INTENSIVE CARE

Sepsis 2021: a review Learning objectives

Catherine Wentowski After reading this article, you should be able to:
C define sepsis and septic shock (Sepsis-3)
David Perez Ingles
C describe clinical criteria required for the early diagnosis of sepsis
Nathan D Nielsen
C formulate an initial treatment plan for the septic patient in a

critical care setting

C describe secondary and adjunctive therapies presently available
Abstract
for refractory sepsis
Sepsis is responsible for tremendous morbidity, mortality and health-
care expenditure worldwide. Over the past decade, the conceptualiza-
tion of sepsis has shifted from one based upon an inflammatory
response to one defined by a dysregulated immune response to infec- Pathophysiology of sepsis
tion and resulting organ dysfunction. The definitions of sepsis and At its onset, sepsis manifests as an overwhelming release of in-
septic shock were revised to improve their diagnostic specificity and flammatory mediators (sometimes referred to as ‘cytokine
facilitate accurate and timely diagnoses at the bedside. The core of storm’) in response to an infection. The immune response to
sepsis management remains early identification and diagnostic infection can be correlated with national defenses: components
testing, early antimicrobial therapy, and early haemodynamic resusci- of the innate immune system or ‘citizens’ (epithelial cells, mac-
tation. Current guidelines recommend that the first steps in treatment rophages, mast cells, innate lymphocytes) at the site of pathogen
and resuscitation should take place within 1 hour from when sepsis is exposure activate and recruit circulating immune cells or
suspected. Additional essential elements in the current sepsis man- ‘troopers’ (neutrophils, NK cells, dendritic cells, platelets,
agement guidelines include using dynamic parameters to assess monocytes, eosinophils). These cells have pathogen-recognition
fluid responsiveness, a conservative fluid strategy following initial receptors (PRRs) on their surface which bind to, and are acti-
resuscitation (with ‘de-resuscitation’ when possible), serial re- vated by, pathogen-associated molecular patterns (PAMPs) on
assessments of haemodynamic status and adaptable treatment bacterial cell walls or damage-associated molecular patterns
plans. This article provides a summary of the most recent clinical ev- (DAMPs) e host biomolecules released when danger is sensed
idence and professional guidelines for the diagnosis and treatment of from a pathogen, burn, trauma, etc. This receptor binding initi-
sepsis in the critical care setting. ates an intracellular signaling cascade resulting in the activation
Keywords Critical care; intensive care; sepsis; septic shock of cytosolic transcription factors such as NF-kB and activator
protein 1 (AP-1), which in turn leads to the production of several
Royal College of Anaesthetists CPD Matrix: 2C03 acute phase reactants, among them cytokines, coagulation fac-
tors and inducible nitric oxide synthetase, thus initiating the
immuno-inflammatory cascade. A subsequent chain reaction
involves the activation of even stronger ‘armed forces’, the
Introduction adaptive immune response. This explosive activation and resul-
In 2017 alone, the global incidence of sepsis was estimated to be tant immune ‘cytokine storm’ is believed to be the causative
48.9 million cases and was associated with 11 million deaths, pathway for septic shock.3
thus representing 19.7% of all deaths worldwide.1 More region- The end-response to an infection is often a combination of pro-
ally, data from the NHS showed 244,158 cases of sepsis with and anti-inflammatory cascades. Once infection resolves, a bal-
48,000 deaths in the UK during the same time period. The annual ance is established between immune up-regulating and down-
economic burden (both direct and indirect) is estimated at a regulating processes, and immune memory is generated to pro-
staggering £7.42 billion.2 More timely and accurate recognition of tect against future exposures. However, when the initial response
at-risk patients is essential to improving patient care and health is excessive or dysregulated, this balancing process itself can
system outcomes. To that end, this article will review the defi- become dysfunctional. As sepsis persists, by about 24e48 hours a
nitions and clinical guidelines for the treatment of sepsis. shift towards a hypo-inflammatory state is observed, and patients
develop features consistent with immunosuppression. This phase
is known as sepsis induced immunoparalysis.4 The majority of
septic shock-related deaths occur during this immunoparalysis
Catherine Wentowski MD MPH Division of Pulmonary and Critical phase. These patients are often not able to clear the initial infec-
Care Medicine, Ochsner Clinic Foundation, New Orleans, LA, USA. tion and, in addition, are predisposed to new infections from
Conflicts of interest: none declared. nosocomial pathogens. T-cell and B-cell exhaustion, apoptosis
David Perez Ingles MD Section of Pulmonary Diseases, Critical Care and anergy are now recognized immunosuppressive mechanisms
and Sleep Medicine, Department of Medicine, University of New observed in patients with fatal sepsis.5,6
Mexico, Albuquerque, NM, USA. Conflicts of interest: none declared.
Nathan D Nielsen MD MSc FCCM Section of Pulmonary Diseases, Definitions
Critical Care and Sleep Medicine, Department of Medicine, University
of New Mexico, Albuquerque, NM, USA. Conflicts of interest: none Over the course of the last three decades, considerable effort has
declared. been expended in improving the recognition, categorization and

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 676 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

