, median (IQ range), min
Lactate measured
22 (10-58)
Blood cultures drawn
50 (30-80)
Antibiotic started
ED admissions
146 (88-223)
Non-ED admissions
34 (0-74)
Arterial catheter inserted
40 (25-60)
CVC inserted
42 (30-95)
ScVO2 measured
70 (50-157)

noncompliant patients (median [range], 5 [3-10] vs 9 [6-19]

days, P = .010) (Fig. 2).
The median times to achieve interventions are shown
in Table 3. Time delays for CVC insertion, ScVO2
measurement, and first antibiotic administration were
significantly shorter in the compliant than in the
noncompliant group. In hypotensive patients and in
patients with a blood lactate concentration greater than 4
mmol/L (36 mg/dL), the total amount of fluid administered after 6 hours was always higher than recommended
in the bundle (mean of 30 8 mL/kg of crystalloids and
the equivalent in colloids).

3.2. Management bundle

The management bundle was correctly applied within
24 hours in 30 (68%) of the 44 patients with severe sepsis
persisting after 6 hours. This compliant group had a
mortality of 23% (6/30) vs 33% (5/14) in the noncompliant
group (P = .29); the odds ratio was 2.22 (CI, 0.54-9.13).
The median length of ICU stay was 6 days (range, 4-11) in
the compliant vs 9 days (range, 6-25) in the noncompliant
group (P = .19).

Table 4

6-h noncompliance

44 (72)
68 12
25 (57)
7.4 2.9
7 (16)
5 (3-10)

17 (38)
69 13
11 (65)
7.8 2.9
7 (41)
9 (6-19)

.77
.57
.64
.047
.010

20 (10-50)
50 (27-75)

30 (16-77)
60 (30-142)

.23
.17

134 (83-159)
39 (0-52)
40 (26-57)
36 (30-60)
67 (42-100)

182 (115-325)
46 (0-136)
42 (23-93)
150 (30-364)
361 (50-420)

.02
.24
.48
.009
.030

Of the 36 patients in septic shock, 32 (89%) received

corticosteroids; 3 of these had been receiving long-term
steroid therapy, and this was not considered as an
intervention related to their septic shock state. Of the 4
other patients who did not receive steroids, 2 did not receive
steroids because of rapid improvement in their condition, and
for the other 2, no reason was given. Activated protein C was
given to 30 (68%) patients; of the other 14 patients, 10 had
contraindications related to an increased risk of bleeding.
Blood glucose concentrations at 24 hours were less than
150 mg/dL in 39 (89%) of the 44 patients. Thirty-one
patients were still being treated by mechanical ventilation
after 24 hours; 11 had acute lung injury or acute respiratory
distress syndrome. None of these patients had a plateau
pressure higher than 30 cm H2O.
In an explorative analysis, patients in whom compliance
with the 24-hour management bundle was already obtained
after 12 hours had a lower mortality rate than those who did
not comply by 12 hours (10% vs 39%, P = .036); the odds
ratio was 6.11 (CI, 1.14-32.78). The length of ICU stay was
also shorter (6 [4-10] vs 9 [6-25] days, P = .055). The
characteristics and results for the 24- and 12-hour groups are
shown in Table 4.

Characteristics of patients in the 24-hour management bundle

No. of patients, n (%)

Age (y)
Shock at admission, n (%)
SOFA score
ICU deaths, n (%)
Length of stay (d)

All patients

24-h
compliance

24-h
noncompliance

44
69 12
33 (75)
8.0 3.0 (4-17)
11 (25)
8 (4-14)

30 (68)
68 10
22 (73)
7.5 (6-9)
6 (20)
6 (4-11)

14 (32)
74 13
11 (79)
8.5 (7-14)
5 (36)
9 (6-25)

.095
1.00
.10
.287
.193

12-h
compliance

12-h
noncompliance

21 (48)
67 10
16 (76)
8 (6-9)
2 (10)
6 (4-10)

23 (52)
72 12
17 (74)
8 (6-9)
9 (39)
9 (6-25)

.133
1.00
.695
.036
.055

SSC guidelines for severe sepsis and septic shock

Fig. 2 Length of stay and compliance with the 6-hour bundle (left
panel) and the 24-hour bundle (right panel). Data are shown as
boxes (interquartile range); line represents the median.

We could not find any significant correlation between the

time delay of any single intervention and ICU mortality (data
not shown).

4. Discussion
The primary purpose of this study was to evaluate the
feasibility of applying the sepsis bundles. Although we found
that both the resuscitation and the management bundle could
be easily applied, the 6-hour bundle was not applied in 28%
of the patients and the 24-hour bundle in 32% of patients.
The second purpose of our study was to analyze which
interventions, in particular, were delayed. The variables that
seemed to influence compliance with the resuscitation bundle
were the time of insertion of the CVC, ScVO2 measurement,
and the time delay to start antibiotic therapy. We found a very
high compliance for glucose control with intensive insulin
therapy in the first 24 hours. Strict blood glucose control is
debatable in view of recent data from the multicenter VISEP
trial [6], but the maintenance of blood glucose to less than
150 mg/dL is probably a reasonable goal [7]. Maintaining the
plateau pressure did not represent a major problem.
Several studies have supported the feasibility of implementing evidence-based protocols for the treatment of septic
shock and suggested improved mortality rates with this
approach [8-10]. In a recent study, Trzeciak et al [8]
demonstrated that early goal-directed therapy (EGDT) end
points can reliably be achieved in septic patients treated in
the ED. However, they did not consider the feasibility of
the bundle as an all or none phenomenon. Moreover, as the

