Download Evidence-Based Review of Moderate to Severe Acquired Brain Injury

Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16. Acute Interventions for Acquired Brain Injury
Matthew J Meyer PhD (Candidate), Robert Teasell MD FRCPC, Joseph Megyesi
MD PhD FRCSC, Nestor Bayona MSc
ERABI
Parkwood Hospital
801 Commissioners Rd E, London ON
519-685-4292 x42630
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Table of Contents
16.1 Management of Intracranial Pressure (ICP) ............................... 10
16.1.1 Non-Pharmacological Treatments ................................................................ 11
16.1.1.1 Head Posture ............................................................................................... 11
16.2.1.2 Hypothermia ................................................................................................ 16
16.1.1.3 Hyperventilation .......................................................................................... 32
16.1.1.4 Cerebrospinal Fluid Drainage ...................................................................... 36
16.1.1.5 Decompressive Craniotomy ........................................................................ 41
16.1.1.6 Continuous Rotational Therapy and Prone Positioning .............................. 50
16.1.2 Pharmacological Treatments ........................................................................ 52
16.1.2.1 Osmolar Therapies....................................................................................... 52
16.1.2.1.1 Hypertonic Saline .......................................................................................................... 52
16.1.2.1.2 Mannitol........................................................................................................................ 61
16.1.2.2 Propofol ....................................................................................................... 66
16.1.2.3 Midazolam ................................................................................................... 69
16.1.2.4 Opioids ......................................................................................................... 71
16.2.2.5 Barbiturates ................................................................................................. 74
16.1.2.6 Cannabinoids ............................................................................................... 81
16.2.2.7 Corticosteroids ............................................................................................ 83
16.1.2.8 Progesterone ............................................................................................... 87
16.1.2.9 Bradykinin Antagonists ................................................................................ 89
16.1.2.10 Dimethyl Sulfoxide..................................................................................... 92
16.2 Prompting Emergence from Coma ............................................ 95
16.2.1 Non-Pharmacological ................................................................................... 95
16.2.1.1 Sensory Stimulation ..................................................................................... 95
16.2.1.2 Music Therapy ........................................................................................... 102
16.2.1.3 Electrical Stimulation ................................................................................. 103
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.2.2 Pharmacological Interventions ................................................................... 105
16.2.2.1 Amantadine ............................................................................................... 105
16.4 Summary ................................................................................ 111
16.5 References ............................................................................. 115
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Table Directory
Table 16.1
Head Posture for the Management of Elevated Intracranial Pressure post-ABI
Table 16.2
Hypothermia for the Acute Management of Elevated ICP Post ABI
Table 16.2a
Summary of RCT Studies of Acute Hypothermia Post ABI
Table 16.2b
Summary of non-RCT Studies of Acute Hypothermia Post ABI
Table 16.3
Hyperventilation for the Treatment of Elevated ICP Post ABI
Table 16.4
Cerebrospinal Drainage for Treatment of Elevated ICP Post ABI
Table 16.5
Decompressive Craniectomy to Control Refractory Elevated ICP Post ABI
Table 16.6
Continuous Rotational Therapy and Prone Positioning in Acute Care
Management Post ABI
Table 16.7
Hypertonic Saline for the Management of ICP Hypertension Post ABI
Table 16.8
Mannitol for the Management of ICP and Hypertension Post ABI
Table 16.9
Propofol for the Management of Acute ABI
Table 16.10
Midazolam for the Management Acute ABI
Table 16.11
Opioids for the Management Acute ABI
Table 16.12
Barbiturates for the Management of Elevated Intracranial Pressure Post ABI
Table 16.13
Cannabinoids as an Acute Therapeutic Strategy Post ABI
Table 16.14
Corticosteroids for the Management of Elevated Intracranial Pressure and
Neuro-protection Post ABI
Table 16.15
Progesterone for Treatment of Acute ABI
Table 16.16
Bradykinin Antagonist as an Acute Therapeutic Strategy Post ABI
Table 16.17
DMSO as an Acute Therapeutic Strategy Post ABI
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Table 16.18
Sensory Stimulation for the Management of Patients in a Coma or Vegetative
State Post ABI
Table 16.18a Summary of Studies of Sensory Stimulation to Promote Emergence from Coma
or Vegetative State Post ABI
Table 16.19
Music and Musicokinetic Therapy for Patients with Coma or Vegetative State
Post ABI
Table 16.20
Electrical Stimulation Post ABI
Table 16.21
Amantadine for Arousal from Post ABI Coma
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Key Points
Study findings suggest a 30° head elevation reduces ICP and improves CPP post TBI.
Elevating the head of the bed post TBI is effective at lowering intracranial pressure in
children, although the impact on CPP was minimal.
Although hypothermia has been shown to reduce elevated ICP by some researchers there is
no solid evidence to support its effectiveness post ABI. More research needs to be done.
Systemic hypothermia increases the risk of pneumonia post ABI.
Tromethamine counteracts the detrimental effects of prolonged hyperventilation for the
control of ICP leading to better outcomes post-ABI.
Hyperoxia may counteract the adverse effects of prolonged hyperventiliation for the control
of ICP post-ABI.
Hyperventilation below 34 torr PaCO2 may cause an increase in hypoperfused brain tissue.
CSF drainage has been found to reduce ICP and increase CPP in those who have sustained an
ABI.
In adults standard trauma craniectomy leads to better control of ICP and better clinical
outcomes at 6 months when compared with limited craniectomy
Resection of a larger bone flap during craniectomy may lead to a greater reduction in ICP,
better patient outcomes and fewer post-surgical complications
Although decompressive cranectomy does reduce ICP in children more research needs to be
conducted investigating its impat on the long term clinical outcomes.
Continuous rotational therapy may not worsen intracranial pressure in severe brain injury
patients
Prone position may increase oxygenation and cerebral perfusion pressure in patients with
acute respiratory insufficiency.
Hypertonic saline reduces ICP more effectively than mannitol.
Hypertonic saline and Ringer’s lactate solution are similar in lowering elevated ICP and result
in similar clinical outcomes and survival up to 6 months post-injury.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
In children, use of hypertonic saline in the ICU setting results in a lower frequency of early
complications and shorter ICU stays compared with Ringer’s lactate in children.
Saline results in decreased mortality rates compared to albumin.
Hypertonic saline may reduce elevated ICP uncontrolled by conventional ICP management
measures.
Hypertonic saline may aid in resuscitation of brain injured patients by increasing cerebral
oxygenation.
Sodium lactate is more effective than mannitol for reducing acute elevations in ICP.
High dose mannitol results in lower mortality rates and better clinical outcomes compared
with conventional mannitol.
Early out of hospital administration of mannitol does not negatively affect blood pressure.
Mannitol may only lower ICP when initial ICP values are abnormally elevated.
Propofol may help to reduce ICP and the need for other ICP and sedative interventions when
used in conjunction with morphine.
Infusions of propofol greater than 4mg/kg per hour should be undertaken with extreme
caution.
Midazolam has no effect on ICP but may result in systemic hypotension.
Bolus opioid administration results in increased ICP.
There is conflicting evidence regarding the effects of opioid infusion on ICP.
Remifentanil results in faster arousal compared to hypnotic based sedation.
There are conflicting reports regarding the efficacy of pentorbarbital for the control of
elevated ICP.
Thiopental is beter than pentobarbital for controlling unmanageable refractory ICP.
Pentobarbital is not better than mannitol for the control of elevated ICP.
Barbiturate therapy plus hypothermia may improve clinical outcomes.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Patients undergoing barbiturate therapy should have their immunological response and
systemic blood pressure monitored.
Dexanabinol is not effective in controlling ICP or in improving clinical outcomes post-ABI.
Methylprednisolone increases mortality rates in ABI patients and should not be used.