algorithmic treatment of sepsis. From the early 1990s until the host response to infection’. Septic shock was redefined as a
mid-2010s, sepsis was defined as a proven or suspected infection ‘subset of sepsis in which profound circulatory failure, cellular,
accompanied by a pronounced systemic inflammatory response and metabolic abnormalities are associated with a greater risk of
syndrome (SIRS) (Table 1). Patients were then further catego- mortality than with sepsis alone’ (Table 2).
rized as having sepsis, severe sepsis or septic shock based on In addition to the new definitions, the consensus committee
organ dysfunction and fluid responsiveness. Though widely also established a set of practical clinical guides to facilitate the
accepted at the time, it failed to detect many cases of sepsis (as diagnosis of sepsis at the bedside. The authors selected the
many as 1 in 8) and a more precise definition was required.7 Sequential Organ Failure Analysis (SOFA) assessment to identify
The Third International Consensus Definitions for Sepsis and organ dysfunction (Table 3). In the absence of organ dysfunction
Septic Shock (‘Sepsis-3’), jointly developed by the Society of or chronic comorbid conditions, a SOFA score is zero. An
Critical Care Medicine (SCCM) and the European Society of increasing SOFA score reflects overall deterioration, with an in-
Intensive Care Medicine (ESICM), were released in 2016 and crease of 2 or more associated with increased mortality (see
noted that sepsis is far more complex than an inflammatory Table 4).
response, involving both pro- and anti-inflammatory processes The SOFA score requires the knowledge of multiple laboratory
and non-immunologic systems in the disease process.4 The term values (e.g. platelets, bilirubin, creatinine) that are not readily
‘severe sepsis’ was eliminated, and sepsis was formally redefined available at the bedside outside of an acute care setting. Thus,
as ‘life-threatening organ dysfunction caused by a dysregulated the more cumbersome SOFA score was adapted to a quick SOFA
(‘qSOFA’) to allow for simple, rapid assessments in the pre-
hospital, clinic or emergency department environment (see
Box 1). In the setting of suspected infection, an assessment on a
Systemic inflammatory response syndromea 3-point qSOFA scale evaluating for encephalopathy, tachypnea
and a systolic blood pressure <100mgHg that results in a score of
Two or more of the following: 2 or greater identifies patients at risk for sepsis. Septic shock, a
subset of sepsis, is recognized as sepsis with concomitant lactic
Temperature <36 C or >38 C
acidosis and persistent vasopressor requirements after adequate
Heart rate >90/min
volume resuscitation.
Respiratory rate >20/min OR PaCO2<32 mmHg
WBC <4000/mm3 OR >12,0000 mm3, or 10%
Management
bands
a
The cornerstone of effective sepsis management is indisputably
Bone RC, Balk RA, Cerra FB, et al. American College of Chest Physicians/So-
ciety of Critical Care Medicine Consensus Conference: definitions for sepsis the timely identification of septic patients and the prompt initi-
and organ failure and guidelines for the use of innovative therapies in sepsis. ation of aggressive but appropriate treatment. Successful treat-
Crit Care Med. 1992;20(6):864e874. ment requires fluid resuscitation with a focus on perfusion and
the early administration of antibiotics. The most recent 2018
Table 1

Comparison of Sepsis-2 and Sepsis-3a

Sepsis-2 Sepsis-3

Publication year 2001 2016

Mechanism (pathophysiology) Physiological inflammatory response (SIRS) to Dysregulated immune response to infection
infection
Spectrum SIRS -> Sepsis -> Severe Sepsis -> Septic Sepsis -> Septic shock
shock -> MODS
Predictive validity for in-hospital mortality 0.64 (95% CI, 0.62e0.66) 0.74 (95% CI, 0.73e0.76)
(area under the curve, p <0.001)
Sensitivity/specificity [ Sensitivity Y Sensitivity
Y Specificity [ Specificity
Definition of septic shock Acute circulatory failure characterized by Subset of sepsis in which underlying
persistent arterial hypotension unexplained by circulatory and cellular metabolism
other causes abnormalities are profound enough to
substantially increase mortality
Practical considerations 3 of 4 SIRS criteria obtainable at bedside Multiple laboratory test results required to
without need for laboratory testing calculate SOFA score; qSOFA can be used as
substitute prior to test availability
a
Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801e810.

Table 2

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 677 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

Sequential (sepsis-related) Organ Failure Assessment score (SOFA)a

Score

System 0 1 2 3 4
Respiratory
PaO2/FiO2, mmHg 400 <400 <300 <200 with respiratory <100 with respiratory
support support
Coagulation
Platelets, 103/mL 150 <150 <100 <50 <20
Liver
Bilirubin, mg/dL <1.2 1.2e1.9 2.0e5.9 6.0e11.9 >12.0
Cardiovascular
MAP70 mmHg MAP<70 mmHg Dopamine<5 or Dopamine 5e15 or Dopamine>15 or
dobutamine (any dose)b norepinephrine 0.1 or norepinephrine>0.1 or
epinephrine0.1b epinephrine>0.1b
Central nervous system
Glasgow Coma Scale scorec 15 13e14 10e12 6e9 <6
Renal
Creatinine, mg/dL <1.2 1.2e1.9 2.0e3.4 3.5e4.9 >5.0
Urine output, mL/day <500 <200

Abbreviations: FiO2-fraction of inspired oxygen; MAP-mean arterial pressure.

a
Vincent JL, Moreno R, Takala J, et al; Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. The SOFA (Sepsis-related
Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med. 1996;22(7):707e710.
b
Catecholamine doses are given as ug/kg.min for at least 1 hour.
c
Glasgow Coma Scale scores range from 3 to 15; higher score indicates better neurologic function.

Table 3

Surviving Sepsis Campaign (SSC) bundle update introduced the presentation. Ideally, microbiological samples from all suspected
‘1-Hour Bundle’.8 This bundle combines elements of the prior 3- foci of infection should be obtained prior to administering anti-
and 6-Hour Bundles into an algorithm that emphasizes the im- biotics. However, if obtaining these samples will delay antibiotic
mediate treatment of septic patients. The first elements of the administration, then antibiotic therapy should always take pri-
bundle are diagnostic: measuring a lactate level and obtaining ority. In multiple studies, the prompt administration of antibi-
blood cultures prior to administration of antibiotics. The otics has been associated with improved survival.11,12
remainder of the bundle addresses management: early antimi- Appropriate empiric antimicrobial therapy is essential for
crobial administration, fluid resuscitation and, if required, decreasing mortality in septic patients and should be guided by
vasopressor support (see Box 2). the clinical presentation and local antimicrobial resistance pat-
The ‘1-Hour Bundle’ has been met with controversy, howev- terns, as well as the patient’s risk factors for particular organisms
er, and there is a lack of high-quality prospective evidence sup- (see Table 5).12,13 Chronic comorbidities (e.g. human immuno-
porting the shift.9 Additionally, in the era of quality metrics, deficiency virus, combined variable immunodeficiency, diabetes
many groups have expressed concerns that compression of the mellitus) and the presence of invasive medical devices (e.g.
timeline could lead to the overdiagnosis of sepsis and thus central venous catheters, urinary catheters) should also be taken
overuse of antibiotics and excessive volume resuscitation.10 into account when selecting initial antibiotic regimens.
When the source of infection is not clear, empiric broad-
Early antimicrobial therapy and source control spectrum antimicrobial therapy should be administered. For
For patients with presumed sepsis, early empiric broad-spectrum the majority of patients, empiric antibiotic coverage should
antibiotic therapy should be initiated within 1 hour of include either an extended spectrum beta lactam, a third- or

