S e v e re S e p s i s an d S e p t i c

S h o c k Tr i a l s ( P ro C E S S ,
A R I S E , P ro M I S e )
What is Optimal Resuscitation?

Tiffany M. Osborn, MD, MPH

KEYWORDS
 Severe sepsis  Septic shock  Resuscitation  Usual care  Lactate
 Hyperlactatemia  Sepsis

KEY POINTS
 Three large international randomized trials (Process, ARISE and Promise) confirmed that
in the general population of patients with severe sepsis and septic shock, early goal-
directed therapy did not confer a mortality benefit compared with usual resuscitation.
The ability to generalize depends on the consistency of treatment provided as part of
usual resuscitation in individual hospitals.
 All 3 trials used the established definitions for identifying septic patients. Until the
SEPSIS-3 definitions are prospectively evaluated, their associated risks and benefits
are unclear.
 Usual care in all 3 trials included early identification using standardized screening proto-
cols, including lactate measurement, early intravascular fluid administration, and early
antibiotics.
 Normotensive patients with lactate level greater than or equal to 4 mmol/L had a similar
mortality to patients with refractory hypotension with a normal lactate level.

INTRODUCTION

Between 2014 and 2015, 3 independent, multicenter, government-funded, random-

ized controlled trials (RCTs) evaluating early goal-directed therapy (EGDT) were pub-
lished. These trials were Protocolized Care for Early Septic Shock (ProCESS) from the
United States,1 Australasian Resuscitation in Sepsis Evaluation (ARISE),2 and Proto-
colised Management in Sepsis (ProMISe)3 in the United Kingdom.
The care of septic patients has progressed significantly over the years; mortality has
significantly decreased but there is still much controversy and confusion. Which defi-
nition should be used to identify septic patients? Given that there was no survival

Section of Acute and Critical Care Surgery, Washington University School of Medicine, 4901
Forest Park Avenue, St. Louis, MO 63108, USA
E-mail address: [email protected]

Crit Care Clin 33 (2017) 323–344

http://dx.doi.org/10.1016/j.ccc.2016.12.004 criticalcare.theclinics.com
0749-0704/17/ª 2017 Elsevier Inc. All rights reserved.
324 Osborn

benefit of EGDT compared with usual resuscitation, how should usual resuscitation be
defined in the context of the data? What interventions do clinicians need to provide
and what is time dependent? This article reviews key findings of the 3 sepsis trials
and reviews the following:
 Background on sepsis care before the original EGDT trial by Rivers and
colleagues.4
 Key elements of ProCESS, ARISE, and ProMISe to assist comprehensive
evaluation.
 Options for operationalization and future direction in the evolving care of septic
patients.

Background
Death is not the enemy but occasionally needs help with timing.
—Peter Safar

Sepsis mortality before 2001 was traditionally high, with reports of mortality ranging
between 30% and 60%.5–9 There was no universal concept of urgency in the treatment
of septic patients. Care was generally fractured, with little collaboration between the
service line silos of the prehospital service, emergency department (ED), the intensive
care unit (ICU), and the wards. Treatment might focus on using vasopressors to
augment blood pressure with less emphasis on end-organ perfusion, resulting in
ischemic limbs and colloquial names such as “leave ’em dead” for levophed.10 In addi-
tion, universal use of ultrasonography in EDs or ICUs was nonexistent during this period.
After observing severe sepsis and septic shock mortality of 50% in local hospitals,
an institutional quality improvement initiative led to a randomized controlled trial eval-
uating EGDT from 1997 to 2000.4 In 2001, Rivers and colleagues4 reported results of a
new, protocolized resuscitation termed EGDT. EGDT was described as a structured
treatment protocol that incorporated elements consistent with consensus guide-
lines.11 EGDT is designed to optimize tissue oxygen transport through early identifica-
tion and time-dependent hemodynamic optimization of oxygen delivery using
continuous monitoring of prespecified physiologic targets.
 Preload: central venous pressure (CVP) was used as a surrogate target for intra-
vascular volume.
 Afterload: mean arterial pressure (MAP) was targeted after volume repletion with
vasoactive agents.
 Contractility and oxygen carrying capacity: central venous oxygen saturation
(ScvO2) guided delivery of inotropes and red blood cell transfusions.
At the time of publication, this protocol was novel because of the absolute mortality
reduction, the time dependency element, and focus on the level of care rather than
location of care. The absolute mortality benefit of 16% (46.5% to 30.5%) suggested
that this was one of the most effective modalities to date.12–14 Subsequent observa-
tional studies supported a mortality benefit of varying degrees.15
Although medicine traditionally functioned in silos, this study emphasized the level
of care rather than location of care. During this period of time, central venous access,
arterial access, and use of inotropes was generally reserved for the ICU. It was unique
to provide several critical care modalities in locations outside of traditional critical care
settings. Although novel in application, the idea was not a unique concept. Dr Peter
Safar16 described critical care as a continuum beginning prehospital, continuing
with ED intervention, and culminating in ICU admission and management.
 Severe Sepsis and Septic Shock 325

Also unique to the era, this study emphasized deliberate, time-dependent recogni-
tion and management of septic patients during the most proximal phase of hospital
presentation in the ED. As part of the effort to reduce global mortality caused by se-
vere sepsis and septic shock, EGDT was incorporated into the first iterations of the in-
ternational management guidelines of the Surviving Sepsis Campaign (SSC).12–14,17
Given that half of the estimated cases of severe sepsis and septic shock originally pre-
sent to the ED,18 the American College of Emergency Physicians representative
worked on the guidelines writing committee, endorsed all iterations of the guidelines,
and actively supported their distribution and implementation.19
Despite incorporation into the SSC guidelines and nonrandomized trials reporting
benefit, adoption was not universal.20,21 Concerns regarding universal adoptions
centered on the ability to generalize, protocol complexity, high control arm mortality,
and financial and infrastructure implications.
To address these concerns, ProCESS, ARISE, and ProMISe, three international, in-
dependent, multicentered, government-funded trials were undertaken. Principal
investigators decided before enrollment to reconcile key protocol design and opera-
tional elements to facilitate combining data on completion for a patient-level meta-
analysis and further evaluation of specific components that would not be possible
based on single trials alone.22

Shock is a rude unhinging of the machinery of life.