Triamcinolone may improve outcomes in patients with a GCS<8 and a focal lesion.
Dexamethasone does not improve ICP levels and may worsen outcomes in patients with ICP >
20mmHg
Glucocorticoid administration may increase the risk of developing first late seizures.
Progesterone decreases 30-day mortality rates.
Progesterone improves GOS and modified FIM scores at 3 and 6 months post-injury.
Some bradykinin antagonists prevent acute elevations in ICP but their effects on long-term
clinical outcomes are uncertain.
Dimethyl sulfoxide may cause temporary reductions in ICP elevations post-ABI.
Sensory stimulation provided by family members improves consciousness for patients with
GCS 6-8.
Sensory stimulation may help to promote emergence from coma or vegetative state post ABI.
Music therapy might be useful in promoting emergence from coma post ABI.
Median nerve electrical stimulation does not improve emergence from coma post-ABI.
Amantadine may improve consciousness and cognitive function in comatose ABI patients.
Dopamine enhancing drugs may facilitate rate of recovery post traumatic brain injury in
children; however, due to the small sample sizes more definitive research is needed.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
List of Common Abbreviations
AANS
ABI
BTF
CBF
CBV
CMRO2
CPP
DRS
EBIC
GCS
GOS
ICP
mmHg
PEDro
RLA
TBI
American Association of Neurological Surgeons
Acquired Brain Injury
Brain Trauma Foundation
Cerebral Blood Flow
Cerebral Blood Volume
Cerebral Metabolic Rate for Oxygen consumption
Cerebral Perfusion Pressure
Disability Rating Scale
European Brain Injury Consortium
Glasgow Coma Scale
Glasgow Outcome Scale
Intracranial Pressure
mm of mercury
Physiotherapy Evidence Database rating scale
Rancho Los Amigos scale
Traumatic Brain Injury
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16. Acute Interventions for Acquired Brain Injury
During the initial stages of a brain injury, there are variable degrees of irreversible damage to
the central nervous system commonly known as the primary injury. Subsequently, a chain of
events is put into motion leading to ongoing injury to the brain caused by edema, hypoxia, and
ischemia which occurs as a result of increased ICP, the release of toxic amounts of excitatory
neurotransmitters like glutamate, and impaired ionic homeostasis (2007). Acute brain injury
treatment therefore focuses on preventing or minimizing the extent of secondary injury by
targeting intracranial hypertension, oxygenation and ion homeostasis in order to reduce cellular
injury.
16.1 Management of Intracranial Pressure (ICP)
High intracranial pressure (ICP) is one of the most frequent causes of death and disability
following severe head injury. It is defined as an ICP reading greater than 20 mmHg within any
intracranial space (subdural, intraventricular, extradural, or intraparenchymal compartments)
(Sahuqillo & Arikan, 2006). Mortality and morbidity following severe brain injury are strongly
associated with increased ICP. Since the consequences of primary brain injury cannot be
reversed, post-injury management primarily focuses on prevention and reversal of secondary
insults to improve outcomes. Following a brain injury, the brain is extremely vulnerable to
secondary ischemia due to systemic hypotension or diminished cerebral perfusion resulting
from elevations in intracranial pressure (Doyle et al., 2001). For these reasons, the acute care of
TBI patients includes the maintenance of adequate blood pressure and management of
anticipated rises in ICP.
Elevated ICP after ABI is generally due to edema or inflammation within the cranial cavity.
There are different physiological mechanisms responsible for the production of this excess fluid
resulting in vasogenic, cytotoxic and interstitial edema (Rabinstein, 2006). Vasogenic edema
results from disruption of the blood brain barrier causing increased permeability and release of
fluid into the extravascular space. Cytotoxic edema is due to failure of cellular ionic pumps
causing increases in intracellular water content. Finally, interstitial edema is the forced flow of
fluid from intraventricular compartments to the parenchyma generally due to an obstruction in
drainage.
Control of ICP is extremely important in patients with traumatic brain injuries (TBI), and
multiple therapies tend to be used to manage ICP. To be effective, treatments need to target
the specific form of edema that is problematic. The degree and timing of ICP elevation are also
important determinants of clinical outcome, so it is important for ICP interventions to act
rapidly to be effective.
Non-surgical therapy includes the use of osmotic and loop diuretics, hypothermia, sedation and
paralysis, controlled hyperventilation and barbiturates. Surgical therapies include
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
ventriculostomy with therapeutic drainage, evacuation of mass lesions, as well as
decompressive craniectomy.
Interest in the promise of potential neuroprotective agents has also begun to spark new
research initiatives. The negative effects associated with cellular level post-traumatic stress
may be a potential target for future therapies. Traditional therapies have included sedatives
such as barbiturates and opiates in an attempt to down regulate cellular metabolism. Newer
initiatives have begun to target free radical production and oxidative stresses, which affect
membrane viability. Some neuroprotective agents that have been suggested include sedatives,
steroids and anti-oxidant solutions.
Guideline Recommendations
In an attempt to standardize acute management of acquired brain injury, several consensus
guidelines of care have been developed. The two most prominent sets of guidelines are those
developed by the American Association of Neurological Surgeons initially in 1995 and most
recently in 2007 (Carney & Ghajar, 2007), and by the European Brain Injury Consortium (EBIC)
in 1997 (Maas et al., 1997). These guidelines, which have gained credibility worldwide, are
widely recognized as influencing clinical practice. With this in mind, we have chosen to add
recommendations made by either organization into our evaluation of each intervention.
However, the conclusions presented here are based on our methodology and have not been
influenced by guideline recommendations.
Although, the EBIC provides descriptive guidelines, they do not incorporate levels of evidence
(Maas et al., 1997). The AANS make recommendations based on levels of evidence as follows
(Carney & Ghajar, 2007):
Level I - Good quality Randomized Control Trial (RCT)
Level II - Moderate quality RCT, good quality cohort, good quality case-control
Level III - Poor quality RCT, moderate or poor quality cohort, moderate or poor quality casecontrol, case-series, databases or registries
16.1.1 Non-Pharmacological Treatments
16.1.1.1 Head Posture
The standard practice in most head injury intensive care units is to elevate the head above the
level of the heart in an effort to reduce intracranial pressure by facilitating venous outflow
without compromising cerebral perfusion pressure (CPP) and cardiac output (Ng et al., 2004). It
has been suggested that head elevation may even slightly improve CPP (Schulz-Stiibner and
Thiex 2006). Placing patients in an elevated head posture also facilitates early provision of
enteral nutrition while at the same time reducing the risk for gastric reflux and pulmonary
aspiration when compared with patients kept in the supine position (Ng et al., 2004). Ng et al.
(2004) note that nursing individuals who sustain a TBI remaining in a flat position reduces the
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
risk for systemic hypotension inherent in a semi-recumbent posture. Furthermore, some
authors argue that a horizontal body position increases cerebral perfusion pressure improving
cerebral blood flow (Winkelman 2000).
The European Brain Injury Consortium stated that no consensus existed regarding the benefits
of head elevation to 30 degrees when compared to the recumbent position (Maas et al., 1997).
The AANS has not referred to head position in their most recent guidelines (Carney and Ghajar
2007).
Individual Studies
Table 16.1 Head Posture for the Management of Elevated Intracranial Pressure post-ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Adult Population
Winkelman
(2000)
RCT
PEDro = 4
N=8 Non-vascular, non-penetrating severe
brain injury (GCS < 8) patients were randomly
assigned to be placed either with the backrest
elevated at 30 degrees or flat (0 degree
elevation). Patients remained in the initial
position for 65-75 min after which they were
placed in the alternative position. ICP and CPP
were assessed at 5, 15, 30 and 60 min after the
change in position.