SEPSIS-3 criteria for sepsis/septic shocka

Sepsis qSOFA2 plus evidence of infection
Septic shock Sepsis plus persistent hypotension requiring administration of vasopressors to maintain a MAP>65 mmHg and a lactate >2mmol/L
despite adequate fluid resuscitation
a
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):801e810.

Table 4

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 678 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

infusion of 30 mL/kg of intravenous crystalloid fluid within 1

Quick Sequential Organ Failure Analysis (qSOFA) scorea hour of sepsis identification if hypotension is present.11 However,
qSOFA criteria:  2 of the following the type of fluid to be used for resuscitation remains an active
Respiratory rate >22/min area of debate. In 2004, the SAFE trial compared clinical out-
Change in mental status comes in critically ill patients receiving volume resuscitation
Systolic blood pressure <100 mmHg with either normal saline or albumin.18 No significant difference
in all-cause mortality was noted between the two groups, though
a
Adapted from: Singer M, Deutschman CS, Seymour CW, et al. The Third In- a sub-group analysis did show unfavourable outcomes when
ternational Consensus Definitions for Sepsis and Septic Shock (Sepsis-
3). JAMA. 2016;315(8):801e810.
albumin was administered to patients with traumatic brain
injury. The combination of the significantly increased expense of
Box 1 albumin infusion and an absence of data to support a clear
benefit led the SSC to recommend crystalloids as the volume
resuscitation agents of choice.
fourth-generation cephalosporin, or a carbapenem. Additional Which crystalloid to use is also a matter of active debate. Two
consideration should be paid to methicillin-resistant Staphylo- recent large randomized controlled trials (RCTs) found mortality
coccus aureus (MRSA) risk factors and, if present, empiric van- benefits, less renal dysfunction, and less need for renal replace-
comycin administration is advised. Combination therapy (the use ment therapy with the use of balanced salt solutions such as
of multiple antibiotics with different pharmacodynamic profiles lactated Ringer’s or PlasmaLyte when compared to 0.9% saline
and mechanisms of action to treat the same organism) is a topic solution.19,20 Because volume resuscitation is a central part of
of some debate, though some early studies reported a synergistic sepsis management, a secondary analysis of the SMART trial was
effect with the addition of an aminoglycoside to a beta-lactam,14 performed focusing on critically ill patients with sepsis. The re-
and later studies assessed the addition of a fluoroquinolone. sults were similar to the prior studies and clearly demonstrated a
Combination therapy has been shown to improve survival in the lower 30-day mortality in septic patients who received balanced
most critically ill patients (i.e. in septic shock where the mortality crystalloids.21
risk is greater than 25%),15 though its role in ‘routine’ sepsis is While adequate resuscitation is essential, fluid overload is
far less clear. In the appropriate clinical context (e.g. suspected associated with increased mortality,22 making perfusion assess-
influenza), empiric anti-viral therapy is often appropriate. In the ment essential to the treatment algorithm. Careful attention to
setting of immunosuppression, total parenteral nutrition or fluid management with de-resuscitation 24e48 hours after suc-
recent abdominal surgery, empiric anti-fungal therapy can be cessful resuscitation decreases the ICU length of stay and in-
considered. Once microbial data are available, prompt de- creases ventilator-free days. Silversides et al. found that a
escalation to a targeted antimicrobial is safe and may even positive fluid balance at day 3 was an independent risk factor for
decrease patient mortality.16 increased mortality.23 Clinicians should aim to achieve an even
Source control has been recognized as an increasingly to net negative fluid balance at 72 hours and keep cumulative
important element of optimal sepsis treatment. The removal of fluid balance low to avoid end-organ dysfunction resulting from
potentially infected invasive devices is highly recommended, volume overload. The REDUCE trial is an ongoing randomized
particularly urinary and central venous catheters. Prompt controlled trial aiming to assess the efficacy and feasibility of
consultation with surgical or interventional radiology services is protocolized de-resuscitation in patients recovering from septic
required to address scenarios such as an empyema, joint infec- shock.24
tion, cholecystitis or intra-abdominal abscess. Generally, the
least invasive intervention method of source control is the Perfusion assessment
preferred option, though timeliness is also a major factor in In the assessment of perfusion in sepsis, critical care practitioners
selecting between competing options. A multidisciplinary have historically focused on normalizing macro-haemodynamic
approach to early intervention (<6 hours) can improve patient parameters. The targeted goals for resuscitation have evolved
mortality.17 from static haemodynamic targets (e.g. heart rate, mean arterial
pressure, central venous pressure) to dynamic variables such as
Fluid resuscitation pulse pressure, stroke volume variation, plethysmography vari-
Prompt fluid resuscitation should be initiated upon the diagnosis ability index, and gauging responses to fluid challenge using IVC
of sepsis. Current guidelines strongly recommend initiating the collapsibility and passive leg raising.

Surviving Sepsis Campaign-1 Hour Bundlea

C Measure lactate level (follow serial measurements if initial level >2mmol/L)
C Obtain blood cultures prior to administering antibiotics
C Administer broad spectrum antibiotics
C Begin rapid administration of 30 mL/kg of crystalloid for hypotension or lactate 4mmol/L
C Start vasopressors if patient is hypotensive during or after fluid resuscitation to maintain a MAP 65mmHg
a
Levy MM, Evans LE, Rhodes A. The surviving sepsis campaign bundle: 2018 update. Intensive care medicine. 2018 Jun 1;44(6):925e8.