—Samuel D. Gross (A System of Surgery)23

PATIENT EVALUATION OVERVIEW

Identification/Inclusion Criteria
Identification is inherently linked to how the patient population is defined (Table 1). All
3 trials used the same established definitions,24 in combination with lactate level
(4 mmol/L) as did the original EGDT trial to identify adult ED patients with severe
sepsis or septic shock (Fig. 1).

Fluid Refractory Hypotension

Fluid refractory hypotension was originally defined in ProCESS as low blood pressure
after a 30-mL/kg fluid bolus, which was later changed to the same threshold, a 1-L
fluid challenge, used by both ProMISe and ARISE. The time of fluid bolus administra-
tion was more than 30 minutes in ProCESS and 60 minutes in ARISE and ProMISe (see
Table 1).

Increased Lactate level

Increased lactate level (>2 mmol/L) is an established indicator of critical illness and
predictor of decreased survival.25–31 Mortality reportedly increases with increasing
hyperlactatemia similar to a dose-response relationship.31 Further, normotensive pa-
tients with a lactate level greater than or equal to 4 mmol/L have a similar mortality to
hypotensive patients who are not hyperlactemic.3,25,26,32,33 Lactate screening was a
component of the protocol in all 3 trials. For consistency with the original EGDT trial,
ProMISe, ProCESS, and ARISE used lactate level greater than 4 mmol/L as an inde-
pendent inclusion criterion. Lactate level greater than 4 mmol/L was the sole identifier
in almost half of the recruited patients (Table 2).
When comparing sole hyperlactatemia (lactate level >4 mmol/L in normotensive pa-
tients), sole refractory hypotension (hypotension with normal lactate), and the two
combined, the inclusion criteria met most often in ProCESS and ProMISe was
326 Osborn

Table 1
Inclusion criteria across trials

ProCESS ARISE ProMISe EGDT

Age >18 y U U U U
Infection: Suspected, U U U U
presumed, or known
Serum lactate screening U U U U
requireda
Serum lactate >4 mmol/L U U U U
Refractory hypotension U U U U
Fluid requirement U U U U
Fluid amountb 20 cc/kg changed 1 liter 1 liter 20 – 30 cc/kg
to 1 liter
SBP <90 mm Hgc U U U U
MAP <65 mm Hg 7 U U 7
Antibiotics required prior 7 U U 7
to randomizationd
a
EDGT lactate screening for enrolment.
b
Amount of fluid required to meet study definition of refractor hypotension.
c
Blood pressure parameter required for inclusion (either SBP or MAP in ProMISe/ARISE).
d
Antibiotics encouraged within 6 hours (ProCESS), no universally accepted standard for timing at
time of EGDT – timing per clinical determination.

hyperlactatemia followed by refractory hypotension. The most common criterion in

ARISE was sole refractory hypotension followed by sole hyperlactatemia. In all 3 trials,
the least common criterion was patients with both refractory hypotension and hyper-
lactatemia (see Table 1).

Fig. 1. Severe sepsis and septic shock definitions were the inclusion criteria. All 3 trials used
the established definitions for inclusion criteria. WBC, white blood cell count.
 Severe Sepsis and Septic Shock 327

Table 2
ProCESS/ARISE/ProMISe prevalence of hyperlactatemia versus hypotension in the usual
resuscitation group

ProCESS ARISE ProMISe SEPSIS-3*

Lactate >4 Lactate >4 Lactate >4 Lactate >4
mmol/L mmol/L mmol/L mmol/L
Hyperlactatemia Onlya (%) 46.7 30.2 44.4 9.9
Refractory Hypotension Onlyb (%) 39.3 53.5 36.3 22.0
Both Refractory Hypotension and 14.0 16.3 19.3 23.6
Hyperlactatemiac (%)
Any Lactate >4 mmol/L (%) 60.7 46.5 63.7 9.9
a
Normotension and hyperlactatemia (lactate level >4 mmol/L).
b
Refractory hypotension after a 1-L fluid challenge without hyperlactatemia.
c
Concurrent refractory hypotension and hyperlactatemia.
Data from Mouncey PR, Osborn TM, Power GS, et al. Trial of early, goal-directed resuscitation for
septic shock. N Engl J Med 2015;372(14):1301–11 [supplemental material] and Shankar-Hari M, Phil-
lips GS, Levy ML, et al. Developing a new definition and assessing new clinical criteria for septic
shock: for the third international consensus definitions for sepsis and septic shock (Sepsis-3).
JAMA 2016;315(8):775–87.

Antibiotic
Antibiotic delivery time is associated with survival in both severe sepsis34 and septic
shock.34,35 Antibiotic initiation was required before randomization in ProMISe and
ARISE (Fig. 2). ProCESS encouraged early antibiotic delivery and, during the era of
the original EGDT trial, there was no universally agreed-on time frame of antibiotic de-
livery except that it be as soon as possible as established by the clinician (see Table 1).

Time Requirements
Time requirements were substantial for all the trials. In the original EGDT trial, ProMISe
and ARISE patients had to meet inclusion criteria within 6 hours of presentation to the

9 9

Fig. 2. Inclusion criteria required in ProCESS, ProMISe, and ARISE. Before patients were ran-
domized in all 3 trials, identification criteria had to be met (requiring early identification/
fluid administration) and antibiotics initiated.
328 Osborn

ED. In ProCESS, patients had to meet inclusion criteria within 12 hours of ED presen-
tation. Once inclusion criteria were met, all trials had up to 2 hours to obtain consent
and enroll the patient (Fig. 3).
Thus, before being randomized into the treatment or usual-resuscitation arms, the
patients required early identification, including either refractory hypotension or hypo-
perfusion (lactate level >4 mmol/L), and early intravenous fluid administration and early
antibiotics (ARISE, ProMISe).

TREATMENT OPTIONS

ProMISe and ARISE had 2 arms to which patients were randomized, consisting of
treatment and usual resuscitation. ProCESS randomized to 1 of 3 arms: treatment,
usual resuscitation, and a protocolized standard therapy arm. The treatment arms
consisted of similar processes to those described in the 2001 EGDT articles
(Fig. 4), which entailed continuous monitoring of prespecified physiologic targets,
including CVP, MAP, and central venous oxygen saturation (ScvO2), to guide delivery
of intravenous fluids, vasoactive drugs, and packed red blood cell transfusions.
The ProCESS protocolized standard care arm was a 6-hour protocol that consisted
of ensuring venous access (peripheral or central) and administration of fluid and vaso-
active agents to a shock index goal and hourly clinical assessment of fluid status and
hypoperfusion (Fig. 5).
How the protocols were delivered and monitored was similar between trials. ARISE
and ProMISe delivered the protocols within existing resources and required each hos-
pital to have a plan for delivery. ProCESS required a designated team outside of the
ED work flow to deliver protocolized care in the two treatment arms.
ProCESS patients had higher acute physiology and chronic health evaluation
(APACHE) II scores, whereas ProMISe had the higher lactate values, followed by ARISE.
ProMISe had the greatest percentage of patients with lactate levels greater than 4 mmol/
L, and all 3 trial had baseline ScvO2 greater than or equal to 70% (Table 3). All 3 trials had
greater fluid, vasopressor, red cell, and dobutamine administration in the EGDT group
compared with usual resuscitation. Central venous catheters and arterial catheters
were placed in 50% to 60% and 60% to 75% of usual resuscitation patients respectively.