No other interventions occurred during the 60
min observation period. Use of backrest
elevation of 30 degrees resulted in significant
improvements in both ICP and CPP
immediately after changes in backrest position
from 0 to 30 degrees (p=0.001) and during
equilibrium at the 30 degree position
(p=0.003).
March et al.,
(1990)
USA
RCT
PEDro = 3
N=4 Head injury patients started laying flat and No significant changes noted on any of the
underwent backrest manipulation into 3
measures among the 4 backrest positions.
different positions: 30 degree head elevation,
30 degree head elevation with knee gatch
raised, and flat to reverse Trendelenburg
position. Subjects were initially placed in the
flat position for 15 minutes, followed by 15
minutes in one of the 3 randomly assigned
alternate backrest positions. All subjects were
assessed under all 4 backrest positions for
changes in ICP, CPP and CBF.
Ledwith et al.,
(2010)
USA
Quasi-RCT
N=30 All participants had sustained either a TBI
or SAH. Each participant was placed in one of
12 body positions, which were randomly
ordered. Positions included, the supine, supine
with knee bent, left lateral position and right
lateral position. In each position the HOP was
elevated to 15, 30 or 45 degrees. Individuals
remained in each position for 2 hours.
Thirty of the 33 participants completed the
study. Results indicate that ICP was
significantly reduced when individuals were
placed in the supine with HOB 45° (p=0.002 –
decrease), left lateral with HOB 15° (p=0.026 –
increase), right lateral with HOB 15° (p=0.002 increase) and the knee elevation with HOB 30°
(p=0.039 - decrease). Oxygenation of brain
tissue also increased while in 4 of the
positions: supine with HOB 30° (p= 0.006 -
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
decrease), supine with HOB 45° (p=0.004 –
decrease), left lateral with HOB 30° (p=0.046, decrease and right lateral with HOB 30°
(p=0.028 - decrease).
Schulz-Stiibner
& Thiex
(2006)
Case Series
N=10 Study participants had sustained either a
severe TBI (n=5) and SAH (n=5). Participants
had their head elevated to 30°. Positive end
expiratory pressure (PEEP) of 5cmH20. PEEP was
increased to 10 and 15cmH20.
When participants had their heads lowered to
a flat position, ICP was increased and CPP was
reduced. Increase in PEEP from 5 to 10 cmH 20
increased ICP without dropping CPP
significantly. When the head was elevated to
30° PEEP was found to increase from 5 to 10
cmH20.
Ng et al.,
(2004)
Singapore
Case Series
N=38 The effect of head elevation on ICP, CPP,
mean arterial pressure, global cerebral
oxygenation, and regional cerebral oxygenation
were compared in a group of closed head injury
patients (GCS ≤ 8) when the head was elevated
30 degrees or when the head was lowered to 0
degrees. Patients acted as their own controls.
ICP was significantly lower at 30 degrees than
at 0 degrees of head elevation (p=0.0005).
Mean arterial pressure remained unchanged.
CPP was slightly but not significantly higher at
30 degrees (p=0.412). Of note those who had
lower ICP at the start of the study were found
to have the greatest decrease in ICP when the
head was elevated 30°. Global cerebral
oxygenation and regional cerebral
oxygenation were not affected significantly by
head elevation.
Meixensberger
et al.,
(1997)
Germany
Case Series
N=22 Acute brain injury patients were initially
placed in a 30 degree body position and
baseline values were collected for 10 min, then
body position was changed to 0 degrees and
values were noted after reaching steady state
conditions (10-15 min). ICP, tissue pO2 (ti-pO2),
and CPP were assessed in both head positions.
Compared with the 30 degree head position
ICP was significantly higher (p<0.001) and CPP
significantly lower (p<0.01) at the 0 degrees
head position. ti-pO2 and mean arterial blood
pressure were unaffected by head position.
Feldman et al.,
(1992)
USA
Case Series
N=22 Patients were treated within 72 hours
after injury. In the first 13 patients the head
was initially elevated to 30 degrees. In the
subsequent 9 patients the head was initially set
at 0 degrees of elevation. Head elevation was
changed to the alternate position after 45
minutes. Cerebral blood flow (CBF), mean
carotid pressure (MCP), ICP, CMRO2, oxygen
saturation in the jugular bulb, cerebrovascular
resistance (CVR), PaCO2, PaO2, arteriovenous
difference of lactate, mean arterial blood
pressure (MABP) and cerebral perfusion
pressure (CPP) were compared in both head
positions.
MCP and ICP were significantly lower at 30
degrees of head elevation than at 0 degrees of
elevation (p=0.085 and p=0.0079
respectively). Furthermore, patients with the
highest ICP at the horizontal position
experienced the greatest reductions in ICP at
30 degree of head elevation (r= -05890). All of
the other physiological parameters were not
significantly affected by the change in head
elevation.
Lee
N=30 Changes in ICP were measured in severe
Compared with the supine 0º position, ICP
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
(1989)
China
Case Series
head injury patients (GCS < 8) with patients in
four different positions: head at 0º; supine with
head of the bed down 30º; three fourths
supine; and three fourths prone without
turning the head.
increased significantly for head down 30º
(p<0.01), three fourths supine (p<0.01) and
three fourths prone (p<0.01) positions,
although there were individual differences in
ICP in response to position changes.
Durward et al.,
(1983)
Canada
Case Series
N=11 Severe brain injury patients (GCS < 8) with
associated intracranial hypertension were
subjected to 4 positions of head elevation: 0
(baseline), 15, 30 and 60º. Each position was
maintained for 5-10 minutes and ICP and CPP)
were assessed.
ICP decreased significantly with elevation of
the head from 0º to 15º (p<0.001). This
decrease in ICP was maintained at 30º
(p<0.001) but was not significantly different
from the ICP at 15º. ICP at head elevation of
60º was not significantly different from 0º of
elevation. Furthermore, CPP was not
significantly affected by 15º or 30º of head
elevation, whereas elevation of the head to
60º caused a significant reduction of CPP
compared with baseline (0º, p< 0.02).
Pediatric Population
Agbeko et al.
(2012)
United Kingdom
RCT
PEDro = 5
N=38 The effect of head elevation on ICP, CPP,
mean arterial pressure, global cerebral
oxygenation, and regional cerebral oxygenation
were compared in a group of closed head injury
patients (GCS ≤ 8) when the head was elevated
30 degrees or when the head was lowered to 0
degrees. Patients acted as their own controls.
ICP was significantly lower at 30° than at 0
degrees of head elevation (p=0.0005). Mean
arterial pressure remained unchanged. CPP
was slightly but not significantly higher at 30
degrees (p=0.412). Of note those who had
lower ICP at the start of the study were found
to have the greatest decrease in ICP when the
head was elevated 30°. Global cerebral
oxygenation and regional cerebral
oxygenation were not affected significantly by
head elevation.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In an earlier RCT Winkelman (2000) reported that head elevation of 30 degrees resulted in
significant improvements in both ICP and cerebral perfusion pressure when compared to a flat
position. This finding was also supported by several other studies (Ng et al., 2004; Lee et al.,
1989; Durward et al.,1983; Feldman et al., 1992). In each of these elevating the head to 30°
significantly decreased ICP, while increasing CPP. Again, in a 2006 study, Schulz-Stiibner noted
that ICP decreased and CPP increased when participants’ heads were elevated to 30°. The
findings of Durward et al. (1983) also suggest that a head elevation of 15 to 30° may yield the
best result. Further they found an elevation of 60° did not lead to changes in ICP but did
increase CPP when compared to not elevating the head. Although these studies lack
randomization, their findings support the work of Winkelman. Despite these positive findings,
March et al. (1990) reported that head elevation to 30° did not improve ICP or CPP compared to
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
no head elevation. Due to the small sample size (n=4) in the March study, it is unclear as to the
benefits of elevating the head to 30° post ABI.