Box 2

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 679 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

Risk factors from select organisms

Pseudomonas aeruginosa (and other MRSA VRE Invasive candidiasis
resistant Gram-negative rods)

Community acquired: C Colonization with MDROs C Liver transplant C Central venous catheter
C IV antibiotic use within 90 days C Recent MRSA infection C Known colonization C Broad spectrum antibiotics
C Colonization with MDROs C Known MRSA colonization C Prolonged use of broad C Plus one of the following
Hospital acquired: C Purulence or abscess of the skin or IV spectrum antibiosis  Parenteral nutrition
C IV antibiotics within 90 days access site C Profound  Dialysis
C 5 or more days of hospitalization C Severe rapidly progressive necro- immunosuppression  Recent abdominal
prior to onset tizing pneumonia surgery
C Requiring acute renal replacement  Necrotizing pancreatitis
therapy  Immunosuppressive
C Septic shock agents
C Colonization with MDROs

Adapted from Derensinski, Stan. ‘Severe Sepsis and Septic Shock Antibiotic Guide.’ Stanford Antimicrobial Safety and Sustainability Program. Stanford Health. May 2017.
http://med.stanford.edu/bugsanddrugs/guidebook/_jcr_content/main/panel_builder_1454513702/panel_0/download_1586531681/file.res/Sepsis%20ABX%202017-05-
25.pdf.

Table 5

However, persistent alterations of microcirculatory blood flow adequate IV fluid resuscitation, vasopressors should be admin-
despite restoring macro-haemodynamic parameters can lead to istered. At present, the SSC recommends noradrenaline/norepi-
organ failure. At least one study has shown that microcirculatory nephrine as the first-line vasopressor.28 If a second agent is
perfusion alterations predict mortality during serious infection, necessary to achieve MAP goals, vasopressin or adrenaline/
whereas mean arterial pressure or cardiac output did not.25 epinephrine are the recommended options.
This dissociation between the macrocirculatory and micro- Patients with septic shock are hypothesized to have a ‘relative
circulatory compartments brings a need to assess end organ tis- vasopressin deficiency’, meaning that levels of vasopressin are
sue perfusion in patients with septic shock and to develop tools lower than expected for a shock state.29 Thus, the addition of low
to analyse microcirculatory blood flow. For example, persistent dose vasopressin or terlipressin can be an effective treatment for
hyperlactatemia has been traditionally considered as a signal of refractory shock. Furthermore, the addition of vasopressin has
tissue hypoperfusion or hypoxia. Therefore, lactate normaliza- been noted to have a ‘dose sparing’ effect on norepinephrine
tion is recommended as a resuscitation target in recent requirements, potentially decreasing the risk of tachyar-
guidelines.11 rhythmia30,31 and other catecholamine-associated side-effects.32
At the same time, peripheral perfusion is a promising alter- In low doses, adrenaline/epinephrine is an inotrope, but at
native target. There is particular interest among the scientific higher doses it can manifest vasoconstrictive properties. A head-
community for using the skin, an easily accessible organ, as a to-head comparison of norepinephrine to epinephrine in septic
surrogate for deeper organ perfusion by using skin temperature shock showed no difference in mortality, but increased adverse
gradients, capillary refill time (CRT), the extent of mottling, and events in the epinephrine arm.31 It is important to note that
the peripheral perfusion index. epinephrine infusions almost invariably cause hyperlactatemia,
In the first major RCT using peripheral perfusion as an assay limiting the usefulness of lactate clearance as an indicator of
for microcirculatory dysfunction (ANDROMEDA-SHOCK), a response to therapy.
CRT-targeted resuscitation strategy did not demonstrate a The choice between adrenaline/epinephrine and vasopressin
reduction in all-cause 28-day mortality when compared to as a second-line agent should be based upon the patient’s current
lactate-targeted fluid resuscitation. However, it did demonstrate haemodynamic status and underlying physiology (including
that CRT could be used to reliably follow dynamic changes in possible septic cardiomyopathy), particularly the presence of (or
regional and microcirculatory flow parameters and hypoxia risk for) tachyarrhythmias or elevated lactate levels.
surrogates during sepsis resuscitation.26 Given the subjectivity of Previously considered the first-line vasopressor for septic
CRT, several new technologies have been developed to more shock, dopamine has fallen out of favour. A 2015 meta-
objectively measure capillary refill time, and other methods for analysis compared the use of norepinephrine to dopamine in
the objective assessment of peripheral perfusion are in develop- patients with septic shock and found decreased all-cause
ment, all in the hope of identifying better microcirculatory targets mortality in the norepinephrine treated patients.33 The abun-
for resuscitation.27 The overall efficacy of microcirculation- dance of data on its arrhythmogenic properties in conjunction
targeted resuscitation is still to be determined, however. with a less favourable haemodynamic profile overall makes
dopamine a less desirable choice in the treatment of septic
Vasopressors shock. Consequently, it should be reserved for a select patient
Mean arterial pressure (MAP) is the primary driving factor population (e.g. septic patients with bradycardia or low risk of
behind systemic perfusion. When hypotension persists despite tachycardia).34

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 680 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