Fig. 3. Timeline overview. a ProMISe and ARISE required inclusion criteria to be met on or
before 6 hours. b ProCESS required inclusion criteria to be met on or before 12 hours. All
3 trials required randomization within 2 hours of meeting inclusion criteria.
 Severe Sepsis and Septic Shock 329

Fig. 4. Early, protocolized resuscitation in all 3 trials were similar to the original EGDT
protocol. (From Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treat-
ment of severe sepsis and septic shock. N Engl J Med 2001;345(19):1368–77; with permission.)

Vasopressors, red cell transfusions, and dobutamine were administered in approxi-

mately half, less than 7%, and less than 4% respectively (Table 4).

The good physician treats the disease; the great physician treats the patient who
has the disease.
—William Osler

EVALUATION OF OUTCOME

In the evaluation of adults with early signs of septic shock who presented to the ED,
ProCESS, ARISE, and ProMISe reported no significant difference in hospital or
90-day mortality among patients receiving 6 hours of EGDT and those receiving usual
330

Fig. 5. ProCESS Trial included a third group “protocol-based standard therapy” with a set of
6-hour resuscitation instructions, but the components were less aggressive than those used
for protocol-based EGDT. CVP, central venous pressure; HCT, hematocrit; IVF, intravenous
fluids; MAP, mean arterial pressure; PRBC, packed red blood cells; SBP, systolic blood pres-
sure; SI, shock index; ScvO2, central venous oxygen saturation. 1Central line should only be
placed and used for venous access. During the 6 h intervention, CVP and ScvO2 measure-
ments are discouraged. If time–sensitive fluid targets can be achieved with smaller IVs
(e.g., one 18 g and one 20 g), that is acceptable. 2Only isotonic fluid should be used (e.g.,
saline, lactated Ringer’s). Colloids are neither encouraged nor excluded. 3Fluid replete/over-
load is defined here as a clinical diagnosis by the treating ProCESS Investigator. Signs and
symptoms of overload include jugular venous distention, rales, and decreased pulse oxime-
try readings. Discontinue all IVF (boluses, background rate) once this occurs, until no longer
deemed fluid replete/overload. 4If patient’s SBP is within 10% of known baseline SBP, AND
patient is not deemed to be clinically hypoperfused, the SBP >100 mmHg target can be
deemed fulfilled. Arterial lines allowed if deemed necessary, but not mandatory. Shock in-
dex 5 heart rate/systolic blood pressure. 5Hypoperfusion is defined here as a clinical diag-
nosis by the treating ProCESS Investigator. Signs and symptoms include, but are not
limited to, MAP <65 despite SBP >100, arterial lactate >4, mottled skin, oliguria, and altered
sensorium. 6Transfuse PRBCs for Hgb <7.5 g/dL. 7From time of prompt by protocol (i.e., not
from time of physician order, or from when intravenous fluid bag hung). (From Process In-
vestigators, Yealy DM, Kellum JA, et al. A randomized trial of protocol-based care for early
septic shock. N Engl J Med. 2014;370(18):1683–93; with permission.)
 Severe Sepsis and Septic Shock 331

Table 3
Baseline values versus resuscitation end points

ProCESS ARISE ProMISe EGDT

EGDT UR EGDT UR EGDT UR EGDT UR
APACHE IIa 20.8 20.7 15.4 15.8 18 17 21.4 20.4
SOFAa — — — — 4 4 — —
SBP (mm Hg) 100.2 99.9 78.8 79.6 77.7 78.4 106 109
SBP, 6 h (mm Hg) — — — — 113.1 110.7 — —
Lactate (mM/L) 4.8 4.9 6.7 6.6 7.0 6.8 7.7 6.9
Lactate, 6 h (mm Hg) — — 2.8 2.9 — — — —
Lactate >4 mM/L (%) 59 60 46.0 46.5 65.4 63.7 79b
CVP — — — — — — — —
CVP, 6 h (mm Hg) — — 11.4 11.9 11.2 11.7 13.8 11.8
ScvO2% 71 — 72.7 — 70.1 — 48.6 49.2
ScvO2% (6 h) — — 75.9 — 74.2 — 77.3 66.0

Abbreviations: UR, usual resuscitation; HR, heart rate; CVP, central venous pressure; ScvO2%, central
venous oxygen saturation (%).
a
Median (interquartile range [IQR]).
b
Value found in Table 3 from Nguyen Crit Care (2016).51

resuscitation (Table 5). In all 3 evaluations, the overall rate of death was lower than
anticipated. ARISE and ProMISe evaluated ED length of stay (LOS) and reported
significantly higher LOS in the usual-resuscitation groups. However, the maximum
median time of 2 hours from ED presentation is significantly lower than the 6-hour na-
tional average of ED LOS for critically ill patients in the United States.36 None of the 3
trials showed a significant difference in hospital LOS. ProMISe reported an increased
ICU LOS in the EGDT group with ProCESS and ARISE reporting no difference. ProM-
ISe was the only trial of the three to evaluate quality of life and resource consumption.
Quality of life of patients with severe sepsis and septic shock (0.60) was substantially
lower than that of age-matched and gender-matched people in the general popula-
tion (0.80). ProMISe also reported a low probability that EGDT was cost-effective
because of increased treatment intensity in the EGDT groups and no difference in
outcomes.
The benefits of EGDT treatment in the original trial were substantial. Why would 3
trials now show such divergent results? Possibilities include:
 The benefits derived in the original single-center RCT might not have been gener-
alizable, and the multiple observational, before/after, historical control studies
were compromised by bias, confounding, sample size, or chance.
 Recruitment time could have encompassed practice changes within the institu-
tions that could have modified the effect benefit.
 The intervention (EGDT) was not blinded in ProCESS, ARISE, or ProMISe.
However, the risk of bias was minimized through central randomization
and the use of a primary outcome not subject to observer bias. In the orig-
inal EGDT RCT, the ED, where the treatment was administered, was not blinded.
However, the ICU practitioners were blinded to who received EGDT in the ED.
 Change of care over the 15 years between trials may have affected the results. The
mortality in the 3 recent trials was substantially reduced compared with those re-
ported by Rivers and colleagues4 (see Table 5). In addition, severity of illness
 332
Osborn
Table 4
Interventions delivered during a 6-hour period