More recently Ledwith et al. (2010) have suggested that no single position is optimal for
improving neurodymanic parameters in those who sustain an ABI. In this study authors placed
individuals who had sustained an ABI (n=33) into one of 12 positions, that were randomly
assigned. Positions included the supine, supine with knee bent, left lateral position and the
right lateral position. While in each of these positions the head of the bed was elevated to 15,
30, or 45 degrees. The level of brain tissue oxygen was significantly affected, notably a decrease
was seen, when participants were placed in the following positions: supine with the head of be
elevated to 45 º, Supine with head of be elevated to 45º, when placed in the right or left lateral
position with the head of the bed elevated to 30º. Further ICP was found to decrease
significantly when placed in the supine position with the head of the bed elevated to 45º
(p<0.002) and when the knee was elevated and the head of bed was elevated to 30º (p<0.039).
Increases in ICP were seen when placed in the left lateral position with the head of the bed
elevated to 15º (p<0.026), the right lateral position with a head of bed elevation of 15º
(p<0.002). Only one position had a significant effect on CPP, the left lateral position with the
head of the bed elevated to 30º (Ledwith et al., 2010).
In one paediatric study examining the benefits of elevating the head of bed by 10 cm
increments in children who had sustained a TBI, ICP was found to decrease (Agbeko et al.,
2012). If the head of bed (HOB) was lowered ICP was found to increase. Cerebral perfusion
pressure (CPP) was not found to change significantly as a result of adjusting the HOB. In total
study authors presented individuals with 66 HOB challenges. Study authors also noted that the
age and height of individuals is necessary when considering the HOB effect on ICP (Agbeko et
al., 2012).
Conclusions
There is Level 2 evidence suggesting a 30° head elevation reduces intracranial pressure with
concomitant increments in CPP.
There is Level 2 evidence to suggest head elevation does reduce ICP in children post TBI;
however, it was not found to have a significant impact of CPP.
Study findings suggest a 30° head elevation reduces ICP and improves CPP post TBI.
Elevating the head of the bed post TBI is effective at lowering intracranial pressure in
children, although the impact on CPP was minimal.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.2.1.2 Hypothermia
Mild to moderate hypothermia is a neuroprotective strategy that has been explored as a
measure to reduce secondary brain injury. It was first proposed as a possible treatment for ABI
more than a half a century ago (Fay 1945) and early reports suggested that hypothermia might
improve clinical outcomes. It is believed that hypothermia is initiated to control for 1) elevated
intracranial pressure (ICP) and 2) it limited the biochemical cascade believed to result in
secondary brain injury (Clifton 2004). Neuroprotective effects include reducing cerebral
metabolism, decreasing the inflammatory response, and decreasing the release of excitotoxic
levels of glutamate and free radicals post-ABI (Marion et al., 1997; Alderson et al, 2004; Globus
et al., 1995). Many studies looking at the benefits of hypothermia report on the impact it has on
ICP (Clifton, 2004).
It should also be noted that prolonged hypothermia is believed to be associated with various
adverse effects including arrhythmias, coagulopathies, sepsis and pneumonia which could
ultimately lead to a poorer clinical outcome (Alderson et al., 2004; Gadkary, and Signorini2004;
Schubert 1995). It has also been suggested that there may be a threshold during re-warming
above which pressure reactivity may reach damaging levels (Lavinio et al., 2007) and that there
is a critical window beyond which hypothermia may be ineffective (Clifton et al., 2009).
Several methods of hypothermic intervention have been suggested. Systemic hypothermia
involves cooling the entire body with cooling blankets (Qiu et al., 2007) and occasionally a
gastric lavage (Marion et al., 1993), while selective hypothermia aims to target the head
specifically using a cooling cap and neckband (Liu et al, 2006). The AANS noted that
prophylactic hypothermia showed no significant association with improved outcomes relative
to normothermic controls but preliminary findings suggested that mortality risks may be seen
when target temperatures were maintained for more than 48 hrs. They also suggested that
hypothermia was associated with significantly higher GOS scores (Bratton et al., 2007a).
Currently there are no EBIC recommendations for hypothermia.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.2 Hypothermia for the Acute Management of Elevated ICP Post ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Hypothermia in the Adult Population
Clifton et al.,
(2011)
USA/Canda
RCT
PEDro=9
N=92 Participants was stratified by centre
(6 centres participated) to either the
hypothermia or normothermia. Those in
the hypothermia groups had their core
body temperatures maintained to 35° C,
while those in the normothermia groups
were maintained at 37° C. Those
receiving hypothermia treatment by
intravenous instillation of up to 2L of cold
crystalloid and received wet sheets or gel
packs to help lower their core body temp.
Those in the normothermia group were
given cooling blankets to maintain their
core body temperature at 37° C.
Overall there were no significant
differences in there number of
individuals between the two groups
who developed complications (mortality
rates and long term outcomes). For
those in the hypothermia groups there
were episodes of increased intracranial
pressure. This resulted in an increase in
the total rate of complications for the
hypothermia group. Surgical patients in
the hypothermia group had poorer
outcomes then those in the
normothermia group (p=0.09). Further
those in the diffuse brain injury group
had more episodes of raised intracranial
pressure then those in the
normothermia group.
Lee et al.,
(2010)
China
RCT
D&B=19
N=45 Participants were randomly
assigned to one of 3 groups, after
undergoing a craniotomy. Those assigned
to group A participated in intracranial
pressure/cerebral perfusion pressure
guided management only, those in group
B received mild hypothermia and ICP/CPP
guided management and group C
received hypothermia and brain tissue
oxygen (Pti O2) with guided ICP/CPP
management
Overall no significant findings were
noted between the 3 groups.
Favourable outcomes were noted in
50% of the normothermia group, 60%
on the hypothermia only group and
71.4% in the Pti O2, and hypothermia
group. Mortality was highest in the
normothermia groups (12.5%) while in
the hypothermia groups it was 6.7% and
in the Pti O2, 8.5%.
Yan et al.,
(2010)
China
RCT
PEDro = 6
N=148 Participants were randomized
between normothermia (37±.05°C) and
hypothermia (32-34°C). All had sustained
a a severe TBI (GCS 3-8). Those in the
hypothermia group were placed on a
cooling bed. Treatment colling lasted 3 to
5 days. Following this patients were
rewarmed, until normal body
temperature was reached. While on the
cooling bed the pressure of the oxygen in
the brain tissue (PbrO2) and cerebral
oxygen saturation (rSaO2) was
monitored. Neuroelectrophysiological
For those with a GCS of 7-8, the PbrO2
levels were significantly greater for
hypothermia group compared to the
normotherapy group. Further the levels
were significantly higher for
hypothermia group, and the shortlatency somatosensory evoked potential
(SLSEP) wave amplitude was
significantly higher for hypothermia
group. For GCS 5-6, PbrO2 levels were
significantly greater for hypothermia
group, rSaO2 levels were significantly
greater for hypothermia group, short-
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
monitoring was also completed.
latency somatosensory evoked potential
(SLSEP) wave amplitude was
significantly higher for hypothermia
group. There were no significant
differences for the GCS 3-4 on any
measurements. GOS for those with a
GCS of 7-8 indicated more in the
hypothermia group made either a good
or moderate receovery (n=17)
compared to the normotherapy group
(n=13). For those with a GCS of 3-4 no
significant differences were noted on
the GOS at follow-up.
Harris et al.,
(2009)
USA
RCT
PEDro = 7
N=21 Severe TBI patients (GCS ≤8) were
randomized to receive either localized
hypothermia treatment (target 33°C) via
a cooling cap or normothermia. Patients
were monitored for brain and systemic
temperature with clinical outcomes of
mortality, GOS and FIM.