There are little data about the use of phenylephrine in patients sensitivity, specificity or cost efficacy necessary to justify use in
with septic shock. Due to its potential to cause splanchnic clinical practice.38
vasoconstriction, and an absence of robust clinical data sup- Procalcitonin (PCT) is the most studied sepsis biomarker.39 A
porting its safety in septic shock, its use should be limited in precursor to the hormone calcitonin, levels are routinely elevated
patients with sepsis.33 in the setting of systemic infection. Thus far, studies show pro-
Angiotensin II (ATII) is the most recent vasopressor to calcitonin to be useful in decreasing the duration of antibiotic
receive European Medicines Agency approval. With the publi- treatment, decreasing costs and length of hospitalization.40 Ele-
cation of the ATHOS-3 trial, a novel option for the treatment of vations of PCT can be seen in uninfected patients, however,
catecholamine resistant vasodilatory shock was introduced. In particularly in patients with renal disease. Based upon available
this multi-centre RCT the group treated with ATII demonstrated data, the SSC suggests using procalcitonin only to limit or dis-
a significant increase in blood pressure and no significant in- continue antibiotic therapy. An elevated procalcitonin in itself is
crease in adverse events.35 ATII appears to be particularly insufficient to determine the initiation (or not) of antibiotics.28
effective in patients with acute kidney injury (AKI), as a subset
analysis of the ATHOS-3 cohort demonstrated a mortality Molecular techniques
benefit in patients with AKI requiring renal replacement ther- Identifying and treating the offending organism in a timely
apy.36 Additionally, results from a recent study showed post- manner is often a challenge. Conventional blood cultures can
cardiac surgery vasoplegia patients had a favourable haemo- take up to 72 hours to provide a definitive result, and recently
dynamic response to ATII, permitting rapid weaning of cate- developed molecular techniques can detect pathogens far more
cholamine vasopressors.37 quickly. Ongoing evaluations of pathogen identification tech-
A subset of patients with septic shock will develop septic car- niques using multiplex polymerase chain reaction (PCR), PCR/
diomyopathy. These patients typically have little cardiac reserve at electrospray ionized mass spectroscopy (ESI-MS) and function-
baseline and are unable to generate a compensatory cardiac output alized nanoparticles coupled with magnetic resonance have
during vasodilatory shock. If cardiac output remains low despite demonstrated the capability of identifying organisms in under 6
use of vasopressors, initiation of inotropic therapy is appropriate. hours with reasonable sensitivity and specificity. Further studies
The Surviving Sepsis Campaign Guidelines recommend dobut- and cost analyses will determine whether these techniques are
amine as the preferred inotrope, though data show similar im- incorporated into general clinical practice, but in the near future
provements in cardiac output with the use of epinephrine as a single these techniques may very well obviate the need for conven-
agent as compared to dobutamine paired with norepinephrine.11 tional blood cultures.41

Diagnostic techniques Adjunctive therapies

Microbiological cultures Steroids

Positive blood cultures are demonstrable evidence of systemic Despite the physiological plausibility for the effectiveness of
infection. If sepsis is suspected, current guidelines recommend steroids in septic shock, the data are conflicting on their efficacy
drawing two sets of blood cultures, both aerobic and anaerobic. in clinical practice. In 2015, a review by Annane et al. analysed
However, blood culture yield is variable and is dependent upon 33 trials and concluded that corticosteroids generally reduce
sampling technique. Care must be taken to adequately prepare mortality among patients with sepsis (though the evidence was
the skin with antiseptic agent and to inoculate each bottle with a rated as low quality).42 Based on this, the 2016 SSC guidelines
minimum of 10 mL of blood. Ideally, cultures should be obtained recommended 200 mg per day of IV hydrocortisone for patients
prior to the administration of antibiotics. Cultures of other bodily who have persistent circulatory failure despite adequate fluid
fluids should also be obtained as clinically indicated (e.g. resuscitation and vasopressor therapy.
sputum, urine, cerebrospinal fluid, etc.). It should be noted, however, that two major steroid trials have
been completed since the last SSC guidelines. The ADRENAL trial
Lactate randomized septic shock patients to receive hydrocortisone
Elevated serum lactate is a marker of disease severity in sepsis. versus placebo. No differences in mortality were noted, but there
The etiology of the rise in lactate is multifactorial, including was a significant improvement in reversal of shock, length of stay
anaerobic metabolism resulting from inadequate oxygen delivery in ICU, ventilator-free days and fewer blood transfusions.43
and accelerated aerobic glycolysis. Several randomized Conversely, another multi-centre, randomized controlled trial
controlled trials reported that a lactate-guided resuscitation administering hydrocortisone plus fludrocortisone (versus pla-
strategy in patients with septic shock reduced mortality.28 As cebo) to patients in septic shock did find a reduction in all-cause
such, the measurement of lactate is a mainstay of sepsis treat- mortality in the treatment group, along with shorter durations of
ment protocols and is included in the current clinical criteria for shock, ICU length of stay and organ support.44 Much like the
the diagnosis of septic shock. evidence, the critical care community remains divided on the use
of steroids in septic shock.
Biomarkers
Numerous novel biomarkers (e.g. Supar, Presepsin, cell-free Vitamin C ( thiamine)
plasma DNA, etc.) are under study with the aim of improving Marik et al. made headlines in 2017 with the publication of a
diagnostic precision and prognostication, and ultimately assess- small single-centre retrospective trial that showed an impressive
ing treatment response. However, most have not shown the mortality benefit with the administration of high dose vitamin C,

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 681 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