ProCESS ARISE ProMISe EGDT

EGDT UR EGDT UR EGDT UR EGDT UR
Arterial Catheter Insertion (%)a Required — 91.4 76.3 74.2 62.2 Required Required
CVC Insertion (%)a,b 93.6 57.9 90 61.9 92.1 50.9 Required Required
Any IV Fluid (mL)a 2805 (1957) 2279 (1881) 1964 (1415) 1713 (1404) 2226 (1443) 2022 (1271) 4981 (2984) 3499 (2438)
Vasopressor (%)a 54.9 44.1 66.6 57.8 53.3 46.6 27.4 30.3
Red Cell Transfusion (%)a 14.4 7.5 13.6 7.0 8.8 3.8 64.1 18.5
Dobutamine (%)a 8 0.9 15.4 2.6 18.1 3.8 13.7 .8

Abbreviations: CVC, central venous catheter; IV, intravenous.

a
Significant difference (P<.001).
b
CVC: ProMISe and ProCESS, any CVC placement; ARISE CVC, EGDT is ScvO2, UR is non-ScvO2.
 Table 5
Outcomes

ProCESS ARISE ProMISe EGDT

EGDT UR EGDT UR EGDT UR EGDT UR
Hospital Death (%) 21 18.9 14.5 15.7 25.6 24.6 30.5 46.5a
Death by 60 d (%) — — — — — — 33.3 49.2a
Death by 90 d (%) 31.9 33.7 18.6 18.8 29.5 29.2 — —
Median ED LOS (IQR)a — — 1.4 (0.5–2.7) 2.0 (1.0–3.8) 1.5 (0.4–3.1) 1.3 (0.4–2.9) — —
Median ICU LOS (IQR) 5.1 (6.3) 4.7 (5.8) 2.8 (1.4–5.1) 2.8 (1.5–5.7) 2.6 (1.0–5.8) 2.2a (0–5.3) — —
HLOSb 11.1 (10) 11.3 (10.9) 8.2 (4.9–16.7) 8.5 (4.9–16.5) 9 (4–21) 9 (4–18) — —

ProCESS used combined in hospital death or death by 60 days.

Severe Sepsis and Septic Shock

 denotes mean with standard deviation.
Abbreviations: HLOS, hospital length of stay; LOS, length of stay.
a
Significant difference (P<.01).
b
Unless otherwise noted, median (IQR).

333
334 Osborn

seemed greater in the Rivers and colleagues4 trial. The ProMISe and ARISE trial
patients had lower APACHE II scores, and all 3 trials had lower lactate levels
and higher Scvo2 (see Table 3). Patients from all 3 trials received antibiotics
very quickly. Receiving antibiotics before randomization was an inclusion criterion
for ProMISe and ARISE, and ProCESS encouraged rapid administration (Fig. 6,
Table 6). Other care elements, such as lung-protective ventilation, that would
now most likely be part of general practice would not have been in 1997 to 2000.
For the general population of patients with severe sepsis and septic shock, rigid
protocols requiring universal application of central venous catheters, red cell transfu-
sion, and inotropes was not effective. Given that the 3 trials had baseline ScvO2 greater
than 70%, it is unclear whether there may be a subset of this patient population that
could benefit from normalizing an abnormally low ScvO2. Important to ensuring individ-
ual intuitional replication of these results is a clear understanding of what usual resus-
citation entailed in ProCESS, ARISE, and ProMISe.

What Was Usual Care?

Often the question arises, “How did all that treatment equate to no care?” The answer
is that usual care, in this case usual resuscitation, did not equate to no care. Usual care
equated to a standardized protocol requiring early identification (including lactate
screening), early intravenous fluids, and early antibiotics (see Fig. 6, Table 6). The hos-
pital mortality in the usual-care groups of ProCESS, ARISE, and ProMISe were 18.9%,
15.7%, and 24.6% respectively (see Table 5). Before translating these results to indi-
vidual institutions, specific parameters for quality improvement projects should
consider the current hospital mortality and the following specific parameters. It is un-
clear that individual institutions would have the same results without similar quality
maintenance processes in place (see Table 6).
 Early identification
 Definition: the foundation of how septic patients are identified centers on how
the disorder is defined. As with most major sepsis evaluations to date, the defi-
nition used in all 3 of these large, prospective RCTs and the original EGDT RCT

Fig. 6. Timeline overview and results for usual care. a Average of median values from all
3 trials. b Antibiotic administration was required before randomization. Longest time to
randomization of usual care was 3 hours (1.6 hours). c Average time to enrollment of
all 3 trials was 2.7 hours.
 Severe Sepsis and Septic Shock 335

Table 6
Early treatment in usual resuscitation patients

PROCESS ARISE ProMISe

Hours to Identification 1.5 (+/ 0.75) 1.3a (0.5–2.4) 1.7 (+/ 1.4)
Identification by lactate aloneb 45% 30% 45%
Hours to antibiotica administration 76.1% less than 3 hoursc 1.12 (.63–1.8) 1.3 a (0.6–2.4)
Fluids prior to randomization (liters) 2.1 (+/ 1.4) 2.6 (+/ 1.3) 2.0 (+/ 1.1)
Hours to randomizationa 3.0 (+/ 1.6) 2.7 (2.0–3.9) 2.5 (1.8–3.5)
Fluids 0–6 h (liters)a 2.8 1.7 2.0
Central venous catheter 58% 62% 51%
Vasopressor use 48% 58% 47%
ED LOSa (h) NR 2.0 (1.0–3.8) 1.3 (.4–2.9)

(+/ ) denote means with standard deviation; a point at which all inclusion criteria met in ProMISe
include antibiotic initiation.
Abbreviation: NR, Not Reported.
a
Median (IQR) unless otherwise noted.
b
Percent of normotensive, usual resuscitation patients identified by lactate >4 mmol/l alone.
c
Related to randomization time.