Beyond the first 3 hours of treatment,
the mean intracranial temperature was
significantly lower in the treatment
group than in the control group
(p<0.05), except at the 6 hour (p=0.08)
and 4 hour (p=0.08) point. The target
temperature of 33°C was rarely
achieved however and no significant
differences were noted on any of the
clinical outcomes between groups.
Qiu et al.,
(2007)
China
RCT
PEDro = 7
N=80 Patients with severe TBI after
unilateral craniotomy were randomized
into a hypothermia group with brain
temperature of 33-35°C maintained for 4
days and a normothermia group. Vital
signs, intracranial pressure, serum
superoxide dismutase levels, GOS scores
and complications were prospectively
analyzed.
Mean ICP of therapeutic group at 24, 48
and 72 hours was lower than control.
(23.49±2.38, 24.68±1.71, and
22.51±2.44 vs 25.87±2.18, 25.90±1.86,
and 24.57±3.95 mmHg; p = 0.000, 0.000,
and 0.003). Mean serum superoxide
dismutase levels were higher at days 3
and 7 in the intervention group
(533.0±103.4 and 600.5±82.9 vs
458.7±68.1 and 497.0±57.3 microg/L;
p=0.000). Percentage of favourable
neurological outcome at 1 year was 70%
vs 47.5% respectively
(P=0.041).Complication, including
pulmonary infection were higher in the
therapeutic group (57.5% vs 32.5%;
p=0.025).
Liu et al.,
(2006)
China
RCT
PEDro = 5
N=66 Brain injured patients (GCS ≤ 8)
were randomly allocated to one of three
groups: 22 to selective brain cooling SBC,
21 to mild systemic hypothermia MSH
and 23 to control. SBC consisted of
Both hypothermia groups showed a
significant decrease in ICP levels relative
to the control group at 24, 48 and 72
hours post injury (p<0.05). Superoxide
dismitase levels were significantly
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
cooing of the head and neck to 33-35°C
using a cooling cap and neckband with
circulating 4°C water. MSH was achieved
using a cooling blanket and refrigerated
ice bags to a rectal temperature of 3335°C. Control received the same
conventional treatment minus the
hypothermia.
higher in the hypothermia groups at 3
and 7 days post injury(p<0.01). The
percentage of patients with good GOS
scores 2 years post injury were 72.7%,
57.1% and 34.8% in the SBC, MSH and
control groups respectively.
Clifton et al.,
(2001)
USA
RCT
PEDro = 6
N=392 Severe (GCS 3-8) comatose nonpenetrating head injury patients were
randomly assigned to be treated with
hypothermia (33ºC) which was initiated
within 6 hours after injury and
maintained for 48 hours or
normothermia (37 ºC). GOS outcome 6
months post-injury was assessed as the
primary outcome.
Poor GOS outcome (severe disability,
vegetative state, death) were reported
in 57% of patients in both groups.
Mortality was not significantly different
between groups (28% in the
hypothermia group and 27% in controls,
p=0.79). Patients in the hypothermia
group had more hospital days with
complications than those in the
normothermia group. Fewer patients in
the hypothermia group had high ICP
compared with those in the
normothermia group.
Jiang et al.,
(2000)
China
RCT
PEDro = 6
N=87 Severe TBI patients (GCS ≤ 8) were
randomly assigned to prolonged mild
hypothermia (33-35ºC for 3-14 days.
Rewarming started when ICP returned to
normal levels) or normothermia (3738ºC). Acute changes in ICP between
groups were compared. Mortality rates
and clinical outcome using the GOS were
assessed 1 year later.
Mortality rate was 25.58% and the rate
of favourable outcome on the GOS
(good recovery/mild disability) was
46.51% in the hypothermia group while
in the normothermia group the
mortality rate was 45.45% and the rate
of favourable outcome was 27.27 %
(p<0.05). Hypothermia caused a
significant reduction in ICP (p<0.01) and
inhibited hyperglycemia (p<0.05).
Shiozaki et al.,
(1999)
Japan
RCT
PEDro=5
N=16 Following randomization
participants in the treatment group mild
hypothermia was begun by cooling the
body surface. Intracranial temperature
was maintained at 33.5 to 34.5°C. This
was continued for the first 48 hours, at
which time the participants were slowly
warmed to 37° C. During the warming
period, all were placed on barbiturates.
Those in the normothermia group had
their intracranial pressure maintained at
36.5 to 37.5° C by surface cooling for 5
days. Barbiturates were infused every 6
to 8 hours during the first 48 hours.
Overall there were no significant
changes in cerebrospinal fluid (CSF)
concentrations between the two groups
when looking at the presence of
excitatory amino acids or cytokines. Nor
was any difference in cardiac arrhythmia
between the two groups. Study results
indicate no significant changes between
the groups regardless of the group
participants were randomized to.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Marion et al.,
(1997)
USA
RCT
PEDro = 5
N=82 Severe closed head injury patients
(GCS 3-7) were randomized to a
hypothermia group (33ºC using cooling
blankets and cold saline gastric lavage
within a mean of 10 hours after injury
and maintained for 24 hours. Patients
were then rewarmed to 37-38.5ºC over
12 hours at a rate of 1 ºC/hour) or to a
normothermia group (37-38.5ºC). GOS
outcomes at 3, 6 and 12 months after
injury were compared.
At 3 months, the hypothermia group
had a significantly more patients with
favorable outcome (good
outcome/moderate disability) on the
GOS than the normothermia group (38%
vs. 17%, p=0.03). At 12 months, a
similar difference in favorable GOS
outcome between groups was observed
(62% vs. 38%, p=0.05). Patients with an
initial GCS of 3-4 did not benefit from
hypothermia, whereas those with scores
of 5-7 did. Among patients with an
initial GCS score of 5-7, significantly
more patients in the hypothermia group
had a favorable GOS outcome at 6
months than those in the normothermia
group (73% vs. 35%, p=0.008). During
the cooling period, the hypothermia
group had significantly lower ICP
(p=0.01), CBF (p=0.05) and heart rate
(p<0.001) and higher CPP (p=0.05)
compared with controls.
Resnick et al.,
(1994)
USA
RCT
PEDro = 5
N=36 Severe head injury patients (GCS ≤
8) were randomly assigned to therapeutic
hypothermia (cooled within 6 hours of
injury to a brain temperature of 32-33 ºC
for 24 hours using cooling blankets and
cold saline gastric lavage. Patients were
then rewarmed over 12 hours) or to a
normothermia group (brain temperature
of 37-38 ºC). At admission and again 24
hours after admission, the development
of delayed traumatic intracerebral
hematoma (DTICH) and the most
common tests of coagulation function:
prothrombin times, partial
thromboplastin times, and platelet
counts were used as outcome measures.
There were no significant differences
between groups in measured
coagulopathy, or in any of the
coagulation parameters. The incidence
of DTICH was not increased by the use
of moderate hypothermia (6/20 of the
patients in the hypothermia group
developed DTICH compared with 5/16
patients in the normothermia group).
The short-term use of hypothermia does
not appear to increase the risk of
intracranial hemorrhage complications
in head injury patients.
Marion et al.,
(1993)
USA
RCT
PEDro = 6
N=40 Severe closed head injury patients
(GCS 3-7) were randomized to
hypothermia (using cooling blankets and
cold saline gastric lavage within 10 hours
after injury and maintained for 24 hours.
Patients were rewarmed to 37-38ºC over
12 hours) or normothermia (37-38ºC).
ICP, CBF, and CMRO2 were assessed as
Hypothermia significantly reduced ICP
(p<0.004) and CBF (p<0.021) during the
cooling period. Mean CMRO2 in the
hypothermia group was significantly
lower during cooling and higher 5 days
after injury compared with the
normothermia group (p<0.001). There
was a trend toward a better outcome 3
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
well as the DRS and the GOS scales 3
months after the injury.
months post-injury in the hypothermia
group as 12 patients in this group
showed moderate, mild or no disability
(GOS score 3-4) whereas only 8 patients
in the normothermia group had
improved to this level (p<0.24). DRS
scores reflected the same trend, with
more patients in the hypothermia group
achieving a better outcome.
