REVIEWS

Sepsis-associated encephalopathy
Teneille E. Gofton and G. Bryan Young
Abstract | Sepsis-associated encephalopathy (SAE) is a diffuse brain dysfunction that occurs secondary to
infection in the body without overt CNS infection. SAE is frequently encountered in critically ill patients in
intensive care units, and in up to 70% of patients with severe systemic infection. The severity of SAE can
range from mild delirium to deep coma. Seizures and myoclonus are infrequent and cranial nerves are almost
always spared, but most severe cases have an associated critical illness neuromyopathy. Development of
SAE probably involves a number of mechanisms that are not mutually exclusive and vary from patient to
patient. Substantial neurological and psychological morbidities often occur in survivors. Mortality is almost
always due to multiorgan failure rather than neurological complications, and is almost 70% in patients with
severe SAE. Further research into the pathophysiology, management and prevention of SAE is needed. This
Review discusses the epidemiology and clinical presentation of SAE. Recent evidence for SAE pathophysiology
is outlined and a diagnostic approach to patients with this syndrome is presented. Lastly, prognosis and
management of SAE is discussed.
Gofton, T. E. & Young, G. B. Nat. Rev. Neurol. 8, 557–566 (2012); published online 18 September 2012; doi:10.1038/nrneurol.2012.183

Introduction
Sepsis-associated encephalopathy (SAE) is commonly is specific for SAE, and SAE diagnosis depends on the
seen in systemically ill patients. The syndrome is defined clinical context and the presence of evidence for an
by diffuse cerebral dysfunction that accompanies sepsis in­fection somewhere in the body.
in the absence of direct CNS infection, structural abnor- The implications of sepsis and SIRS on cerebral func-
mality or other types of encephalopathy (for example, tion are profound, and the body of knowledge regard-
hepatic or renal encephalopathy), as detected by clinical ing SAE is growing. Further research, however, is clearly
or standard laboratory tests. Patients presenting with SAE necessary to improve recovery from SAE and reduce
show evidence of severe systemic infection with features long-term consequences on cerebral function. This
of sepsis or systemic inflammatory response syndrome article provides an overview of current knowledge on
(SIRS).1 SAE manifests as a spectrum of disturbed cere­ the epidemiology and clinical presentation of SAE. We
bral function ranging from mild delirium to coma. As review recent evidence for the pathophysiology of SAE
mortality is increased with severity of SAE,2 early iden- and present a diagnostic approach to patients with this
tification and management of patients with SAE are disease. Lastly, prognosis of SAE and a management
important to reduce associated morbidity and mortality. strategy for the disorder are discussed.
Delirium is often the first manifestation of sepsis, provid-
ing a useful diagnostic clue. SAE should trigger a search Epidemiology
for infection and prompt initiation of appropriate therapy. Sepsis is a SIRS that accompanies an identified or sus-
Cerebral dysfunction seen in SAE reflects the sys- pected infection—a definition established by Bone
temic metabolic, inflammatory and haemodynamic dis- and colleagues. 4 A group of experts attending the
turbances that are associated with SIRS, rather than a International Sepsis Definitions Conference in 2001
direct CNS abnormality. SAE is effectively a diagnosis of defined SIRS as a syndrome presenting with two or more
exclusion, which can be made after meningitis, encepha- of the following symptoms: increased or decreased body
litis, septic emboli from endocarditis and noninfective temperature (>38 °C or <36 °C), hyperventilation (res- Departments of Clinical
systemic inflammation (as seen in pancreatitis, burns piratory rate >20 breaths per min or an arterial partial Neurological Sciences
or trauma) have been ruled out. Despite lack of direct pressure of carbon dioxide <32 mmHg) and abnormal and Medicine, London
Health Sciences
CNS infection, laboratory evidence of CNS dysfunction white blood cell count (>12,000 or <4,000 cells/μl).5 Centre, University
is common in SAE and may manifest as abnormalities in Severe sepsis is defined as sepsis together with organ Hospital, University of
Western Ontario, 339
EEG measurements, somatosensory evoked potentials, dysfunction and hypo­perfusion or hypotension,4 and Windermere Road,
levels of biomarkers of CNS injury (such as neuron-­ septic shock is defined as sepsis together with arterial London, ON N6A 5A5,
specific enolase [NSE] and protein S100b), and in results hypotension despite appropriate fluid resuscitation. 4 Canada (T. E. Gofton,
G. B. Young).
of neuro­radiological tests.3 None of these manifestations Patients with sepsis and encephalo­pathy were previously
diagnosed as having septic encephalo­pathy. However, Correspondence to:
G. B. Young
Competing interests this term has fallen out of favour, because it suggests an bryan.young@
The authors declare no competing interests. active infection within the CNS. lhsc.on.ca