was based on the established definition of severe sepsis and septic shock24
and lactate screening.
 Structured screening protocol: all 3 trials had structured screening of ED pa-
tients on arrival and during the ED stay with a standardized tool. Screening
tools allowed consistency among providers.
 Lactate screening: using lactate as a component of sepsis screening was
essential. Sole hyperlactatemia (lactate level >4 mmol/L) identified up to half
of the eligible patients. In addition, point-of-care lactate measurement was
found to be effective in further reducing time to identification.
 Time to identification: the median or mean time to identification from ED pre-
sentation was less than 2 hours in all 3 trials (see Fig. 6, Table 6).
 Treatments provided to patients before randomization:
 Early intravenous antibiotics: antibiotics administration before randomization
was required in ARISE and ProMISe and was encouraged in ProCESS. The
median time to antibiotic administration was 1 to 2 hours from ED presentation
in ARISE and ProMISe and less than 3 hours in all 3 trials.
 Early fluid administration: before randomization, usual-care patients received 2
to 3 L of fluid between the 3 trials.
 Treatments provided to usual resuscitation patients after randomization:
 In all 3 trials, usual resuscitation patients received monitoring, investigations,
and treatment, considered appropriate by the treating clinicians.
 ED LOS: median ED LOS was 1 hour (ProMISe) to 2 hours (ARISE) before
admission (see Table 5).

PATIENT IDENTIFICATION: WHAT DEFINITION SHOULD BE USED?

Representatives from 2 critical care organizations (Society of Critical Care Medicine

and the European Society of Intensive Care Medicine) recently formed a task force
to derive new sepsis definitions for screening and identification.32,37,38 After
completing their work, the definitions were sent to other organizations for review
and endorsement. The organizations should be commended for their efforts to devise
336 Osborn

definitions that incorporate large data evaluation in conjunction with expert opinion.
This set of sepsis definitions was the first to include rigorous evaluations resulting in
derivation and validation of new sepsis definitions and supporting documents.32
However, the proposed definitions have not been universally accepted.39–41 The
guidelines were not endorsed by the American College of Chest Physicians, the Infec-
tious Disease Society of America, the American College of Emergency Physicians, or
the Latin American Sepsis Institute, among others. The concerns can be categorized
into 2 areas: clinical and operational.

CLINICAL
Screening
Its authors propose Quick Sequential Organ Failure Assessment (qSOFA) as a tool to
screen infected patients for severe sepsis or septic shock. Infected patients with 2 or
more criteria of hypotension, altered mental status, and increased respiratory rate
were found to be at risk for the primary outcome of death and the secondary outcome
of death with or without ICU stay of 3 days or more (Table 7).
The traditional paradigm has been to treat the sick but cast a wide net, screening for
patients who may deteriorate and treat to arrest progression.1–3,42–47 Many clinicians
consider infected patients with hypotension and/or altered mental status to no longer
be at risk of decline but within an active deterioration phase of organ system dysfunc-
tion and/or failure.42,43,48,49 In the ICU, the higher level of specificity compared with
sensitivity is understood. If a patient at risk for deterioration is missed, the person is
still within a monitored hospital setting and ICU resources can be managed. However,
misidentifying a patient at risk of decline on the floor or in the ED can result in a
different outcome. The investigators show eloquently that qSOFA is associated with
high risk for poor outcomes.38 However, as its authors state, the applicability as a
screening tool remains a question for prospective evaluation and further

Table 7
Comparisons of established, SEPSlS-3 definitions and SSC Guidelines application

Established Definition
(Used by CMS) Sepsis - 3 Definition
Sepsis Presumed/known 2 SOFA criteria (present
infection + 2 SIRS or increased)
Includes: hypotension +
normal lactate (shock)
Severe Sepsis Sepsis + End organ Not a category
dysfunction, Lactate >4
mmol/L
Septic shock Sepsis + Refractory Vasopressors AND
hypotension (+/ lactate >2 mmol/L
lactate)
Mortality Observed Mortality Sepsis 5 low acuity Sepsis 5 higher acuity
5 Expected mortality
Ratio Observed mortality low Observed mortality higher
Example Expected mortality low Expected mortality low

Mortality ratio, national quality metrics based on established definitions (expected mortality).
When clinicians apply a low acuity diagnosis (sepsis) to a higher acuity patient (SEPSlS-3 definition
of sepsis), the observed mortality will be higher than expected. Results in similar care appearing
worse based on different definitions applied to the same patients.
 Severe Sepsis and Septic Shock 337

discrimination of patients who currently are identified and treated who are not consid-
ered within this paradigm.

Application: Screening and Identification of Sepsis and Septic Shock

The SEPSIS-3 definitions categorize sepsis as infected patients with an increase of
greater than or equal to 2 Sequential Organ Failure Assessment (SOFA) points
from baseline in the ICU (area under the curve, 0.79). Outside of the ICU, it is recom-
mended that clinicians obtain information for SOFA calculations if 2 qSOFA criteria
are attained.37 Both systems were validated in secondary databases (Table 8).
The new definitions categorize septic shock as hypotension requiring vaso-
pressors and increased lactate level (>2 mmol/L). This combination was highly
predictive of poor outcomes with an estimated mortality of 42%32,37 (Table 9).
There is concern that use of the SEPSIS-3 definitions may foster under-
recognition of partially compensated septic shock phenotypes. Cryptic shock, as
shown by increased lactate level alone, is a phenotype that portends a worse
outcome rather than simply a bad infection. Some clinicians are concerned about
patients who may be missed, especially in the early phases of care when durable
long-term improvements are most possible.
It is unclear why lactate, an established indicator of critical illness,25–31 was not
retained in the qSOFA screening criteria. It is possible that hyperlactatemia is not as
good an indicator of decline in septic patients as was previously considered. As its au-
thors point out, it is also possible that the infrequent lactate evaluations may have
contributed. Only 9% of infected patients from the derivation set and validation set
had lactate evaluation on the day of admission. A post-hoc evaluation of qSOFA
and lactate levels was done in the Kaiser Permanente Northern California health sys-
tem data set, in which 57% of patients with suspected infection had lactate levels
measured on the day of admission; however, only 20% had lactate level increases
greater than or equal to 2 mmol/L.38
In ProCESS, ARISE, and ProMISe, in which lactate measurement was one of the in-
clusion criteria, lactate level greater than 4 mmol/L was the sole identifier in 30% to
45% of patients (see Table 6). Further, when evaluating patients across 240 ICUs
across the United Kingdom, the prevalence of lactate level greater than 4 mmol/L in
normotensive infected patients presenting to the ICU was 44.5%. This finding is
consistent with ProCESS, ARISE, and ProMISe.
The mortality of normotensive hyperlactemic patients in the ARISE data set was
26% and in the Sepsis-3 article the prevalence for lactate level greater than
4 mmol/L was 9.9%; however, the mortality was 29.9%. The mortality for hyperlacta-
temia (29.9%) is equivalent to the mortality of hypotensive patients (29.7%); however,
hypotension is part of the definition but lactate level is not a component until the pa-
tient has concurrent hypotension (see Table 9).
For example, consider a 55-year-old woman, absent preexisting organ dysfunc-
tion (baseline SOFA 0), who is infected from a urinary source. Initially she is normo-
tensive and her lactate level is 5 mmol/L. She would not fit the proposed definition
paradigm offered by Sepsis-3. In addition, she might not have lactate level evalu-
ated because it is only considered in the proposed Sepsis-3 definitions if the patient
is hypotensive.
If this same woman presents with hypotension (MAP of 60 mm Hg) after receiving
fluids (SOFA 1), she may not be considered septic until vasopressors are initiated
and not considered in septic shock until her lactate result is known, if it is increased.
This patient’s classification depends on clinical decisions or processes rather than
clinical presentation and progression.
 338
Osborn
Table 8
Sequential [Sepsis-Related] Organ Failure Assessment Score