Shiozaki et al.,
(1993)
Japan
RCT
PEDro = 5
N=33 Severe head injury patients (GCS ≤
8) in whom ICP remained > 20 mmHg
despite high dose barbiturate therapy
were randomly assigned to mild
hypothermia (33.5 –34.5 ºC via surface
cooling using water circulating blankets
for 2 days. Patients were then rewarmed
slowly and core temperature maintained
between 35.5 –36.5 ºC for 24 hours. If
ICP again became elevated, patients were
re-cooled to 34 ºC) or no hypothermia.
ICP, CPP and GOS were assessed.
In the control group 12/17 died from ICP
hypertension, while mild hypothermia
significantly reduced ICP (p<0.01) and
increased CPP (p<0.01). Fifty percent of
the patients in the hypothermia group
survived compared with only 18% in the
control group (p<0.05), while 31% in the
hypothermia group and 71% in the
control group died from uncontrollable
ICP (p<0.05). Thirty-eight percent of the
patients in the hypothermia group had
good GOS outcome (good recovery/mild
disability) at 6 months compared with
only 6% in the control group.
Tokutomi et al.,
(2009)
Japan
Non-RCT
N=61 Patients treated with hypothermia
to 35 ºC were compared to those treated
with hypothermia to a target
temperature of 33 ºC. Patients were
monitored for ICP levels, CPP, serum
potassium levels, c reactive protein,
mortality and complications.
Patients in both groups exhibited
decreases in ICP below 20 mmHg with
no differences in the incidence of
intracranial hypertension or low CPP.
Patients cooled to 35 ºC showed
significant improvements in serum
potassium concentrations and Creactive protein levels with a trend
toward decreased mortality and fewer
complications.
Qiu et al.,
(2006)
China
Non-RCT
N=90 Severe TBI patients (GCS ≤ 8) were
divided between selective brain cooling
(SBC) group (cooling cap at 4ºC and a
neck band with ice straps) and a control
group. Patients were treated for 3 days
and monitored for ICP. They were then
followed up on at 6 months post-injury.
At 24, 48 and 72 hours ICP was lower for
the SBC group compared to the
normothermia group (19.14 ± 2.33 vs
23.41 ± 2.51, 19.72 ± 1.73 vs 20.97 ±
1.86 and 17.29 ± 2.07 vs 20.13 ± 1.87
mmHg respectively. P<0.01). There was
also significant difference in GOS good
neurological outcome rates 6 months
after injury (68.7% SBC vs 46.7%
Control, P<0.05)
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Carhuapoma et al.,
(2003)
USA
Case Series
N=6 Acute brain injury patients (GCS 5-9)
who experienced fever (≥ 38.0 ºC) that
was refractory to conventional
pharmacological antipyretic treatment
were treated with hypothermia (36.5ºC).
Prophylactic hand warming using instant
Hot Packs was used to prevent shivering
at the time of cooling.
Fever was significantly reduced to 36.9
ºC after 120 min (p<0.001). Core
temperature remained controlled and
unchanged during the remainder of the
treatment (36.6 ºC, p = 0.5; 36.6 ºC,
p=0.09, and 36.5 ºC, p=0.09 at 180, 300
and 600 minutes respectively).
Tokutomi et al.,
(2003)
Japan
Case series
N=31 Severe brain injury patients (GCS ≤
5) were cooled to 33 ºC. Hypothermia
was induced by surface cooling with
water-circulating blankets. Patients were
slowly rewarmed after 48-72 hours of
hypothermia. Changes in ICP and CPP
were assessed.
Incidence of elevated ICP decreased
significantly with hypothermia
(p>0.0001). ICP decreased significantly
at brain temperatures < 37 ºC and
decreased more sharply at 35-36 ºC, but
no differences were observed at
temperatures < 35 ºC. CPP peaked at
35-35.9ºC and decreased with further
decreases in temperature.
Yamamoto et al.,
(2002)
Japan
Non-RCT
N=84 Severe TBI patients (GCS 3-7) were
assigned to receive mild therapeutic
hypothermia (33-35 ºC) for at least 36
hours to a maximum of 7 days according
to the severity of brain injury or to a
group which was treated without
hypothermia. GOS outcome 3 months
after the injury was compared between
groups. The hypothermia group was
subdivided into 2 subgroups according to
their GOS outcome: GOOD (good
recovery/mild disability) and POOR
(severe disability/vegetative/death).
The mild hypothermia group had
significantly better outcome and
significantly lower mortality compared
with the control group (p<0.05). The
patients in the good outcome subgroup
were significantly younger, and their
cerebral perfusion pressure was
significantly higher during hypothermia
compared with the poor outcome
subgroup (p<0.05).
Polderman et al.,
(2002)
Netherlands
Non-RCT
N=136 Severe (GCS ≤ 8) head injury
patients in whom intracranial pressure
ICP remained > 20 mmHg despite
conventional therapy were assigned to
be treated with moderate hypothermia
(32-34 ºC) using water-circulating
blankets (n=64) or no hypothermic
treatment (n=72). If ICP remained ≤ 20
mmHg for 24 hrs, the patient was slowly
rewarmed (1ºC per 12 hrs). If ICP
increased to > 20 mmHg, the
temperature was again decreased until
ICP decreased below 20 mmHg. Mortality
and neurological outcome were assessed
at 6 months using the GOS.
The average duration of hypothermia
was 4.8 days (range 1-21 days). ICP
decreased markedly in all patients
during cooling. Actual mortality rates
were significantly lower in the
hypothermia group compared with the
control group (62% vs. 72%, p<0.05).
The number of patients with good
neurological outcome on the GOS was
also higher in the hypothermia group
(15.7% vs. 9.7%, p<0.02).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Gal et al., (2002)
Czech Republic
Non-RCT
N=30 Severe head injury patients (GCS 38) were assigned to a hypothermic group
(core temperature reduced to 34 ºC
within 15 hours of injury and maintained
for 72 hours using forced air cooling
combined with circulating water mattress
cooling) or a normothermic group (36.537.5 ºC). Changes in ICP, CPP, and GOS
outcome at 6 months were assessed.
The 2 groups were similar in terms of
injury severity, age, sex ratios, and CT
findings. There was a significant
reduction in ICP in the hypothermia
group compared with the normothermia
group (p=0.0007). Significant increase
in CPP during hypothermia (p=0.0007)
with unchanged (p=0.90) systolic arterial
pressure values compared with
normothermia. No significant
differences between groups in GOS
outcome 6 months post-injury (p=0.084)
although 87% of patients in the
hypothermia group reached good
neurological recovery (GOS 4-5)
compared with only 47% in the control
group.
Chen et al., (2001)
China
Non-RCT
N=30 Severe head injury patients were
assigned to moderate hypothermia (3335 ºC over 10 hours for 3-10 days) or
routine treatment. Mortality, endothelin
(ET) in serum, neuron-specific enolase
(NSE) in serum, free radical scavenger
superoxide dismutase (SOD) was
compared between groups.
Mortality was significantly lower in the
hypothermia group than in the control
group. Hypothermia greatly reduced ET
and NSE and increased SOD. No obvious
differences found in arterial pH, pO2,
+
+
pCO2, serum K , Na , or Cl between
groups.
Tateishi et al.