NATURE REVIEWS | NEUROLOGY VOLUME 8 | OCTOBER 2012 | 557

© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS

Key points SAE had significantly higher heart rate, blood lactate
and serum sodium levels, as well as lower platelet count,
■■ Sepsis-associated encephalopathy (SAE) is an early feature of infection in the
body and might appear before other systemic features of sepsis are obvious serum albumin levels and serum pH.2 The source and
■■ SAE has a spectrum of degrees of severity, ranging from delirium to deep coma the aetiology of infection are also important factors:
■■ Hyperventilation can be an early feature and paratonic rigidity might be the biliary tract or intestinal infections are associated with
only neurological finding, besides delirium, in early SAE; advanced disease is greater risk of SAE, and the most commonly implicated
associated with critical illness polyneuropathy in 70% of cases organisms include Staphylococcus aureus, Enterococcus
■■ No specific markers for SAE exist, so diagnosis relies on exclusion of primary feacium, Acinetobacter spp., Pseudomonas aeruginosa
CNS infection and other causes of encephalopathy
and Stenotrophomonas maltophilia. 2 This study has
■■ Although early stages of SAE might have a uniform pathogenesis, multiple
mechanisms, which can operate alone or in combination, could be involved
limitations including a small sample size and being a
in advanced disease retrospective analysis, but the exclusion criteria were
■■ Morbidity and mortality increase with disease severity; early investigation rigo­rous, which ensured that patients with encephalo­
and prompt treatment of underlying infection when delirium presents are, pathy of alter­native aetiologies were excluded from the
therefore, important analysis. Over­all, it seems that patients with advanced-
stage SAE have a higher burden of systemic illness than
those with early-stage disease, and that the source of
Few studies have specifically addressed the incidence infection and causative organism can be important in
and prevalence of SAE.1,2 The studies are varied and most the development of SAE.
do not have a uniform study cohort. Indeed, even the
operational definition of SAE is problematic, as patients Clinical presentation and diagnosis
might develop encephalopathy in the early phases of SAE is one of many causes of delirium. Patients with
infection without meeting the criteria for sepsis, or a SAE can be expected to present with a level of con-
confused state can develop in patients with a remote sciousness that is out of keeping with the degree of any
infection, such as an abdominal, walled-off abscess that is sedative treatment they are receiving, and they might
not ‘septic’. One of the challenges in studying SAE is that show disturbances of sleep–wake cycles or evidence of
without a specific diagnostic test, it remains a syndrome hallucinations, restlessness or agitation, among other
diagnosed by exclusion. symptoms commonly seen in delirium. Apart from the
SAE is thought to be the most common cause of abnormal mental status and gegenhalten (paratonic
encephalopathy in the medical–surgical intensive care rigidity), the findings in the neurological examination
unit (ICU),6 and over half of patients with sepsis have are unremarkable. However, 70% of advanced cases
encephalopathy.7 Incidence of encephalopathy is higher of SAE have an associated critical illness neuromyo-
in patients who have bacteraemia and evidence of renal, pathy.7 With advanced-stage SAE, cranial nerve func-
hepatic or multiorgan failure. As sepsis is a leading cause tion is invariably spared,13 which is a useful factor in
of ICU admissions and 20–50% of patients with sepsis the differential diagnosis of Guillain–Barré syndrome.
have delirium, SAE is highly prevalent in the ICU.8 70% Hyperventilation in SAE is often due to respiratory
of patients with bacteraemia have neurological symp- alkalosis in the early, delirious phase and to metabolic
toms ranging from lethargy to coma, and >80% have acidosis in advanced-stage sepsis.14 As patients with SAE
abnormalities on EEG.9,10 Approximately half (46%) of have sepsis by definition, pulse rate is also commonly,
patients with bacteraemia have SAE.1,9 Patients with but not invariably, elevated.5
acutely altered mental status associated with encephalo­ SAE is often suspected in patients presenting with
pathy have higher mortality rates (49%) than patients acutely altered mental status accompanied by sepsis or
with pre-existing mental status changes (41%) or normal septic shock (Figure 1). When evaluating a patient with
mental status (26%).11 A recent international study (the sepsis and delirium, the primary goal should be to inves-
Delirium Epidemiology in Critical Care study) involv- tigate and eliminate the possibility of a primary CNS
ing 497 patients in ICU showed that the prevalence of pathology that might cause delirium. Some common
delirium was 32.3% and that sepsis was the leading cause primary neurological disturbances that can lead to
of medical illnesses that required ICU admission.12 delirium include cerebral abscess, encephalitis, menin-
In another study from China, Zhang et al.2 retrospec- gitis, stroke, nonconvulsive status epilepticus or, more
tively reviewed patients admitted to ICU over a 3‑year rarely, primary CNS vasculitis.15,16 Frequent confound-
period. The exclusion criteria included pre-existing or ing factors, such as sedative medications, can make
chronic liver or kidney failure, severe electrolyte imbal- evaluation of mental status challenging.
ances, blood glucose disturbances, CNS infections or Patients suspected of having SAE should undergo a
pre-existing CNS disease, previous cardiac arrest with thorough investigation through clinical history, medi-
resuscitation or previous treatment with sedative drugs. cation review, physical examination, laboratory tests,
Out of 323 patients, 41 had SAE and the rest had sepsis neuro­imaging and EEG to assess anatomical abnormali-
without encephalopathy. This analysis demonstrated that ties, bearing in mind that criteria for sepsis might not
patients with higher APACHE II (Acute Physiology and yet be met (Box 1), but aiming to determine whether
Chronic Health Evaluation II) scores and lower scores on the patient has an infection. The clinical history should
the Glasgow coma scale (GCS) are more likely to have include a search for evidence of prior CNS pathology,
SAE.2 Compared with patients without SAE, those with hepatic or renal dysfunction and prior episodes of

558 | OCTOBER 2012 | VOLUME 8  www.nature.com/nrneurol

© 2012 Macmillan Publishers Limited. All rights reserved
 REVIEWS

delirium associated with severe illness or recent infec- Febrile, delirious or

tions. A detailed review of medications may provide comatose patient
information on recent or discontinued medications,
such as benzodiazepines, other psychoactive agents or
Other conditions (such as
recent antibiotics, which might cause encephalopathy NMS, MH, thyroid storm,
Underlying No drug fever, hypothalamic or
or delirium. A history of overdose of or withdrawal brainstem stroke, heat
infection?
from psychoactive medications might suggest an alter- prostration, cancer, trauma,
pancreatitis and seizures)
native aetiology for reduced level of consciousness or Yes
delirium, and a recent history of infection requiring
antibiotics can help to pinpoint a source of sepsis, and CNS infection, Yes
Appropriate therapy
thereby lead to diagnosis of SAE. A thorough physical endocarditis?
examination should be done to investigate a source of No
infection (for example, infection of a decubitus ulcer or
other types of cellulitis or rash), to assess the evidence
SAE diagnosis
for organ dysfunction (for example, scleral icterus and Treat infection
jaundice suggest hepatic dysfunction) and to identify any Provide supportive care
focal neurological abnormalities, including hemi­paresis Figure 1 | Diagnostic approach to patients showing
or cranial nerve abnormalities, which might suggest a symptoms of SAE. A systematic approach helps to avoid
focal neurological process, such as an abscess or a stroke. missing other potential contributors to delirium.
Importantly, the clinician should attend promptly to any Abbreviations: MH, malignant hyperthermia; NMS,
changes in the mental status of the patient, especially to neuroleptic malignant syndrome; SAE, sepsis-associated
problems associated with attention or concentration. encephalopathy.
Multiple screening tools for detection of delirium
are available, but none is specific for SAE. For practical Box 1 | Laboratory and pathological features of SAE
purposes, we advocate using the Confusion Assessment
Method (CAM) for patients outside the ICU. The CAM Neuroimaging
Normal
has recently been validated for use in diagnosis of
Multiple ischaemic strokes
delirium in this setting and shows excellent sensitivity
Nonspecific subcortical white matter lesions
(94–100%) and positive predictive value (91–94%).17,18
Cerebrospinal fluid analysis
However, the version of the CAM that was developed for
Elevated protein levels
ICU patients (CAM-ICU) had only about 41–47% sensi- Normal glucose levels and cell culture analysis
tivity when used by regular ICU nurses,19,20 although the
EEG recordings
specificity was excellent (98%). Patients with hypo­active Normal
delirium are most likely to be missed with the CAM. Theta waves
Unstructured delirium assessments or the Intensive Delta waves
Care Delirium Screening Checklist seem to have greater Triphasic waves
sensitivity for detecting delirium in the ICU than does Burst–suppression pattern
the CAM, but outcomes of patients missed with the Pathological findings
CAM-ICU seem to be similar to those of general ICU Minimal to no abnormalities
patients.21 Thus, the most appropriate delirium screen- Subtle neuronal drop-out
ing tool for ICU remains controversial. The use of coma Cerebral ischaemia
Haemorrhages
scales in SAE has not been explored, but the FOUR (Full
Multifocal necrotizing leukoencephalopathy
Outline of UnResponsiveness) scale—a clinical grading Microabscesses
scale used for assessing level of consciousness—is more Abbreviation: SAE, sepsis-associated encephalopathy.
suitable for intubated ICU patients than is the GCS. 22
Nonetheless, the GCS can be useful in predicting the
course of SAE.7 urine culture, and cerebrospinal fluid [CSF] analysis and
No specific biomarker exists for SAE and it remains culture) are required. Evidence of bacteraemia or other
largely a clinical diagnosis. However, a myriad of haematological infection should also be sought.
abnor­mali­t ies are seen in patients with SAE, includ- The absence of bacteraemia does not exclude SAE, and
ing ab­normalities in laboratory tests that do no directly identification of a pathological organism is not always
assess neurological function and, by definition, patients possible. Common reasons for difficulty in identifying
have sepsis. A comprehensive metabolic panel of analy- a disease-causing organism include prior antibiotic use,
ses including a complete blood count, measurement of which might hinder detection of specific organisms, or
expanded electrolytes (sodium, potassium, chloride, presence of an occult abscess. By definition, evidence of
magnesium, phosphate and calcium) and serum enzyme CNS infection in the CSF rules out a diagnosis of SAE.
levels (alanine and aspartate aminotransferases, alkaline However, elevations of CSF protein levels are expected
phosphatase and γ-glutamyltransferase), renal function and are thought to occur as a result of increased perme-
tests (serum creatinine), and investigations for infec- ability of the blood–brain barrier (BBB)10 and movement
tion (blood cultures, respiratory or sputum culture, of proteins from blood into the CSF.