Score
System 0 1 2 3 4
Respiration
PaO2/FiO2, mm Hg (kPa) 400 (53.3) <400 (53.3) <300 (40) <200 (26.7) with <100 (13.3) with
respiratory support respiratory support
Coagulation
Platelets, 103/mL 150 <150 <100 <50 <20
Liver
Bilirubin, mg/dL (mmol/L) <1.2 (20) 1.2–1.9 (20–32) 2.0–5.9 (33–101) 6.0–11.9 (102–204) >12.0 (204)
Cardiovascular MAP 70 mm Hg MAP <70 mm Hg Dopamine <5 or Dopamine 5.1–15 or Dopamine >15 or
dobutamine epinephrine 0.1 or epinephrine >0.1 or
(any dose)a norepinephrine 0.1a norepinephrine >0.1a
Central nervous system
Glasgow Coma Scale 15 13–14 10–12 6–9 <6
scoreb
Renal
Creatinine, mg/dL <1.2 (110) 1.2–1.9 (110-170) 2.0–3.4 (171–299) 3.5–4.9 (300–440) >5.0 (440)
(mmol/L)
Urine output, mL/d — — — <500 <200

Abbreviations: FiO2, fraction of inspired oxygen; MAP, mean arterial pressure; PaO2, partial pressure of oxygen.
a
Catecholamine doses are given as mg/kg/min for at least 1 hour.
b
Glasgow Coma Scale scores range from 3–15; higher score indicates better neurological function.
Adapted from Singer M, Deutschman CS, Seymour CW, et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA
2016;315(8):801–10.
 Table 9
Hyperlactatemia and refractory hypotension: prevalence and mortality

ICNARC
Sepsis-3 Definition Sepsis-3 Definition Sepsis-3 Definition Lactate >4 mmol/L
N 5 18,840 Refractory Hypotension Only Lactate >2–4 mmol/L Lactate >4 mmol/L N 5 12,004
Hyperlactatemia Only — 4106 out of 18,840a (21) 1856 out of 18,840 (9.9) 5339 out of 12,004 (44.5)
N (% total)
Hyperlactatemia Only — 1086 out of 4108a (26.4) 555 out of 1856 (29.9) 1397 out of 5339 (26.2)
Mortality N (%)
Refractory Hypotension Onlyb 4135 out of 18,840 (22.0) — — 2186 out of 12,004 (18.2)
N (%)
Refractory Hypotension Onlyb 1226 out of 4135 (29.7) — — 1397 out of 5339 (26.2)
Mortality N (%)
Both Refractory Hypotension — 4295 out of 18,840 (22.8) 4448 out of 18,840 (23.6) 4479 out of 12,004 (37.3)
and Hyperlactatemiac N (%)
Both Refractory Hypotension — 1498 out of 4295 (34.2) 2198 out of 4448 (49.4) 2485 out of 4479 (55.5)

Severe Sepsis and Septic Shock

and Hyperlactatemiac
Mortality N (%)

Authors; ARISE data adapted from Table S22 (Mouncey et al,3 2015).
a
Combined groups 4 and 5.
b
Combined groups 2 and 6.
c
Combined groups 1 and 3.
Adapted from Table 3 in Shankar-Hari M, Phillips GS, Levy ML, et al. Developing a new definition and assessing new clinical criteria for septic shock: for the Third
International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315(8):775–87.

339
340 Osborn

If lactate assessment is only considered proposed in the Sepsis-3 definitions when

combined with hypotension, will some clinicians, especially in resource-limited areas,
stop looking for it before hypotension? It was reported32,37 that hyperlactatemia in
normotensive patients occurred more often than refractory hypotension alone. How-
ever, given that mortality for both was high (see Table 9), which care opportunities
will be missed and how would they be affected?

Operationalization and the Law of Unintended Consequence

Clinicians using the new Sepsis-3 definition will seem to provide worse care for pa-
tients with sepsis than clinicians using the established definition. This difference
may occur in any country or environment where a mortality index is used to assist
with monitoring value-based care or care quality. This difference may exist despite
the quality of care provided.
One method of national quality and value monitoring is the mortality index. The mor-
tality index (MI) is the ratio of observed deaths to expected deaths.