thiamine and hydrocortisone (VCTHC) to patients with septic Anticoagulants (including thrombomodulin and
shock.45 Since 2017, however, several groups have attempted to heparin)
replicate these results without success. Fujii et al. published a The use of thrombomodulin for treatment of sepsis induced
prospective, open-label RCT in 2020 that ultimately concluded disseminated intravascular coagulation (DIC) has been well-
that the cocktail of VCTHC did not significantly change the established in Japan for several years,58 and the SCARLET trial
duration of vasopressor use or improve mortality.46 Likewise, sought to determine if the benefit would extend to a more het-
Chang et al. failed to detect a mortality benefit in their single- erogeneous population. Unfortunately, this large international
blinded RCT.47 The long-awaited VICTAS trial, meant to be the RCT failed to demonstrate any mortality benefit in patients with
largest multi-centre, randomized, double-blinded trial on the sepsis associated coagulopathy treated with thrombomodulin.59
topic, was terminated early due to futility, failing to detect a Other anticoagulants, such as antithrombin and heparin have
significant difference in mortality, duration of vasopressor use or also been studied,60,61 but neither are recommended as part of
ventilator days (though it may have been underpowered to detect routine treatment at this time, though prophylactic anticoagulant
a mortality difference).48 Overall, the results from follow-up use remains recommended in the SSC Guidelines.28
studies on vitamin C have been underwhelming to say the least.
Intravenous thiamine has not been shown to improve mor- Intravenous immunoglobulin
tality in septic shock, whether alone or in combination with Acquired immunosuppression leading to low endogenous
vitamin C or hydrocortisone.49 immunoglobulin levels may impact mortality in patients with
sepsis.62 Thought to mitigate the inflammatory response and
Blood purification techniques augment immune mechanisms, intravenous immunoglobulin
Endotoxin is circulating in 50% of patients in septic shock and is (IVIG) has the potential to modulate both pro- and anti-
associated with poor outcomes.50 The EUPHRATES trial inflammatory processes. Thus far, however, IVIG (especially
attempted to neutralize endotoxin via haemoperfusion with a isolated IgG IVIG) has not been shown to improve mortality in
polymyxin fibre device.51 EUPHRATES was the largest RCT to patients with sepsis.63 While the infusion of a polyclonal IVIG
study this technology but was terminated early due to failure to containing IgG, IgA and IgM showed more promising results
achieve the primary mortality endpoint. Though EUPHRATES including a modest reduction in mortality, reliable data regarding
failed to show a mortality benefit in the overall original trial timing and dosing of IVIG are still too insufficient to warrant its
population, a post-hoc analysis found that polymyxin B haemo- routine use.64,65 As such, IVIG is not currently recommended as a
perfusion decreased mortality in a subset of patients with mod- therapeutic agent for the treatment of sepsis.
erate endotoxin activity as opposed to extreme endotoxin activity
(EAA).52 Based on these findings, the TIGRIS trial is currently Conclusions
enrolling patients to evaluate similar haemoperfusion techniques
Sepsis and septic shock are leading contributors to worldwide
in patients with moderate EAA septic shock.53 Other techniques,
morbidity and mortality. The current definitions of sepsis and
such as extracorporeal plasma filtration to remove pro- and anti-
septic shock are based upon an expanded understanding of
inflammatory mediators from the bloodstream are under evalu-
sepsis as a dysregulated immune response as opposed to solely
ation.54 While the concept remains appealing, at this time data
an inflammatory process. Early recognition and prompt man-
are lacking to support the adoption of any blood purification
agement, such as the implementation of the 1-Hour Bundle, are
technologies for sepsis.
essential to improving patient outcomes. The core elements of
Glycaemic control any treatment algorithm remain the same: antibiotics, source
Glycaemic control in sepsis is a much-debated topic. On one control, intravenous fluid resuscitation and vasopressors,
hand, hyperglycaemia can exert an immunosuppressive effect tempered with the recognition that sepsis is not a static process,
diminishing the body’s ability to mount a competent immune and that frequent reassessment and revision of the treatment
response. On the other, the anti-inflammatory effects of insulin plan are essential. A number of potential targeted adjunctive
may inhibit cellular autophagy and negatively impact a patient’s therapies and novel diagnostic techniques remain on the
capacity to clear infection.55 horizon, though none yet have a role in routine clinical
Evidence shows that severe hyperglycaemia in patients with practice. A
sepsis is associated with increased 30-day mortality.56 On the
other hand, the NICE-SUGAR trial, a large, multi-centre, ran-
domized controlled trial compared intensive insulin control (81 REFERENCES
e108 mg/dL) to conventional insulin control (<180 mg/dL), 1 Rudd KE, Johnson SC, Agesa KM, et al. Global, regional, and
found increased mortality due to hypoglycaemia with intensive national sepsis incidence and mortality, 1990-2017: analysis for
insulin therapy.57 As both hyperglycaemia and hypoglycaemia the Global Burden of Disease Study. Lancet (London, England)
are associated with adverse outcomes, current recommendations 2020; 395: 200e11.
are to start insulin after blood glucose levels exceed 180 mg/dL 2 Hex N. The cost of sepsis care in the UK. 2017, http://allcatsgrey.
(10 mmol/L) with the goal of maintaining levels less than org.uk/wp/dowload/health_economics/YHEC-Sepsis-Report-17.
180 mg/dL. There is no specific lower threshold target, but 02.17-FINAL.pdf.
avoidance of hypoglycaemia and wide swings in glucose levels is 3 Wasyluk W, Zwolak A. Metabolic alterations in sepsis. J Clin Med
advised. 2021; 10.

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 682 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