NATURE REVIEWS | NEUROLOGY VOLUME 8 | OCTOBER 2012 | 559

© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS

a b
Fp1–F3 Fp1–A1
F3–C3 F7–A1
C3–P3 T3–A1
P3–O1 T5–A1
Fp2–F4 O1–A1
F3–A1
F4–C4 C3–A1
C4–P4 P3–A1
P4–O2 Fz–A1
Fp1–F7 Cz–A2
F7–T3 Pz–A1
T3–T5 Fp2–A2
T5–O1 F8–A2
Fp2–F8 T4–A2
F8–T4 T6–A2
T4–T6 O2–A2
F4–A2
T6–O2
300 μV C4–A2 300 μV
X3–X4 P4–A2
c 1s 1s
d
Fp1–F7 Fp1–F7
F7–T3 F7–T3
T3–T5 T3–T5
T5–O1 T5–O1
Fp2–F8 Fp1–F3
F8–T4 F3–C3
T4–T6 C3–P3
T6–O2 P3–O1
Fp1–F3
Fp2–F4
F3–C3
F4–C4
C3–P3
P3–O1 C4–P4
Fp2–F4 P4–O2
F4–C4 Fp2–F8
C4–P4 F8–T4
P4–O2 T4–T6
140 μV T6–O2 140 μV
X1–X2
1s 1s

Figure 2 | EEG recordings of patients with encephalopathy. a | EEG recordings with bipolar longitudinal montage from a
patient (56-year-old female) with encephalopathy, demonstrating triphasic waves. b | EEG recordings with a referential
montage (ipsilateral ear reference) from the same patient as in (a), demonstrating triphasic waves. c | EEG recordings (with
bipolar longitudinal montage) from a comatose patient (76-year-old male) with encephalopathy. A mixture of delta and theta
frequency waveforms predominate. d | EEG recording (with bipolar longitudinal montage) from a comatose patient (50-year-
old male) with encephalopathy. Generalized suppression is seen throughout all cerebral regions.

Neuroimaging in patients with SAE gives variable but with laboratory evidence of bacteraemia, EEG reveals
results. Some patients have normal brain MRI scans brain abnormalities in 50% of cases.10 These changes on
despite having SAE. In other patients, acute abnor- EEG are resolved when sepsis is treated.
malities are seen on MRI, including multiple ischaemic Mortality is increased in patients with severe abnor-
strokes or white matter lesions in the centrum semiovale malities on EEG.10 In particular, patients who have tri­
(primarily at the level of Virchow–Robin spaces).23 The phasic wave or burst–suppression patterns on EEG have
severity of these CNS lesions is associated with the sever- higher mortality rates than those with abnormal theta
ity of sepsis and correlates inversely with GCS scores.23 or delta wave patterns.10,25 Notably, these abnormal EEG
A neuro­i maging study in critically ill patients with findings are not only seen in patients with SAE, but are
delirium (not specifically SAE) demonstrated an associa- also frequently recorded in patients with other types of
tion between long duration of delirium and small brain encephalo­pathy 26 and, therefore, do not shed light on the
volume at 3 months after hospital discharge.24 Small aetiology of SAE. Nevertheless, they provide objective
brain volume was also associated with increased cogni- evidence of cerebral disturbance and its severi­ty in SAE.
tive impairment at 12 months after discharge from hos- The clinical manifestations of SAE can be detected
pital.24 Further studies are required to determine whether before presentation of strong evidence of sepsis or SIRS.
small brain volume occurs before or as a consequence of In our experience, changes in cognitive or mental status
delirium during critical illness. that are associated with SAE can present in susceptible
In patients with sepsis, the EEG data show progres- patients up to 36–48 h before other systemic symptoms
sive slowing of brain activity with increasing severity of of sepsis or SIRS become apparent. Such patients, who
SAE.10 Mild encephalopathy is associated with slowing have an otherwise normal neurological examination and
of brain activity in the theta range and severe encephalo­ neuroimaging data and no other aetiology factors associ-
pathy is usually associated with excessive delta waves and, ated with coma, might show unexplained reductions in
less commonly, with a burst–­suppression pattern of activ- level of consciousness. However, functional neurological
ity (Figure 2; Table 1).10 A triphasic wave pattern is seen investigations, such as EEG recordings, can demonstrate
on EEG in approximately 20% of patients with sepsis.10 changes suggestive of sepsis, including diffuse slowing of
Even in patients showing no clear encephalo­pathy (with brain activity and abundant triphasic waves. Thus, we
relatively preserved cognition) on physical examination, propose that SAE is an early feature of systemic infection

560 | OCTOBER 2012 | VOLUME 8  www.nature.com/nrneurol

© 2012 Macmillan Publishers Limited. All rights reserved
 REVIEWS

that presents before the Bone et al.4 criteria for sepsis Table 1 | Changes in EEG recordings in patients with SAE*
are satisfied.
Degree of EEG findings (% of patients)
Overall, SAE remains a diagnosis of exclusion. In encephalopathy
patients presenting with symptoms suggestive of SAE, Normal Theta Delta Triphasic Burst–suppression
waves waves waves pattern
a rigorous investigation for other treatable aetiologies
of encephalopathy, such as systemic organ dysfunction, None 50 38 12 0 0

stroke, intracranial haemorrhage, meningoencephalitis Mild 0 47 54 0 0

or other metabolic disturbance, should be undertaken to Severe 0 10 40 20 30
avoid overlooking the cause of encephalopathy. *Generated from data provided by Young et al.86 Abbreviation: SAE, sepsis-associated encephalopathy.