(1998)
Japan
Case Series
N=9 Severe brain injury patients (GCS ≤ 8)
with ICP ≥ 20 mmHg despite conventional
therapeutic measures were subjected to
a maximum of 6 days of mild
hypothermia induced by repeated
intragastric cooling using a nasoduodenal
tube of iced half-saline infused during 15
– 30 min supplemented with surface
cooling. This process was repeated if
necessary to reduce and maintain ICP <
20 mm Hg. Changes in ICP and jugular
venous oxygen saturation at baseline and
3 hours after beginning of hypothermia
were assessed. Platelet count, C-reactive
protein, and GOS at 6-12 months after
discharge were also evaluated.
Hypothermia was induced within 24
hours of admission and continued for 20
– 118 hours (mean = 68 h). Lowest brain
temperature obtained ranged from 33 –
35 ºC. Significant reduction in ICP 3
hours after cooling (p<0.05), however 4
patients experienced systemic infection
complications. Increased C-reactive
protein and decreased platelet count
were observed in all patients during
hypothermia. 7/9 patients showed good
recovery or moderate disability
according to the GOS 6 – 12 months
after discharge.
Hypothermia in the Pediatric Population
Adelson et al.,
(2013)
N=77 Children were randomly assigned
to either the hypothermia group or the
When looking at the main outcome,
mortality at 3 months, researchers
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
USA
RCT
D&B=27
normothermia group. Children in the
hypothermia group were cooled with
iced saline to 34-45° c then surfaced
cooled to 32-33° C. This was maintained
for a 48 hour period at which time all
were rewarmed. Those in the
nomothermia group had their
temperatures maintained at 36.5-37.5°C.
If temperature rose to 38°C cooling
blankets and rectal acetaminophen was
given.
found no difference between the
groups. Six of 39 in the hypothermia
group died and 2 of 38 in the
normothermia group died. When
looking at global functioning no
significant differences were found
between the two groups. Due to the
lack of significant findings between the
two groups, the study was halted much
earlier then planned.
Hutchison et al.,
(2013)
Canada
RCT
N=205 Children with TBI were
randomized to receive either
hypothermia (32.5±0.5°C) or
normothermia (37±0.5°C) treatment.
Patients 7 or older showed more
unfavorable outcomes in the
hypothermia group than in the
normothermia group (P=0.06). Patients
with intracranial pressure lower than 20
mm Hg also showed a higher risk of
unfavorable outcomes when receiving
hypothermia treatment as opposed to
normothermia. No benefits were noted
in either intermediate or long-term
outcomes after hypothermia treatment.
Bayir et al.,
(2009)
USA
RCT
PEDro = 6
N=28 Infants and children with severe TBI
(GCS ≤8) were randomized to receive
either hypothermia (32-33°C) or
normothermia (36.5-37.5°C). Patients
were monitored for measures of
oxidative stress.
When looking at the total antioxidant
reserve, those in the normothermai
group showed reduction reduction vs
those in the hypothermia group post
injury. An inverse relationship between
cerebrospinal fluid (CSF) total
antioxidant reserve and temperature
after injury (p=0.022) was found.
Glutathione concentration in CSF and
temperature were inversely related
after injury (p=0.002). The effect of
temperature was not significant on CSF
prot-SH (p=0.104) or CSF F2-isoprostane
levels (p=0.104).
Li et al.,
(2009)
China
RCT
PEDro = 6
N=22 Children (6-108 months) who were
severely brain injured
(GCS ≤8) were randomized to either
localized hypothermia treatment using a
cooling cap (34.5±0.2°C) or
normothermia (38.0±0.5°C).
Hypothermia treated group showed
lower ICP values which were statistically
significant at 8, 24, 48 and 72 hours
(P<0.05). In the hypothermia group,
neuron-specific enolase levels were
lowered (P<0.05), S-100 levels were
lowered (P<0.01) and creatine kinase-BB
levels were also lowered (P<0.0001) in
comparison with the normothermia
Follow-up to the
2008 study
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
group.
Hutchison et al.,
(2008)
RCT
Canada
PEDro= 9
N=225 Children aged 1-17 years with a
TBI were randomly assigned to either a
hypothermia (n=108) or nonhypothermia group (n=117). The
intervention group maintained a mean
temperature of 31.9-34.3 degrees while
the control group maintained a normal
temperature (36.5-37.4 degrees).
At 6 months, 22% of the nonhypothermia group had only sustained
unfavorable outcomes in comparison to
31% of the hypothermia group. A total
of 23 deaths were recorded among the
hypothermia group in contrast to the 14
deaths among the non-hypothermia
group.
A total of 205 children completed the
study.
Adelson et al.,
(2005)
USA
RCT
PEDro = 6
N=48 Children aged 0-13 yrs within 6
hours of injury, GCS < or = 8, abN CT.
Hypo 1: Phase II multicenter, randomized
controlled trial. HYPO 32- 33deg vs
NORM 36.5-37.5 deg.
Hypo 2: single institution > 6 hours post
injury, unknown time of injury or 13-18
years.
Mortality, coagulopathy, arrhythmia,
infection, intracerebral hemorrhages,
mean ICP or CPP no significant
difference. Glascow Outcome Score
(GOS), Vineland Adaptive Behaviour
Scale and Child Health Questionnaire at
3 and 6 months no significant difference
(*study designed as a safety trial). Phase
III clinical trial required to determine
efficacy.
Biswas et al.,
(2002)
USA
RCT
PEDro = 7
N=21 Children who met study inclusion
criteria were randomly assigned to either
the hypothermia group or the
normothermia group. All were
administered sedative, analgesics and
neuromuscular blocking agents during
the first 48 hrs of hospital admission.
Those assigned to the hypothermia group
began the cooling process immediately
following enrollment into the study.
Rectal temperature was lowered and
maintained at 32° to 34°C. This
temperature was maintained for a period
of 48 hrs, at which time participants were
slowly rewarmed. Those in the
normothermia group received standard
treatment.
Results indicate there were no
statistically significant differences
between the groups when looking at ICP
changes (p=0.73), nor was there any
significant differences were looking the
overall ICP levels between the two
groups (p=.77). A trend in lower ICP
levels was noted in the hypothermia
th
th
group during the 13 to the 28 hour
during treatment. CPP levels for each
group decreased slowly during the first
60 hrs, then began to increase. The rate
of increase was not significant for either
group. ICP levels were within the normal
range for approximately 90% (or more)
of the time each day for the
hypothermia group compared to the
normothermia group.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al. 2002)
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Table 16.2a Summary of RCT Studies of Acute Hypothermia Post ABI
Authors
Methods
Results
Adult Population
Clifton et al.,
N=97 Hypothermia (35°C) vs
- Results of the GOS indicate poor outcomes
(2011)
Normothermia 37°C)
were higher in the hypothermia group
compared to the normothermia group.
Lee et al.,
N=45 study consisted of 3 groups: A)
-Groups B and C had lower ICP levels than
(2010)
ICP/CPP guided management, B)
Group A.
ICP/CPP guided management with mild + fewer deaths were noted in Group B
hypothermia and C) mild hypothermia
+ PtiO2. increased as ICP levels decreased.
and PtiO2.
Yan et al.,
N=148 Hypothermia (32-34°C) vs.
+ oxygenation & electrophysiologic
(2010)
normothermia (37±.05°C)
measures (GCS 5-6 and 7-8)
ND oxygenation or electrophysiologic
measures (GCS 3-4)
ND clinical outcomes (GOS 1-7 years)
Harris et al.,
N=21 Hypothermia via cooling cap vs.
+ cerebral temperature
(2009)
normothermia
ND mortality, FIM, GOS
Qiu et al.,
N=80 Mild hypothermia (33-35°C) for 4 + for decreased ICP
(2007)
days after decompressive craniotomy.