4 Singer M, Deutschman CS, Seymour CW, et al. The third inter- 22 Acheampong A, Vincent JL. A positive fluid balance is an inde-
national consensus definitions for sepsis and septic shock pendent prognostic factor in patients with sepsis. Crit Care
(Sepsis-3). JAMA 2016; 315: 801e10. (London, England) 2015; 19: 251.
5 Martin MD, Badovinac VP, Griffith TS. CD4 T cell responses and 23 Silversides JA, Major E, Ferguson AJ, et al. Conservative fluid
the sepsis-induced immunoparalysis state. Front Immunol 2020; management or deresuscitation for patients with sepsis or acute
11: 1364. respiratory distress syndrome following the resuscitation phase of
6 Sjaastad FV, Kucaba TA, Dileepan T, et al. Polymicrobial sepsis critical illness: a systematic review and meta-analysis. Intensive
impairs antigen-specific memory CD4 T cell-mediated immunity. Care Med 2017; 43: 155e70.
Front Immunol 2020; 11: 1786. 24 Messmer A, Pfortmueller C. Protocolised early de-resuscitation in
7 Kaukonen KM, Bailey M, Pilcher D, Cooper DJ, Bellomo R. Sys- septic shock (REDUCE) 2021. Inselspital, Bern, Switzerland:
temic inflammatory response syndrome criteria in defining severe University Hospital Bern. Identification No. NCT04931485.
sepsis. New Engl J Med 2015; 372: 1629e38. Recruiting. Retrieved from: https://clinicaltrials.gov/ct2/show/
8 Levy MM, Evans LE, Rhodes A. The surviving sepsis campaign NCT04931485.
bundle: 2018 update. Intensive Care Med 2018; 44: 925e8. 25 De Backer D, Donadello K, Sakr Y, et al. Microcirculatory alter-
9 Kalantari A, Rezaie SR. Challenging the one-hour sepsis bundle. ations in patients with severe sepsis: impact of time of assess-
West J Emerg Med 2019; 20: 185e90. ment and relationship with outcome. Crit Care Med 2013; 41:
10 Marik PE, Farkas JD, Spiegel R, Weingart S. POINT: should the 791e9.
surviving sepsis campaign guidelines be retired? Yes. Chest 26 Herna ndez G, Ospina-Tasco  n GA, Damiani LP, et al. Effect of a
2019; 155: 12e4. resuscitation strategy targeting peripheral perfusion status vs
11 Levy MM, Evans LE, Rhodes A. The surviving sepsis campaign serum lactate levels on 28-day mortality among patients with
bundle: 2018 update. Crit Care Med 2018; 46: 997e1000. septic shock: the ANDROMEDA-SHOCK randomized clinical trial.
12 Kumar A, Ellis P, Arabi Y, et al. Initiation of inappropriate antimi- JAMA 2019; 321: 654e64.
crobial therapy results in a fivefold reduction of survival in human 27 Castro R, Kattan E, Ferri G, et al. Effects of capillary refill time-vs.
septic shock. Chest 2009; 136: 1237e48. lactate-targeted fluid resuscitation on regional, microcirculatory
13 Rhee C, Kadri SS, Dekker JP, et al. Prevalence of antibiotic- and hypoxia-related perfusion parameters in septic shock: a
resistant pathogens in culture-proven sepsis and outcomes randomized controlled trial. Ann Intensive Care 2020; 10: 150.
associated with inadequate and broad-spectrum empiric anti- 28 Rhodes A, Evans LE, Alhazzani W, et al. Surviving sepsis
biotic use. JAMA Netw Open 2020; 3: e202899. campaign: international guidelines for management of sepsis and
14 Giamarellou H. Aminoglycosides plus beta-lactams against gram- septic shock: 2016. Intensive Care Med 2017; 43: 304e77.
negative organisms. Evaluation of in vitro synergy and chemical 29 Landry DW, Levin HR, Gallant EM, et al. Vasopressin deficiency
interactions. Am J Med 1986; 80: 126e37. contributes to the vasodilation of septic shock. Circulation 1997;
15 Kumar A, Zarychanski R, Light B, et al. Early combination anti- 95: 1122e5.
biotic therapy yields improved survival compared with mono- 30 McIntyre WF, Um KJ, Alhazzani W, et al. Association of vaso-
therapy in septic shock: a propensity-matched analysis. Crit Care pressin plus catecholamine vasopressors vs catecholamines
Med 2010; 38: 1773e85. alone with atrial fibrillation in patients with distributive shock: a
16 Routsi C, Gkoufa A, Arvaniti K, et al. De-escalation of antimicro- systematic review and meta-analysis. JAMA 2018; 319:
bial therapy in ICU settings with high prevalence of multidrug- 1889e900.
resistant bacteria: a multicentre prospective observational cohort 31 Myburgh JA, Higgins A, Jovanovska A, et al. A comparison of
study in patients with sepsis or septic shock. J Antimicrob Che- epinephrine and norepinephrine in critically ill patients. Intensive
mother 2020; 75: 3665e74. Care Med 2008; 34: 2226.
17 Grek A, Booth S, Festic E, et al. Sepsis and shock response team: 32 Hartmann C, Radermacher P, Wepler M, Nuûbaum B. Non-he-
impact of a multidisciplinary approach to implementing surviving modynamic effects of catecholamines. Shock (Augusta, Ga.)
sepsis campaign guidelines and surviving the process. Am J Med 2017; 48: 390e400.
Qual: Off J Am Coll Med Qual 2017; 32: 500e7. 33 Avni T, Lador A, Lev S, Leibovici L, Paul M, Grossman A. Vaso-
18 Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R. pressors for the treatment of septic shock: systematic review and
A comparison of albumin and saline for fluid resuscitation in the meta-analysis. PLos One 2015; 10: e0129305.
intensive care unit. New Engl J Med 2004; 350: 2247e56. 34 Einav S, Helviz Y, Ippolito M, Cortegiani A. Vasopressor and
19 Semler MW, Self WH, Wanderer JP, et al. Balanced crystalloids inotrope treatment for septic shock: an umbrella review of re-
versus saline in critically ill adults. New Engl J Med 2018; 378: views. J Crit Care 2021; 65: 65e71.
829e39. 35 Khanna A, English SW, Wang XS, et al. Angiotensin II for the
20 Self WH, Semler MW, Wanderer JP, et al. Balanced crystalloids treatment of vasodilatory shock. New Engl J Med 2017; 377:
versus saline in noncritically ill adults. New Engl J Med 2018; 378: 419e30.
819e28. 36 Tumlin JA, Murugan R, Deane AM, et al. Outcomes in patients
21 Brown RM, Wang L, Coston TD, et al. Balanced crystalloids with vasodilatory shock and renal replacement therapy treated
versus saline in sepsis. A secondary analysis of the with intravenous angiotensin II. Crit Care Med 2018; 46: 949e57.
SMART clinical trial. Am J Respir Crit Care Med 2019; 200: 37 Klijian A, Khanna AK, Reddy VS, et al. Treatment with angiotensin
1487e95. II is associated with rapid blood pressure response and