Pathophysiology
The pathophysiology of SAE has not been established,
Microscopic
but several likely mechanisms have been proposed. These brain
mechanisms are not mutually exclusive and might be injury
involved to varying degrees in different patients, often Inflammatory Blood–brain
cytokines barrier
acting in concert or at various stages of the systemic compromise
illness. This section provides a brief overview of the pro-
cesses that are thought to have a considerable role in the
pathogenesis of SAE (Figure 3). Sepsis-associated
encephalopathy

Microscopic brain injury

Specific neuropathological changes have been described Altered Altered
in patients who died from septic shock. In a prospective cerebral cerebral
metabolism microcirculation
postmortem study of patients with sepsis, cerebral lesions
Altered
were reported which in one patient, were compatible neurotransmission
with multifocal necrotizing leukoencephalo­pathy.27 Fur­
ther prospective data show that cerebral changes associ-
Figure 3 | Aetiology of SAE. The aetiology of SAE is
ated with sepsis include evidence of cere­bral ischaemia, multifactorial. Factors that contribute to SAE pathogenesis
ha­emorrhages and micro-abscesses (Figure 4).28 are shown in circles. The role of each factor varies from
patient to patient and depends on the clinical situation.
BBB function and cerebral microcirculation Abbreviation: SAE, sepsis-associated encephalopathy.
Adequate function of the cerebral microcirculation is
important for maintenance of normal cerebral function. models of sepsis have demonstrated astrocytic dysfunc-
Increasing evidence provides support for altered cerebral tion with detachment of astrocytes from the micro­
microcirculation during sepsis. This alteration probably vasculature.36,37 Taken together, altered blood flow at the
has an important role in SAE pathogenesis. In a sheep microvascular level and BBB dysfunction affect cerebral
model of peritonitis, Taccone et al.29 showed that septic function and probably contribute to the development
shock is associated with reductions in the total perfused and progression of SAE.
cerebral vessel density and in the proportion of per­
fused small vessels in the brain. Similarly, in patients Neurotransmission and cerebral metabolism
with sepsis, disturbances occur in cerebral autoregula- Inflammatory and metabolic changes have been pro-
tion and in response of cerebral blood vessels to carbon posed to be associated with sepsis and to lead to altera-
dioxide concentration.30 Several studies of the effects of tions in cerebral neurotransmission.38,39 Van Gool et al.38
sepsis on carbon-dioxide-induced cerebral vasoreactivity proposed that cytokines such as tumour necrosis factor
have produced conflicting results.31–33 One group found that are produced systemically during sepsis activate
that cerebral vasoreactivity was impaired in patients with cerebral microglia which, in turn, release inflamma-
sepsis,31 whereas others showed that sepsis did not have tory mediators within the CNS.38 These inflammatory
any significant effect on cerebral vascular function.32,33 mediators cause aberrant neuronal function and, thereby,
Rat models of sepsis show increased permeability of delirium and SAE.
the BBB in sepsis, as detected by increased horseradish Cholinergic neurotransmission has a role in this
peroxidase staining of CNS samples taken from septic inflam­matory process and can usually reduce the dura­
rats, compared with controls.34 All the above-mentioned tion of delirium.38 If cholinergic function is impaired—
changes have been documented in patients with SAE, for example, by neurodegenerative disease, pre-existing
but not all patients with SAE have these abnormalities: cholinergic dysfunction or anticholinergic m ­ edications
only some, if any, of these changes might be found in a —a more severe and long-lasting delirium could result.38
given patient. Conversely, a double-blind, placebo-controlled trial
Together with altered cerebral microcirculation, dys- showed that delirium lasts longer and mor­t ality is
function of the BBB alters the extracellular milieu of the increased in patients treated with the cholinesterase
brain. Patients with sepsis have higher levels of protein in inhibitor rivastigmine, compared with the control group,
the CSF than do control patients without sepsis.35 Animal who did not have delirium.40 Given these findings, the

NATURE REVIEWS | NEUROLOGY VOLUME 8 | OCTOBER 2012 | 561

© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS

a b for effects of SAE on long-term cognitive function also

exists, as discussed below.
In sepsis, cerebral metabolism is altered. Experimental
evidence shows that oxidative stress occurs in the early
stages of sepsis, which results in decreased ATP syn­
thesis.46 One study showed an increase in the ratio of
superoxide dismutase to catalase activity in rats during
early stages of sepsis47 and another study showed evi-
dence of mitochondrial dysfunction during sepsis in
c d
mice.48 Some evidence from animal and tissue culture
models shows sepsis-associated effects on the endothe-
lium, which are probably caused by increased superoxide
dismutase activity and reduced expression of endothelial
nitric oxide synthase.49 Coagulation is activated in micro-
vascular beds, leading to tissue ischaemia, and increased
nitric oxide production has been proposed to damage
mitochondria.49 Abundant studies in various animal
models of sepsis support mitochondrial injury caused
by oxidative stress.50,51 Although these studies have not
Figure 4 | Histological changes in the brains of patients with sepsis. a | Coronal focused on the brain, cerebral mitochondria are also
section of the brain from a patient who died of sepsis. This section demonstrates likely to be negatively affected by sepsis, especially in the
evidence of cerebral ischaemia and small areas of haemorrhage. b | Areas of context of increased BBB permeability.
microabscess are seen on histological preparation of the brain from the patient in
In a rat model of sepsis, alterations were also shown
(a), with magnification of ×40. c | The histopathological section in (b) visualized with
a higher magnification (×200). d | Regions of cerebral ischaemia seen on the same in cerebral protein expression. Proteins related to cell
histological preparation as in (a) and (b), with magnification of ×200. Photos courtesy structure, energy production, signalling and cell death
of D. Ramsay, London Health Sciences Centre and University of Western Ontario. (chaperonins, 78 kDa glucose-regulated protein, and glial
fibrillary acidic protein [GFAP]) can be downregulated
hypothesis that disturbance of neurotransmitter function 24 h after onset of sepsis.52 Results of a recent study in
is an aetiological factor for delirium should be more criti- a rat model of sepsis investigating the role of hydrogen
cally assessed in the future. The failure of rivastigmine to sulphide, a novel gaseous transmitter in the CNS, suggest
improve patient outcomes does not rule out the possibil- that increased hydrogen sulphide synthesis occurs
ity that disturbances in cholinergic function contribute in CNS in the setting of sepsis and is associated with
to development of delirium. increased levels of inflammatory cytokines, in­cluding
In studies of sepsis, IL‑1β levels are increased in vitro tumour necrosis factor and IL‑1β.53
and are associated with hippocampal dysfunction in
rats. 41 Examination of CSF proteins in patients with Serum biomarkers
sepsis and delirium suggests that plasma concentrations Several serum biomarkers are elevated in patients with
of essential amino acid precursors (such as tryptophan septic shock. These biomarkers include procalcitonin,
and tyrosine) for synthesis of neurotransmitters, includ- IL‑6, NSE and S100b.54,55 The majority of studies investi-
ing serotonin, dopamine and noradrenaline, are abnor- gating serum biomarkers for sepsis have been performed
mal (considerably higher or lower than normal levels) in patients with a diagnosis of sepsis, but not specifically
during sepsis.42 However, the same study showed that SAE. One study investigated the levels of NSE, S100b and
phenylalanine levels were normal in these patients.42 GFAP in paediatric cases of sepsis with encephalopathy.54
The role of altered levels of serum amino acids in These markers are specific to the CNS and their presence
sepsis and implications for possible neurotransmitter in serum reflects CNS injury.56 Children with sepsis had
imbalance in the brain is complicated, controversial higher levels of serum NSE, S100b and GFAP than did
and still unclear. One study demonstrated an increase controls, and the serum levels of both NSE and S100b
in levels of aromatic amino acids compared with levels were highest in children who did not survive sepsis. 54
of branched-chain amino acids in the plasma of patients Similar findings have been demonstrated in adults, with
with SAE.43 Subsequently, a similar small observational 42% of 170 patients with sepsis showing an increase in
study showed that only phenylalanine levels were S100b serum levels, and 53% of patients showing an
increased in the plasma and CSF of patients with SAE.44 increase in serum NSE levels during the first 72 h after
In addition to abnormal neurotransmission during hospital admission.57 Presence of S100b in serum of the
the acute period of sepsis, long-term effects on neuro- patients reflects glial cell injury and abnormal BBB func-
transmission have been shown. In a rat model of sepsis, tion, whereas presence of NSE, an intraneuronal enzyme,
animals that recovered showed loss of neurons in the reflects neuronal injury. All these markers have also been
hippo­campus and the prefrontal cortex and decreased studied in comatose patients who have survived cardiac
cholinergic innervation of the postrolandic cortical arrest, and in patients with traumatic brain injury.56,58,59
areas.45 These changes result in behavioural effects that Biomarkers are potentially useful for identification of
persist despite resolution of sepsis. Clinical evidence septic illness in general, but they have been considered