459
median time to achieve some of the EGDT end points in
their study was around 6 hours, it is reasonable to suppose
that the bundle was not completed in many patients in the
time recommended.
In our study, compliance with the bundles was higher than
previously reported. Gao et al [4] reported compliance with
the resuscitation bundle in 52% of patients and with the
management bundle in 30% of patients; mortality was
significantly reduced when 6-hour compliance was obtained,
but not for the 24-hour compliance. Similarly, we noted a
significant reduction in mortality in the group of patients in
whom the resuscitation bundle was achieved, and we also
found a significant reduction in the length of stay in the
compliant group. A similar trend was observed for the
management bundle, although the differences were not
statistically significant, probably because of the small
number of patients. Importantly, the 2 groups had similar
baseline characteristics, including age, presence of shock on
admission, and severity scores on admission.
We could not find any significant correlation between the
time delay to perform any single intervention and ICU
mortality. This reinforces the usefulness of the concept of a
bundle as a group of interventions that can improve outcome
if executed together, whereas single interventions do not
seem to have a strong impact.
The third aim of the study was to evaluate whether some
interventions could be applied earlier than recommended.
There is evidence in the literature to support the early use of
antibiotics, early resuscitation, and the early administration
of drotrecogin alfa (activated). In a recent study, Kumar et al
[11] reported that antimicrobial administration within the
first hour of hypotension in patients with septic shock was
associated with increased survival to hospital discharge.
Rivers et al [12] showed that resuscitation with EGDT
improved outcomes in patients with severe sepsis and septic
shock. In the recent ENHANCE study [13], patients who
received drotrecogin alfa (activated) in the first 24 hours had
a better outcome than those who started the drug infusion
after the first 24 hours. In our study, we noted a lower ICU
mortality and a shorter length of ICU stay in patients who
complied with the management bundle, which includes
administration of drotrecogin alfa (activated) and corticosteroids, within 12 hours rather than the recommended 24
hours. Indeed, corticosteroid administration is in the 24-hour
(management) bundle, but the same SSC guidelines [2]
suggest that corticosteroids should be administered if the
patient is still hypotensive despite fluid resuscitation and
vasopressors. Our findings would suggest that an earlier use
may be beneficial.
Our study has some limitations. First, it was conducted in a
single, large, well-staffed department of intensive care, and
the results may not be readily applicable to every institution;
in addition, we benefit from a shock lab, which may improve
the speed with which some interventions are achieved.
Second, the study included a relatively small number of
patients. Third, to evaluate the timing of interventions, we had

460
to select a time point and chose to use the time of admission to
the ICU or ED, although we are aware that this may not
accurately represent the time of diagnosis or onset of sepsis.
On the basis of these observations, we believe that the
guidelines could be implemented more rapidly. Fluid
resuscitation, blood cultures, antibiotic administration, and
serum lactate concentration measurement are interventions
that should be initiated very rapidly, certainly within the first
hour. If the patient continues to have severe sepsis after the
resuscitation phase (first 6 hours), in the following 6 hours,
we should consider steroids and drotrecogin alfa (activated)
infusion, thus effectively advancing the management bundle
by 12 hours. In this study, we found this faster management
to be feasible, and it was associated with better outcomes.

5. Conclusions
In this observational study, we found both the resuscitation and the management bundle feasible and easy to apply
at the bedside. Compliance with the 6-hour bundle was
associated with a reduction in ICU mortality and length of
stay, whereas for the 24-hour bundle, a stricter time limit for
the interventions (12 hours) could result in a better impact
on outcome.

References
[1] Vincent JL, Sakr Y, Sprung CL, et al. Sepsis in European intensive care
units: results of the SOAP study. Crit Care Med 2006;34:344-53.

M. Zambon et al.
[2] Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign
guidelines for management of severe sepsis and septic shock. Crit Care
Med 2004;32:858-73.
[3] Levy MM, Pronovost PJ, Dellinger RP, et al. Sepsis change bundles:
converting guidelines into meaningful change in behavior and clinical
outcome. Crit Care Med 2004;32:S595-7.
[4] Gao F, Melody T, Daniels DF, et al. The impact of compliance with 6hour and 24-hour sepsis bundles on hospital mortality in patients with
severe sepsis: a prospective observational study. Crit Care 2005;9:
R764-70.
[5] Levy MM, Fink MP, Marshall JC, et al. 2001 SCCM/ESICM/ACCP/
ATS/SIS International Sepsis Definitions Conference. Intensive Care
Med 2003;29:530-8.
[6] Brunkhorst FM, Kuhnt E, Engel C, et al. Intensive insulin therapy in
patient with severe sepsis and septic shock is associated with an
increased rate of hypoglycemiaresults from a randomized multicenter study (VISEP). Infection 2005;33(Suppl 1):19 [abst].
[7] Malhotra A. Intensive insulin in intensive care. N Engl J Med
2006;354:516-8.
[8] Trzeciak S, Dellinger RP, Abate NL, et al. Translating research to
clinical practice: a 1-year experience with implementing early goaldirected therapy for septic shock in the emergency department. Chest
2006;129:225-32.
[9] Kortgen A, Niederprum P, Bauer M. Implementation of an evidencebased standard operating procedure and outcome in septic shock.
Crit Care Med 2006;34:943-9.
[10] Shapiro NI, Howell MD, Talmor D, et al. Implementation and
outcomes of the Multiple Urgent Sepsis Therapies (MUST) protocol.
Crit Care Med 2006;34:1025-32.
[11] Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before
initiation of effective antimicrobial therapy is the critical determinant
of survival in human septic shock. Crit Care Med 2006;34:1589-96.
[12] Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in
the treatment of severe sepsis and septic shock. N Engl J Med
2001;345:1368-77.
[13] Vincent JL, Bernard GR, Beale R, et al. Drotrecogin alfa (activated)
treatment in severe sepsis from the global open-label trial ENHANCE.
Crit Care Med 2005;33:2266-77.