Observed deaths
MI 5
Expected deaths

There has been no decision to date that third parties will change from using
the established definitions. Centers for Medicare & Medicaid Services stated
at the time of the SEP-3 definition articles release32,37 that they had no current
plans to change the definitions they were currently using. Because the established
definition of sepsis has a low mortality, the expected deaths are low. In the
proposed definition, patients who would normally be classified as having septic
shock (hypotension plus normal or no lactate measurement) would now be
classified as having sepsis. Although the observed deaths would be higher, the
expected deaths (based on the established, older sepsis definition) would
remain low.
For example:

Observed deaths
MI 5
Expected deaths

Infection hypotension1no=normal lactate Observed Mortality

MI 5 5
Above patient “Septic shock” diagnosis Expected Death 30%1

Infection hypotension1no=normal lactate Observed Mortality

MI 5 5
Above patient “Sepsis” diagnosis Expected Death 15%1

Percentages presented with the plus symbol (1) are estimates for example pur-
poses only. Septic shock mortality is based on SSC database. Expected mortality
for sepsis is based on 2016 global report.50

REVISED SEPSIS BUNDLE

With the publication of 3 sepsis RCTs that do not show superiority of EGDT in patients
with sepsis who have received timely antibiotics and fluid resuscitation compared
with controls, the Surviving Sepsis Campaign guidelines revised the sepsis bundles
to the following (http://www.survivingsepsis.org/SiteCollectionDocuments/SSC_
Bundle.pdf):
 Severe Sepsis and Septic Shock 341

To be completed within 3 hours of time of presentationa:

1. Measure lactate level.
2. Obtain blood cultures before administration of antibiotics.
3. Administer broad-spectrum antibiotics.
4. Administer 30 mL/kg crystalloid for hypotension or lactate level greater than or equal to
4 mmol/L.
To be completed within 6 hours of time of presentation:
5. Apply vasopressors (for hypotension that does not respond to initial fluid resuscitation) to
maintain an MAP greater than or equal to 65 mm Hg.
6. In the event of persistent hypotension after initial fluid administration (MAP <65 mm Hg) or
if initial lactate level was greater than or equal to 4 mmol/L, reassess volume status and
tissue perfusion and document findings according to Table 1.
7. Remeasure lactate if initial lactate level is increased.
Document reassessment of volume status and tissue perfusion with:
Either:
 Repeat focused examination (after initial fluid resuscitation) by licensed independent
practitioner, including vital signs and cardiopulmonary, capillary refill, pulse, and skin findings.
Or 2 of the following:
 Measure CVP
 Measure ScvO2
 Bedside cardiovascular ultrasonography
 Dynamic assessment of fluid responsiveness with passive leg raise or fluid challenge
a
Time of presentation is defined as the time of triage in the ED or, if presenting from another
care venue, from the earliest chart annotation consistent with all elements of severe sepsis or
septic shock ascertained through chart review.

LONG-TERM RECOMMENDATIONS
1. Identification: it is suggested that it may be prudent to wait for prospective valida-
tion of the proposed Sepsis-3 definitions before universally implementing them
clinically and use the established definitions currently. If clinicians or hospitals think
that using the proposed definitions will facilitate patient care, they should consider
using them.
2. Usual care in ProCESS, ProMISe, and ARISE for patients with severe sepsis and
septic shock entailed early identification with standardized screening protocols,
including lactate measurement, early fluid administration for hypotensive patients,
and early antibiotic administration within 3 hours.
3. Data measurement: there should be some form of data collection and monitoring
for on-going quality-assessment purposes.

SUMMARY

Usual resuscitation may have evolved over the last 15 years since the original EGDT
trial. In-hospital and 90-day survival of patients receiving usual resuscitation were
similar to those achieved with EGDT when they were identified early through a
screening program including lactate measurement and received early intravenous
fluids and antibiotics. The addition of continuous ScvO2 monitoring and strict protoco-
lization did not improve outcomes. It is unclear whether there are subpopulations that
might benefit from additional treatment.
342 Osborn

REFERENCES

1. Process Investigators, Yealy DM, Kellum JA, et al. A randomized trial of protocol-
based care for early septic shock. N Engl J Med 2014;370(18):1683–93.
2. ARISE Investigators, Anzics Clinical Trials Group, Peake SL, et al. Goal-directed
resuscitation for patients with early septic shock. N Engl J Med 2014;371(16):
1496–506.
3. Mouncey PR, Osborn TM, Power GS, et al. Trial of early, goal-directed resuscita-
tion for septic shock. N Engl J Med 2015;372(14):1301–11.
4. 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(19):1368–77.
5. Salvo I, de Cian W, Musicco M, et al. The Italian SEPSIS study: preliminary results
on the incidence and evolution of SIRS, sepsis, severe sepsis and septic shock.
Intensive Care Med 1995;21(Suppl 2):S244–9.
6. Brun-Buisson C, Doyon F, Carlet J, et al. Incidence, risk factors, and outcome of
severe sepsis and septic shock in adults. A multicenter prospective study in inten-
sive care units. French ICU Group for Severe Sepsis. JAMA 1995;274(12):968–74.
7. Zeni F, Freeman B, Natanson C. Anti-inflammatory therapies to treat sepsis and
septic shock: a reassessment. Crit Care Med 1997;25(7):1095–100.
8. Sands KE, Bates DW, Lanken PN, et al. Epidemiology of sepsis syndrome in 8
academic medical centers. JAMA 1997;278(3):234–40.
9. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in
the United States: analysis of incidence, outcome, and associated costs of care.
Crit Care Med 2001;29(7):1303–10.
10. Nasraway SA. Norepinephrine: no more “leave ’em dead”? Crit Care Med 2000;
28(8):3096–8.
11. Practice parameters for hemodynamic support of sepsis in adult patients in
sepsis. Task force of the American College of critical care medicine, Society of
critical care medicine. Crit Care Med 1999;27(3):639–60.
12. Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines for
management of severe sepsis and septic shock. Intensive Care Med 2004;30(4):
536–55.
13. Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international
guidelines for management of severe sepsis and septic shock: 2008. Crit Care
Med 2008;36(1):296–327.
14. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: international
guidelines for management of severe sepsis and septic shock, 2012. Intensive
Care Med 2013;39(2):165–228.
15. Otero RM, Nguyen HB, Huang DT, et al. Early goal-directed therapy in severe
sepsis and septic shock revisited: concepts, controversies, and contemporary
findings. Chest 2006;130(5):1579–95.
16. Safar P. Critical care medicine–quo vadis? Crit Care Med 1974;2(1):1–5.
17. Huang DT, Osborn TM, Gunnerson KJ, et al. Critical care medicine training and
certification for emergency physicians. Ann Emerg Med 2005;46(3):217–23.
18. Wang HE, Shapiro NI, Angus DC, et al. National estimates of severe sepsis in
United States emergency departments. Crit Care Med 2007;35(8):1928–36.
19. Osborn TM, Nguyen HB, Rivers EP. Emergency medicine and the Surviving
Sepsis Campaign: an international approach to managing severe sepsis and
septic shock. Ann Emerg Med 2005;46(3):228–31.
 Severe Sepsis and Septic Shock 343