562 | OCTOBER 2012 | VOLUME 8  www.nature.com/nrneurol

© 2012 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Table 2 | Interventions for SAE

Intervention Study population Rationale Outcome
Branched-chain amino acid Patients To restore the amino acid balance in Improved serum amino acid
infusions43 with SAE* the blood and the neurotransmitter imbalance and led to recovery
balance in the brain from encephalopathy‡
Glutamate-release inhibitor Rat model To reduce excitotoxicity in the brain Reduced brain oedema, blood–
(riluzole)78 of sepsis brain barrier permeability and
histological damage§
Antioxidant (ascorbate)79 Various animal To reduce microvascular damage Improved capillary blood flow,
models of sepsis caused by free radicals reduced vascular permeability
and improved arteriolar
responsiveness to drugs§
Plasmapheresis or plasma Patients To reduce levels of inflammatory Improved survival in patients
filtration absorption80 with sepsis cytokines in the blood treated with plasmapheresis
Antagonists of cytokine receptors81 To block cytokine receptors No improvement in survival with
cytokine receptor blockers§
Activated protein C82 Patients To reduce clotting in microcirculation Worse outcome and more
with sepsis bleeding in patients§
β2 adrenergic receptor agonists Pig model of SAE To reduce cerebral oedema Reduced perivascular oedema
(dopexamine)83 in the brain§
*Patients were simultaneously treated with antibiotics. ‡Observational study, thus not placebo-controlled. §Placebo-controlled study. Abbreviation: SAE, sepsis-
associated encephalopathy.

to have insufficient specificity and sensitivity to be clini- suspicion for fungal infection as a potential cause of SAE,
cally useful in differentiating SAE from other encephalo­ is required. Once a causative organism has been identi-
pathies.60 Serum NSE and S100b protein levels, however, fied, narrowing the spectrum of anti­biotics is appropri-
are sometimes elevated in patients with SAE.3 The clini- ate, keeping in mind the sensitivities of the organisms to
cal importance of elevated serum levels of NSE and the antibiotics and the local anti­biotic resistance patterns.
S100b in patients with SAE, and the associations of these Symptomatic management of delirium (for example,
markers with the severity of SAE, have been questioned, with antipsychotic therapy) is also necessary.64
given that not all studies have shown consistent results.61 Sedative medication is a common confounder in assess-
A study showed that increased serum levels of IL‑8 ment of neurological status of patients in ICU. Withdrawal
were associated with delirium in patients with inflam- of sedative treatment and judicious use of benzodiazepine
mation, whereas IL‑10 and amyloid‑β (Aβ)1–42 and Aβ1–40 and opiate antagonists (such as flumazenil or naloxone,
levels were elevated in patients with noninflammatory respectively) can be helpful, but often cannot completely
delirium.62 Patients with inflammatory delirium, who resolve delirium. Triphasic waves on EEG are not com-
later showed cognitive impairment, had acutely elevated monly found in drug-induced encephalo­pathy (although
Aβ levels.57 Systemic elevations in serum procalcitonin they might be seen with baclofen, lithium, levodopa and
and IL‑6 levels are also seen in patients with severe pentobarbital intoxication),65 but are features of septic
sepsis.55 Neither procalcitnonin nor IL‑6, however, is a and metabolic (especially hepatic or uraemic) encephalo­
specific biomarker for CNS injury. Taken together, the pathy.26 Somatosensory evoked potentials are usually
above-mentioned studies support the hypothesis that more resistant than EEG patterns to sedative agents or
SAE is, in fact, a result of direct CNS injury. other psychoactive medications. Management of delirium
in patients with SAE requires judicious use of sedative
Treatment and management drugs. In general, lorazepam, a highly potent benzo­
The mainstay of management of SAE hinges on early diazepine, should be avoided. In a placebo-controlled
detection of delirium—which is often the first manifes- trial comparing the effect of dexmedetomidine to that of
tation of sepsis—determination of the underlying cause, lorazepam on clinical outcome in sepsis, patients treated
accurate and prompt treatment of the infection, and pro- with dexmedetomidine had more encephalo­pathy-free
vision of supportive care. Because SAE is not a conse- days, shorter time on the ventilator and lower mortality
quence of direct CNS infection, treatment focus remains than those treated with lorazapam.66
an appropriate management of the systemic infection, Supportive care of patients with SAE should prob-
sepsis and the sequelae of SIRS.63 Early detection of SAE, ably include EEG monitoring to detect nonconvulsive
even before sepsis is diagnosed, is important in prompt- seizures, although the incidence of these seizures in our
ing early investigation for and treatment of infection, prospective monitoring study was low (less than 10%),
thereby avoiding substantial morbidity and mortality with no patients qualifying as having status epilepti-
associated with advanced-stage SAE. cus.67 Thus, prophylactic use of antiepileptic drugs is
In the absence of an identified source or causative not justified.
organism, broad-spectrum antiobiotics are necessary. A Although no definitive therapy exists for SAE, a
prescription of antibiotics covering both gram-negative num­ber of interventions have been tried or suggested in
and gram-positive bacteria, as well as a high degree of humans and animal models of SAE (Table 2). More-direct