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 683 Ó 2021 Published by Elsevier Ltd.
 INTENSIVE CARE

vasopressor sparing in patients with vasoplegia after cardiac septic shock and elevated endotoxin level: the EUPHRATES
surgery: a post-hoc analysis of angiotensin II for the treatment of randomized clinical trial. JAMA 2018; 320: 1455e63.
high-output shock (ATHOS-3) study. J Cardiothorac Vasc Anesth 52 Klein DJ, Foster D, Walker PM, Bagshaw SM, Mekonnen H,
2021; 35: 51e8. Antonelli M. Polymyxin B hemoperfusion in endotoxemic septic
38 Pierrakos C, Velissaris D, Bisdorff M, Marshall JC, Vincent JL. shock patients without extreme endotoxemia: a post hoc
Biomarkers of sepsis: time for a reappraisal. Crit Care (London, analysis of the EUPHRATES trial. Intensive Care Med 2018; 44:
England) 2020; 24: 287. 2205e12.
39 Schuetz P, Beishuizen A, Broyles M, et al. Procalcitonin (PCT)- 53 Foster D. Efficacy of polymyxin B hemoperfusion (PMX) for
guided antibiotic stewardship: an international experts consensus endotoxemic septic shock in a randomized, open-label study
on optimized clinical use. Clin Chem Lab Med 2019; 57: 1308e18. (TIGRIS) 2021. Spectral Diagnostics. Identification No.
40 Bouadma L, Luyt C-E, Tubach F, et al. Use of procalcitonin to NCT03901807. Retrieved from: https://clinicaltrials.gov/ct2/show/
reduce patients’ exposure to antibiotics in intensive care units NCT03901807.
(PRORATA trial): a multicentre randomised controlled trial. Lancet 54 Pickkers P, Payen D. What’s new in the extracorporeal treatment
2010; 375: 463e74. of sepsis? Intensive Care Med 2017; 43: 1498e500.
41 Vincent J-L, Brealey D, Libert N, et al. Rapid diagnosis of infection 55 Van Niekerk G, Davis T, Engelbrecht A-M. Hyperglycaemia in
in the critically ill, a multicenter study of molecular detection in critically ill patients: the immune system’s sweet tooth. Crit Care
bloodstream infections, pneumonia, and sterile site infections. Crit 2017; 21: 202.
Care Med 2015; 43: 2283. 56 van Vught LA, Wiewel MA, Klein Klouwenberg PM, et al.
42 Annane D, Bellissant E, Bollaert PE, Briegel J, Keh D, Kupfer Y. Admission hyperglycemia in critically ill sepsis patients: associ-
Corticosteroids for treating sepsis. Cochrane Database Syst Rev ation with outcome and host response. Crit Care Med 2016; 44:
2015; 2015: Cd002243. 1338e46.
43 Venkatesh B, Finfer S, Cohen J, et al. Adjunctive glucocorticoid 57 Investigators N-SS. Intensive versus conventional glucose control
therapy in patients with septic shock. N Engl J Med 2018; 378: in critically ill patients. New Engl J Med 2009; 360: 1283e97.
797e808. 58 Aikawa N, Shimazaki S, Yamamoto Y, et al. Thrombomodulin alfa
44 Annane D, Renault A, Brun-Buisson C, et al. Hydrocortisone plus in the treatment of infectious patients complicated by dissemi-
fludrocortisone for adults with septic shock. New Engl J Med nated intravascular coagulation: subanalysis from the phase 3
2018; 378: 809e18. trial. Shock (Augusta, Ga.) 2011; 35: 349e54.
45 Marik PE, Khangoora V, Rivera R, Hooper MH, Catravas J. Hy- 59 Vincent JL, Francois B, Zabolotskikh I, et al. Effect of a recom-
drocortisone, vitamin C, and thiamine for the treatment of severe binant human soluble thrombomodulin on mortality in patients
sepsis and septic shock: a retrospective before-after study. Chest with sepsis-associated coagulopathy: the SCARLET randomized
2017; 151: 1229e38. clinical trial. JAMA 2019; 321: 1993e2002.
46 Fujii T, Luethi N, Young PJ, et al. Effect of vitamin C, hydrocor- 60 Warren BL, Eid A, Singer P, et al. Caring for the critically ill patient.
tisone, and thiamine vs hydrocortisone alone on time alive and High-dose antithrombin III in severe sepsis: a randomized
free of vasopressor support among patients with septic shock: the controlled trial. JAMA 2001; 286: 1869e78.
VITAMINS randomized clinical trial. JAMA 2020; 323: 423e31. 61 Jaimes F, De La Rosa G, Morales C, et al. Unfractioned heparin
47 Chang P, Liao Y, Guan J, et al. Combined treatment with hydro- for treatment of sepsis: a randomized clinical trial (The HETRASE
cortisone, vitamin C, and thiamine for sepsis and septic shock: a Study). Crit Care Med 2009; 37: 1185e96.
randomized controlled trial. Chest 2020; 158: 174e82. 62 Martin-Loeches I, Muriel-Bombín A, Ferrer R, et al. The protective
48 Sevransky JE, Rothman RE, Hager DN, et al. Effect of vitamin C, association of endogenous immunoglobulins against sepsis
thiamine, and hydrocortisone on ventilator- and vasopressor-free mortality is restricted to patients with moderate organ failure. Ann
days in patients with sepsis: the VICTAS randomized clinical trial. Intensive Care 2017; 7: 44.
JAMA 2021; 325: 742e50. 63 Werdan K, Pilz G, Bujdoso O, et al. Score-based immunoglobulin
49 Qian X, Zhang Z, Li F, Wu L. Intravenous thiamine for septic G therapy of patients with sepsis: the SBITS study. Crit Care Med
shock: a meta-analysis of randomized controlled trials. Am J 2007; 35: 2693e701.
Emerg Med 2020; 38: 2718e22. 64 El-Nawawy A, El-Kinany H, Hamdy El-Sayed M, Boshra N. Intra-
50 Marshall JC, Walker PM, Foster DM, et al. Measurement of venous polyclonal immunoglobulin administration to sepsis syn-
endotoxin activity in critically ill patients using whole blood drome patients: a prospective study in a pediatric intensive care
neutrophil dependent chemiluminescence. Crit Care (London, unit. J Trop Pediatrics 2005; 51: 271e8.
England) 2002; 6: 342e8. 65 Kreymann KG, de Heer G, Nierhaus A, Kluge S. Use of polyclonal
51 Dellinger RP, Bagshaw SM, Antonelli M, et al. Effect of targeted immunoglobulins as adjunctive therapy for sepsis or septic shock.
polymyxin B hemoperfusion on 28-day mortality in patients with Crit Care Med 2007; 35: 2677e85.

ANAESTHESIA AND INTENSIVE CARE MEDICINE 22:11 684 Ó 2021 Published by Elsevier Ltd.