20. McIntyre LA, Hebert PC, Fergusson D, et al, Canadian Critical Care Trials Group.
A survey of Canadian intensivists’ resuscitation practices in early septic shock.
Crit Care 2007;11(4):R74.
21. Sivayoham N. Management of severe sepsis and septic shock in the emergency
department: a survey of current practice in emergency departments in England.
Emerg Med J 2007;24(6):422.
22. ProCESS/ARISE/ProMISe Methodology Writing Committee, Huang DT, Angus DC,
Barnato A, et al. Harmonizing international trials of early goal-directed resuscita-
tion for severe sepsis and septic shock: methodology of ProCESS, ARISE, and
ProMISe. Intensive Care Med 2013;39(10):1760–75.
23. Gross SD. A system of surgery. Philadelphia: 1872.
24. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and
guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM
consensus conference committee. American College of Chest Physicians/Society
of Critical Care Medicine. Chest 1992;101(6):1644–55.
25. Ranzani OT, Monteiro MB, Ferreira EM, et al. Reclassifying the spectrum of septic
patients using lactate: severe sepsis, cryptic shock, vasoplegic shock and dys-
oxic shock. Rev Bras Ter Intensiva 2013;25(4):270–8.
26. Puskarich MA, Trzeciak S, Shapiro NI, et al. Outcomes of patients undergoing
early sepsis resuscitation for cryptic shock compared with overt shock. Resusci-
tation 2011;82(10):1289–93.
27. Singer AJ, Taylor M, Domingo A, et al. Diagnostic characteristics of a clinical
screening tool in combination with measuring bedside lactate level in emergency
department patients with suspected sepsis. Acad Emerg Med 2014;21(8):853–7.
28. Puskarich MA, Trzeciak S, Shapiro NI, et al. Whole blood lactate kinetics in pa-
tients undergoing quantitative resuscitation for severe sepsis and septic shock.
Chest 2013;143(6):1548–53.
29. Arnold RC, Shapiro NI, Jones AE, et al. Multicenter study of early lactate clear-
ance as a determinant of survival in patients with presumed sepsis. Shock
2009;32(1):35–9.
30. Shapiro NI, Fisher C, Donnino M, et al. The feasibility and accuracy of point-of-
care lactate measurement in emergency department patients with suspected
infection. J Emerg Med 2010;39(1):89–94.
31. Shapiro NI, Howell MD, Talmor D, et al. Serum lactate as a predictor of mortality in
emergency department patients with infection. Ann Emerg Med 2005;45(5):
524–8.
32. Shankar-Hari M, Phillips GS, Levy ML, et al. Developing a new definition and as-
sessing new clinical criteria for septic shock: for the third international consensus
definitions for sepsis and septic shock (Sepsis-3). JAMA 2016;315(8):775–87.
33. Levy MM, Dellinger RP, Townsend SR, et al. The Surviving Sepsis Campaign: re-
sults of an international guideline-based performance improvement program tar-
geting severe sepsis. Intensive Care Med 2010;36(2):222–31.
34. Ferrer R, Martin-Loeches I, Phillips G, et al. Empiric antibiotic treatment reduces
mortality in severe sepsis and septic shock from the first hour: results from a
guideline-based performance improvement program. Crit Care Med 2014;
42(8):1749–55.
35. 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 sep-
tic shock. Crit Care Med 2006;34(6):1589–96.
36. McCaig LF, Nawar EW. National hospital ambulatory medical care survey: 2004
emergency department summary. Adv Data 2006;(372):1–29.
344 Osborn

37. Singer M, Deutschman CS, Seymour CW, et al. The third international consensus
definitions for sepsis and septic shock (Sepsis-3). JAMA 2016;315(8):801–10.
38. Seymour CW, Liu VX, Iwashyna TJ, et al. Assessment of clinical criteria for sepsis:
for the third international consensus definitions for sepsis and septic shock
(Sepsis-3). JAMA 2016;315(8):762–74.
39. Vincent JL, Martin GS, Levy MM. qSOFA does not replace SIRS in the definition of
sepsis. Crit Care 2016;20(1):210.
40. Machado FR, Salomao R, Pontes de Acevedo LC, et al. Latin American
Sepsis Institute (LASI). Why LASI did not endorse the new definitions of sepsis
published today in JAMA 2016. Available at: http://ilas.org.br/assets/arquivos/
upload/declaracao%20sepse%203.0%20ILAS%20-%20English%20version.pdf.
Accessed October 10, 2016.
41. Simpson SQ. New sepsis criteria: a change we should not make. Chest 2016;
149(5):1117–8.
42. Levy MM, Rhodes A, Phillips GS, et al. Surviving Sepsis Campaign: association
between performance metrics and outcomes in a 7.5-year study. Crit Care
Med 2015;43(1):3–12.
43. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: international
guidelines for management of severe sepsis and septic shock: 2012. Crit Care
Med 2013;41(2):580–637.
44. Levy MM, Artigas A, Phillips GS, et al. Outcomes of the Surviving Sepsis
Campaign in intensive care units in the USA and Europe: a prospective cohort
study. Lancet Infect Dis 2012;12(12):919–24.
45. Levy MM, Dellinger RP, Townsend SR, et al. The Surviving Sepsis Campaign: re-
sults of an international guideline-based performance improvement program tar-
geting severe sepsis. Crit Care Med 2010;38(2):367–74.
46. Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international
guidelines for management of severe sepsis and septic shock: 2008. Intensive
Care Med 2008;34(1):17–60.
47. 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(3):
858–73.
48. Osborn TM, Phillips G, Lemeshow S, et al. Sepsis severity score: an internation-
ally derived scoring system from the Surviving Sepsis Campaign database*. Crit
Care Med 2014;42(9):1969–76.
49. Marshall JC, Cook DJ, Christou NV, et al. Multiple organ dysfunction score: a reli-
able descriptor of a complex clinical outcome. Crit Care Med 1995;23(10):
1638–52.
50. Fleischmann C, Scherag A, Adhikari NK, et al. Assessment of global incidence
and mortality of hospital-treated sepsis. Current estimates and limitations. Am J
Respir Crit Care Med 2016;193(3):259–72.
51. Nguyen HB, Jaehne AK, Jayaprakash N, et al. Early goal-directed therapy in se-
vere sepsis and septic shock: insights and comparisons to ProCESS, ProMISe,
and ARISE. Crit Care 2016;20(1):160.