NATURE REVIEWS | NEUROLOGY VOLUME 8 | OCTOBER 2012 | 563

© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS

therapies aimed at containing the CNS abnormalities, patients with SAE. Whether the additional presence or
such as microcirculatory changes, altered BBB perme- severity of SAE affects cognition and quality of life is not
ability and abnormal neurotransmission, will hopefully known. In a literature review pertaining to long-term
be developed in the future. These therapies would con- cognitive sequelae of SAE, the authors concluded that
ceivably help to prevent permanent impairment of cogni- further research is necessary to determine the patho-
tion and cerebral function following recovery from sepsis physiology of cognitive impairment following SAE.77
and SAE. Increased knowledge will facilitate the development of
targeted interventions that can be administered while
Prognosis the patient is in the ICU and might limit the long-term
As discussed above, SAE is known to cause both consequences of SAE on daily life.
acute and chronic changes in cognition. The most
easily detected changes occur acutely and manifest as Conclusions
encephalo­pathy and delirium. Mounting evidence in In conclusion, we propose that SAE can be an early
both animal models and in human studies, however, sug- feature of infection in the body before the ‘Bone et al.’
gests that substantial long-term cognitive sequelae are criteria for sepsis are satisfied. Furthermore, as SAE does
associated with SAE. Animal models have shown long- not require evidence of brain pathology, its earliest mani-
term changes in behaviour, learning and memory follow- festations are similar to SIRS-related encephalopathy and
ing SAE.45,68 One study demonstrated altered behaviour are likely to be largely due to inflammatory mediators
and long-term memory deficits in rats in an open-field (cytokines). In cases of severe SAE, prompt initiation of
and radial maze test.45 This study also showed reduced effective therapy at an early stage should avoid serious
cholinergic innervation of the postrolandic cortex, the morbidity and mortality. Late phases of SAE are probably
hippocampus and the prefrontal cortex. 45 The hippo­ heterogeneous in terms of involved mechanisms, and the
campal neuronal loss was postulated to be the basis of the resultant laboratory abnormalities, brain dysfunction
long-term memory deficits. In vitro findings of enhanced and even structural lesions can be diverse, including
γ‑aminobutyric acid-related activity and altered hippo­ ischaemic lesions, encephalopathy due to failure of other
campal neuronal membrane excitability following organs, haemato­logical disorders (such as disseminated
exposure to lipopolysaccharide-induced chronic inflam- intravascular coagulation) with brain purpura, and the
mation in a rat model of SAE provides further evidence associated critical illness neuromyopathy.
for hippocampal synaptic dysfunction in sepsis.41 These SAE is a clinical syndrome that is prevalent in patients
studies suggest that both neuronal loss and alterations in in ICU and carries high mortality in severe cases owing
cholinergic signalling have a role in the long-term effects to its association with multiorgan failure. Moreover,
of SAE. survivors of severe SAE have high morbidity. The brain
We previously discussed the impact of SAE on mortal- is not merely a passive bystander in a reversible septic
ity in the ICU.11 In survivors, it is now well-established illness, but is one of the organs profoundly affected in an
that sepsis can lead to long-term cognitive impair- early and progressive manner. Recognition of delirium
ment 69–74 and lower health-related quality of life.70 In par- in ICU patients as an early feature of SAE is important,
ticular, patients rated the quality of life as being lower in and aggressive investigation and treatment of infec­tion
the domains of physical function, general health, vitality and associated systemic effects, such as hypotension, is
and social function.70 In a follow-up study of critically ill the most effective strategy to reduce mortality and mor-
patients with acute respiratory distress syndrome 6 years bidity associated with the disease. More research and
after discharge from the ICU, 40% of patients had residual clinical trials are needed to develop stratgies to address
disability and 23% of patients had cognitive impairment SAE without inhibiting the immune response and aggra-
(particularly in attention skills).75 All patients with cog- vating systemic infection. Some promising avenues that
nitive impairment also had residual disability, whereas might be considered are use of free radical scavengers,
only a small proportion of cognitively intact patients had restoration of the amino acid balance in the blood, and
residual disability. The patients with cognitive impair- monitoring of and treating microcirculatory abnormali-
ment also had the lowest quality of life scores, especially ties in the brain and other organs. Clearly, there is much
in the domains of physical and social function.75 work to be done.
One study showed that, in addition to cognitive seque-
lae, at 1 year after discharge, 29% of survivors of acute
respiratory distress syndrome had substantial levels of Review criteria
depression and anxiety.72 In this study, the most signifi-
We searched the MEDLINE, PubMed, Google Chrome and
cant correlates of poor quality of life were length of stay Cochrane Library databases using the following search
in ICU and emotional morbidity.72 In another study, it terms: “sepsis-associated encephalopathy”, “septic
was found that survivors of severe sepsis had considerable encephalopathy”, “sepsis AND encephalopathy”, “sepsis
social and emotional morbidity compared with an age- AND delirium”, and “sepsis AND coma”. The search was
matched and sex-matched group who had been admitted restricted to the past 20 years, but was not restricted
to special care units with myocardial infarction.76 to types of articles or language of publication. We also
The above-mentioned data were obtained in survi- searched the reference lists of selected full-text reviews
and original articles to identify further references.
vors of severe illness in the ICU, but not specifically in

564 | OCTOBER 2012 | VOLUME 8  www.nature.com/nrneurol

© 2012 Macmillan Publishers Limited. All rights reserved
 REVIEWS

1. Young, G. B. et al. The encephalopathy patients focusing on relevant clinical outcomes. GABAergic inhibition in hippocampal CA1
associated with septic illness. Clin. Invest. Med. J. Crit. Care 27, 212–217 (2012). pyramidal neurons. Hippocampus 15, 656–664
13, 297–304 (1990). 22. Wijdicks, E. F., Bamlet, W. R., Maramattom, B. V., (2005).
2. Zhang, L. N. et al. Epidemiological features and Manno, E. M. & McClelland, R. L. Validation of a 42. Pandharipande, P. P. et al. Plasma tryptophan
risk factors of sepsis-associated new coma scale: The FOUR score. Ann. Neurol. and tyrosine levels are independent risk factors
encephalopathy in intensive care unit patients: 58, 585–593 (2005). for delirium in critically ill patients. Intensive Care
2008–2011. Chin. Med. J. (Engl.) 125, 828–831 23. Sharshar, T. et al. Brain lesions in septic shock: Med. 35, 1886–1892 (2009).
(2012). a magnetic resonance imaging study. Intensive 43. Freund, H. R., Ryan, J. A. Jr & Fischer, J. E.
3. Iacobone, E. et al. Sepsis-associated Care Med. 33, 798–806 (2007). Amino acid derangements in patients with
encephalopathy and its differential diagnosis. 24. Gunther, M. L. et al. The association between sepsis: treatment with branched chain amino
Crit. Care Med. 37, S331–S336 (2009). brain volumes, delirium duration, and cognitive acid rich infusions. Ann. Surg. 188, 423–430
4. Bone, R. C., Sprung, C. L. & Sibbald, W. J. outcomes in intensive care unit survivors: The (1978).
Definitions for sepsis and organ failure. Crit. VISIONS cohort magnetic resonance imaging 44. Mizock, B. A. et al. Septic encephalopathy.
Care Med. 20, 724–726 (1992). study. Crit. Care Med. 40, 2022–2032 (2012). Evidence for altered phenylalanine metabolism
5. Levy, M. M. et al. 2001 SCCM/ESICM/ACCP/ATS/ 25. Oddo, M., Carrera, E., Claassen, J., Mayer, S. A. & and comparison with hepatic encephalopathy.
SIS International Sepsis Definitions Conference. Hirsch, L. J. Continuous electroencephalography Arch. Intern. Med. 150, 443–449 (1990).
Crit. Care Med. 31, 1250–1256 (2003). in the medical intensive care unit. Crit. Care Med. 45. Semmler, A. et al. Long-term cognitive
6. Bleck, T. P. et al. Neurologic complications of 37, 2051–2056 (2009). impairment, neuronal loss and reduced cortical
critical medical illnesses. Crit. Care Med. 21, 26. Kaplan, P. W. The EEG in metabolic cholinergic innervation after recovery from
98–103 (1993). encephalopathy and coma. J. Clin. Neurophysiol. sepsis in a rodent model. Exp. Neurol. 204,
7. Eidelman, L. A., Putterman, D., Putterman, C. & 21, 307–318 (2004). 733–740 (2007).
Sprung, C. L. The spectrum of septic 27. Sharshar, T. et al. Multifocal necrotizing 46. Azevedo, L. C. Mitochondrial dysfunction during
encephalopathy. Definitions, etiologies, and leukoencephalopathy in septic shock. Crit. Care sepsis. Endocr. Metab. Immune Disord. Drug
mortalities. JAMA 275, 470–473 (1996). Med. 30, 2371–2375 (2002). Targets. 10, 214–223 (2010).
8. Ely, E. W. et al. Delirium as a predictor of 28. Sharshar, T. et al. The neuropathology of septic 47. Barichello, T. et al. Oxidative variables in the rat
mortality in mechanically ventilated patients in shock. Brain Pathol. 14, 21–33 (2004). brain after sepsis induced by cecal ligation and
the intensive care unit. JAMA 291, 1753–1762 29. Taccone, F. S. et al. Cerebral microcirculation is perforation. Crit. Care Med. 34, 886–889
(2004). impaired during sepsis: an experimental study. (2006).
9. Bolton, C. F., Young, G. B. & Zochodne, D. W. Crit. Care 14, R140 (2010). 48. d’Avila, J. C. et al. Sepsis induces brain
The neurological complications of sepsis. Ann. 30. Taccone, F. S. et al. Cerebral autoregulation is mitochondrial dysfunction. Crit. Care Med. 36,
Neurol. 33, 94–100 (1993). influenced by carbon dioxide levels in patients 1925–1932 (2008).
10. Young, G. B., Bolton, C. F., Archibald, Y. M., with septic shock. Neurocrit. Care 12, 35–42 49. Cepinskas, G. & Wilson, J. X. Inflammatory
Austin, T. W. & Wells, G. A. The (2010). response in microvascular endothelium in
electroencephalogram in sepsis-associated 31. Terborg, C. et al. Dysfunction of vasomotor sepsis: role of oxidants. J. Clin. Biochem. Nutr.
encephalopathy. J. Clin. Neurophysiol. 9, reactivity in severe sepsis and septic shock. 42, 175–184 (2008).
145–152 (1992). Intensive Care Med. 27, 1231–1234 (2001). 50. Brealey, D. et al. Mitochondrial dysfunction in
11. Sprung, C. L. et al. Impact of encephalopathy on 32. Matta, B. F. & Stow, P. J. Sepsis-induced along-term rodent model of sepsis and organ
mortality in the sepsis syndrome. The Veterans vasoparalysis does not involve the cerebral failure. Am. J. Physiol. Regul. Integr. Comp.
Administration Systemic Sepsis Cooperative vasculature: indirect evidence from Physiol. 286, R491–R497 (2004).
Study Group. Crit. Care Med. 18, 801–806 autoregulation and carbon dioxide reactivity 51. Crouser, E. D., Julian, M. W., Blaho, D. V. &
(1990). studies. Br. J. Anaesth. 76, 790–794 (1996). Pfeiffer, D. R. Endotoxin-induced mitochondrial
12. Salluh, J. I. et al. Delirium epidemiology in critical 33. Fulesdi, B. et al. Cerebral vasoreactivity to damage correlates with impaired respiratory
care (DECCA): an international study. Crit. Care acetazolamide is not impaired in patients with activity. Crit. Care Med. 30, 276–284 (2002).
14, R210 (2010). severe sepsis. J. Crit. Care 27, 337–343 (2012). 52. Hinkelbein, J. et al. Alterations in cerebral
13. Wilson, J. X. & Young, G. B. Progress in clinical 34. du Moulin, G. C., Paterson, D., Hedley-Whyte, J. metabolomics and proteomic expression during
neurosciences: sepsis-associated & Broitman, S. A. E. coli peritonitis and sepsis. Curr. Neurovasc. Res. 4, 280–288
encephalopathy: evolving concepts. Can. J. bacteremia cause increased blood–brain (2007).
Neurol. Sci. 30, 98–105 (2003). barrier permeability. Brain Res. 340, 261–268 53. Chen, D. et al. Effects of hydrogen sulfide on a
14. MacKenzie, I. M. The haemodynamics of human (1985). rat model of sepsis-associated encephalopathy.
septic shock. Anaesthesia 56, 130–144 (2001). 35. Davies, D. C. Blood–brain barrier breakdown in J. Huazhong Univ. Sci. Technolog. Med. Sci. 31,
15. Douglas, V. C. & Josephson, S. A. Altered mental septic encephalopathy and brain tumours. 632–636 (2011).
status. Continuum (Minneap. Minn.) 17, J. Anat. 200, 639–646 (2002). 54. Hsu, A. A. et al. Neurological injury markers in
967–983 (2011). 36. Papadopoulos, M. C. et al. Faecal peritonitis children with septic shock. Pediatr. Crit. Care
16. Miu, D. K. & Yeung, J. C. Incidence of post- causes oedema and neuronal injury in pig Med. 9, 245–251 (2008).
stroke delirium and 1‑year outcome. Geriatr. cerebral cortex. Clin. Sci. (Lond.) 96, 461–466 55. Fioretto, J. R. et al. Interleukin‑6 and
Gerontol. Int. http://dx.doi.org/10.1111/ (1999). procalcitonin in children with sepsis and septic
j.1447-0594.2012.00871.x. 37. Nishioku, T. et al. Detachment of brain pericytes shock. Cytokine 43, 160–164 (2008).
17. Khan, B. A. et al. Delirium in hospitalized from the basal lamina is involved in disruption of 56. Berger, R. P. et al. Serum neuron-specific
patients: implications of current evidence on the blood–brain barrier caused by enolase, S100B, and myelin basic protein
clinical practice and future avenues for research lipopolysaccharide-induced sepsis in mice. Cell. concentrations after inflicted and noninflicted
—a systemic evidence review. J. Hosp. Med. Mol. Neurobiol. 29, 309–316 (2009). traumatic brain injury in children. J. Neurosurg.
http://dx.doi.org/10.1002/jhm.1949. 38. van Gool, W. A., van de Beek, D. & 103, 61–68 (2005).
18. Inouye, S. K. et al. Clarifying confusion: the Eikelenboom, P. Systemic infection and delirium: 57. Nguyen, D. N. et al. Elevated serum levels of
confusion assessment method. A new method when cytokines and acetylcholine collide. Lancet S‑100beta protein and neuron-specific enolase
for detection of delirium. Ann. Intern. Med. 113, 375, 773–775 (2010). are associated with brain injury in patients with
941–948 (1990). 39. Jacob, A., Brorson, J. R. & Alexander, J. J. Septic severe sepsis and septic shock. Crit. Care Med.
19. Reade, M. C., Eastwood, G. M., Peck, L., encephalopathy: inflammation in man and 34, 1967–1974 (2006).
Bellomo, R. & Baldwin, I. Routine use of the mouse. Neurochem. Int. 58, 472–476 (2011). 58. Lamers, K. J. et al. Protein S‑100B, neuron-
Confusion Assessment Method for the Intensive 40. van Eijk, M. M. et al. Effect of rivastigmine as an specific enolase (NSE), myelin basic protein
Care Unit (CAM-ICU) by bedside nurses may adjunct to usual care with haloperidol on (MBP) and glial fibrillary acidic protein (GFAP) in
underdiagnose delirium. Crit. Care Resusc. 13, duration of delirium and mortality in critically ill cerebrospinal fluid (CSF) and blood of
217–224 (2011). patients: a multicentre, double-blind, placebo- neurological patients. Brain Res. Bull. 61,
20. van Eijk, M. M. et al. Routine use of the controlled randomised trial. Lancet 376, 261–264 (2003).
confusion assessment method for the intensive 1829–1837 (2010). 59. Wijdicks, E. F. et al. Practice parameter:
care unit: a multicenter study. Am. J. Respir. Crit. 41. Hellstrom, I. C., Danik, M., Luheshi, G. N. & prediction of outcome in comatose survivors
Care Med. 184, 340–344 (2011). Williams, S. Chronic LPS exposure produces after cardiopulmonary resuscitation (an
21. Tomasi, C. D. et al. Comparison of CAM-ICU and changes in intrinsic membrane properties and a evidence-based review): report of the Quality
ICDSC for the detection of delirium in critically ill sustained IL‑beta‑dependent increase in Standards Subcommittee of the American

NATURE REVIEWS | NEUROLOGY VOLUME 8 | OCTOBER 2012 | 565

© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS

Academy of Neurology. Neurology 67, 203–210 69. Soriano, F. G. Quality of life: late sequela in 78. Toklu, H. Z. et al. The effects of riluzole on
(2006). sepsis. Crit. Care Med. 33, 262–263 (2005). neurological, brain biochemical, and histological
60. Pierrakos, C. & Vincent, J. L. Sepsis biomarkers: 70. Heyland, D. K. & Kutsogiannis, D. J. Quality of changes in early and late term of sepsis in rats.
a review. Crit. Care 14, R15 (2010). life following critical care: moving beyond J. Surg. Res. 152, 238–248 (2009).
61. Piazza, O., Russo, E., Cotena, S., Esposito, G. & survival. Intensive Care Med. 26, 1172–1175 79. Wilson, J. X. Mechanism of action of vitamin C in
Tufano, R. Elevated S100B levels do not (2000). sepsis: ascorbate modulates redox signaling in
correlate with the severity of encephalopathy 71. Gordon, S. M., Jackson, J. C., Ely, E. W., endothelium. Biofactors 35, 5–13 (2009).
during sepsis. Br. J. Anaesth. 99, 518–521 Burger, C. & Hopkins, R. O. Clinical identification 80. Busund, R., Koukline, V., Utrobin, U. &
(2007). of cognitive impairment in ICU survivors: Nedashkovsky, E. Plasmapheresis in severe
62. van den Boogaard, M. et al. Biomarkers insights for intensivists. Intensive Care Med. 30, sepsis and septic shock: a prospective,
associated with delirium in critically ill patients 1997–2008 (2004). randomised, controlled trial. Intensive Care Med.
and their relation with long-term subjective 72. Hopkins, R. O., Weaver, L. K., Chan, K. J. & 28, 1434–1439 (2002).
cognitive dysfunction; indications for different Orme, J. F. Jr. Quality of life, emotional, and 81. Goldie, A. S. et al. Natural cytokine antagonists
pathways governing delirium in inflamed and cognitive function following acute respiratory and endogenous antiendotoxin core antibodies
noninflamed patients. Crit. Care 15, R297 (2011). distress syndrome. J. Int. Neuropsychol. Soc. 10, in sepsis syndrome. The Sepsis Intervention
63. Dellinger, R. P. et al. Surviving Sepsis Campaign: 1005–1017 (2004). Group. JAMA 274, 172–177 (1995).
international guidelines for management of 73. Hopkins, R. O. et al. Neuropsychological 82. Marti-Carvajal, A. J., Sola, I., Lathyris, D. &
severe sepsis and septic shock: 2008. Intensive sequelae and impaired health status in Cardona, A. F. Human recombinant activated
Care Med. 34, 17–60 (2008). survivors of severe acute respiratory distress protein C for severe sepsis. Cochrane Database
64. Markowitz, J. D. & Narasimhan, M. Delirium and syndrome. Am. J. Respir. Crit. Care Med. 160, of Systematic Reviews, Issue 3. Art. No.:
antipsychotics: a systematic review of 50–56 (1999). CD004388. http://dx.doi.org/10.1002/
epidemiology and somatic treatment options. 74. Hough, C. L. & Curtis, J. R. Long-term sequelae 14651858.CD004388.pub5.
Psychiatry (Edgmont) 5, 29–36 (2008). of critical illness: memories and health-related 83. Moss, R. F., Parmar, N. K., Tighe, D. &
65. Blume, W. T. Drug effects on EEG. J. Clin. quality of life. Crit. Care 9, 145–146 (2005). Davies, D. C. Adrenergic agents modify cerebral
Neurophysiol. 23, 306–311 (2006). 75. Rothenhausler, H. B., Ehrentraut, S., Stoll, C., edema and microvessel ultrastructure in porcine
66. Pandharipande, P. P. et al. Effect of Schelling, G. & Kapfhammer, H. P. The sepsis. Crit. Care Med. 32, 1916–1921 (2004).
dexmedetomidine versus lorazepam on outcome relationship between cognitive performance and
in patients with sepsis: an a priori-designed employment and health status in long-term Acknowledgements
analysis of the MENDS randomized controlled survivors of the acute respiratory distress The authors thank D. Ramsay of the University of
trial. Crit. Care 14, R38 (2010). syndrome: results of an exploratory study. Gen. Western Ontario for providing the neuropathological
67. Young, G. B. & Doig, G. S. Continuous EEG Hosp. Psychiatry 23, 90–96 (2001). images used in Figure 4.
monitoring in comatose intensive care patients: 76. Lazosky, A., Young, G. B., Zirul, S. & Phillips, R.
epileptiform activity in etiologically distinct Quality of life after septic illness. J. Crit. Care 25, Author contributions
groups. Neurocrit. Care 2, 5–10 (2005). 406–412 (2010). Both authors made substantial contributions to
68. Barichello, T. et al. Long-term cognitive 77. Comim, C. M. et al. Cognitive impairment in the researching data for article, discussion of content,
impairment in sepsis survivors. Crit. Care Med. septic brain. Curr. Neurovasc. Res. 6, 194–203 writing, and reviewing and/or editing of the
33, 1671 (2005). (2009). manuscript before submission.

566 | OCTOBER 2012 | VOLUME 8  www.nature.com/nrneurol

© 2012 Macmillan Publishers Limited. All rights reserved