+ for increased superoxide dismutase
+ for better GOS outcomes at 1 year
+ for increased pulmonary complications
Liu et al.,
N=66 Mild Hypothermia (33-35°C),
+ for reduced ICP relative to control
(2006)
Selective brain cooing (33-35°C brain
+ for increased SOD relative to control
temp)
+ for better outcomes 2 years post injury
relative to control
Jiang et al.,
N=87 Mild hypothermia (33-35ºC) for
+ lower mortality
(2000)
up to 14 days vs. normothermia.
+ for favourable GOS
+ for reduction in ICP
Clifton et al.,
N=392 Mild hypothermia (33 ºC) within ND for GOS outcome
(2001)
6 hrs of injury and maintained for 48
ND for mortality
hrs vs. normothermia.
- for medical complications
+ for number of patients with elevated ICP
Marion et al.,
N=82 Mild hypothermia (33ºC for 24
+ favorable GOS only for patients with initial
(1997)
hours) vs. normothermia.
GCS of 5-7.
+ reduction in ICP, CBF and HR and higher
CPP.
Marion et al.,
N=40 Mild hypothermia (32-33ºC) for < + for reduction in ICP and CBF and CRMO2
(1993)
24 hours vs. normothermia.
ND for favorable GOS or DRS outcome
Resnick et al.,
N=36 Mild hypothermia (32-33 ºC) for
ND for caugolopathy parameters
(1994)
24 hours vs. normothermia.
Shiozaki et al.,
N=33. Mild hypothermia (33.5 –34.5
+ for reduction in ICP and increase in CPP
(1993)
ºC) for 2 days vs. normothermia.
+ for survival
Pediatric population
Adelson et al.,
N=77 Hypothermia (34-35°C) vs
ND between groups when looking at
(2013)
Normothermia (36.5-37.5°C)
mortality or global function (GOS) at 3
months
Hutchinson et
N=205 Hypothermia (32.5±0.5°C) vs
- >7 more unfavorable outcomes in those
al., (2013)
normothermia (37±0.5°C)
over 7
ND in either intermediate or long-term
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Authors
Bayir et al.,
(2009)
Hutchison et al.,
(2008)
Li et al.,
(2008)
Methods
N=28 Infants and children.
Hypothermia (32-33°C) vs.
normothermia (36.5-37.5°C).
N=225 Children (1-17)
Hypothermia (32.5°C for 24 hr) vs.
normothermia (37°C).
N=22 Children (6 to 108 months)
Hypothermia via cooling cap
(34.5±0.2°C) vs. normothermia
(38.0±0.5°C)
N=21 Hypothermia vs. normothermia
Results
outcomes after hypothermia treatment.
+ attenuation of oxidative stress
ND unfavorable outcome at 6- months
+ ICP reduction
+ CSF biochemical markers
Biswas et al.,
No differences between groups
(2002)
ND = No difference between groups; + = Improvement compared with control; - = Impairments
compared with control
Table 16.2b Summary of non-RCT Studies of Acute Hypothermia Post ABI
Authors
Methods
Results
Tokutomi et al.,
N=61 Hypothermia to 35 ºC vs.
+ serum potassium levels and C-reactive
(2009)
hypothermia to 33 ºC
protein levels
ND ICP control, mortality, complications
Qiu et al.,
N=90 Selective brain cooling (33-35
+ lower ICP
(2006)
ºC) for 3 days vs control.
+ for positive GOS outcome 6 months post
injury
Carhuapoma et al., N=6 Mild hypothermia (36.5ºC).
+ for reduction in fever after 120 min of
(2003)
hypothermia
Tokutomi et al.,
N=31 Moderate hypothermia (33ºC)
+ for reduction in ICP at brain temperatures
(2003)
of 35-37ºC. ND at temperatures < 35ºC.
CPP peaked at 35-35.9ºC and decreased
with lower temperatures.
Gal et al.,
N=30 Mild hypothermia (34ºC) for
+ for reduction in ICP and increase in CPP
(2002)
72 hours vs. normothermia.
ND for systolic BP
ND in GOS outcome at 6 months
ND for favourable GOS or DRS outcome
Polderman et al.,
N=136 Mild hypothermia (32-34ºC)
+ for reduction in ICP
(2002)
vs. no hypothermia.
+ for lower mortality
+ for good GOS outcome
Yamamoto et al.,
N=84 Mild hypothermia (33-35 ºC)
+ for GOS outcome
(2002)
for 36 hrs – 7 days according to
+ for lower mortality
severity of injury vs. no
+ for CPP in patients with good GOS
hypothermia.
outcome
Chen et al.,
N=30 Mild hypothermia (33-35ºC)
+ lower mortality
(2001)
for 3-10 days vs. routine treatment.
+ for reduction in ET and NSE and increased
SOD
Tateishi et al.,
N = 9. Mild hypothermia (33-35 ºC)
+ for reduction in ICP 3 hrs after cooling
(1998)
for up to 6 days.
+ for increased C-reactive protein and
decreased platelet counts
+ for favourable GOS outcome 6-12 months
post-discharge
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
ND = No difference between groups; + = Improvement compared with control
Discussion
Due to the inconsistent findings in a variety of clinical trails, Clifton and colleagues (2011)
revisited the treatment effects of hypothermia post TBI. In this RCT patients (n=97) were
randomized to either a hypothermia group (n=52) or a normothermia group (n=45).
Participants ranged in age from 16 to 45 years and all had sustained a non-penetrating brain
injury and had a GCS <8. Overall no significant findings were noted between the two groups.
Study findings lead authors to conclude that early hypothermia induction did not act as a
neuroprotectant in those with a diffuse brain injury.
Lee et al., 2010 investigated the efficacy of hypothermia on ICP and CPP in 45 individuals post
TBI. Participants in this study were randomly assigned to one of 3 groups: group A (n=16)
received intracranial pressure/cerebral perfusion (ICP/CPP) pressure guided management only
(normothermia group); group B (n=15) received mild hypothermia and ICP/CPP guided
management; group C (n=14) received mild hypothermia and PtiO2 guided with ICP/CPP
management. Body temperature was monitored throughout the study. ICP levels were lower
for those in groups B and C compared to group A. Overall mean GOS was highest for those in
group C while mean ICP was lowest for this group. There were not significant differences in the
number of deaths in each group. Study authors concluded that mild hypothermia coupled with
PtiO2 was beneficial post TBI (Lee et al., 2011).
In the study conducted by Yan and colleagues 148 participants were randomized to
hypothermia or normothermia. Participants were divided into three groups by GCS (3-4, 5-6, 78) prior to randomization. Cerebral oxygenation and electrophysiological markers were
monitored. They noted that both oxygenation and electrophysiological outcomes were
improved in patients who received hypothermia and had initial GCS 7-8, but not in the group
who had a GCS of 3-4. Those with GCS of 5-6 the effect was unclear (Yan et al., 2010).
Harris et al. (2009) conducted a randomized clinical trial to evaluate the impact of selective
brain cooling using a cooling cap on cerebral temperature and clinical outcomes. Although they
reported that cerebral temperatures were indeed decreased by hypothermia, no differences
were noted between groups on mortality, FIM gain, or GOS outcome.
Qiu et al. (2007) conducted an RCT of mild hypothermia in severe TBI patients who had
undergone craniotomy. Mean ICP of therapeutic group at 24, 48 and 72 hours was lower than
controls. Mean serum superoxide dismutase levels were higher at days 3 and 7 in the
intervention group (P=.000). Percentage of favourable neurological outcome on the GOS scale
at 1 year was 70% vs 47.5% in the therapeutic group and control group respectively (P=.041).
Complications, including pulmonary infection were higher in the therapeutic group (57.5% vs
32.5%; p=.025). Qiu et al. (2007) reported that complications were managed without severe
sequelae and they noted that mild hypothermia was an effective treatment after
decompressive craniotomy.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Liu et al. (2006) performed a randomized controlled study to assess the use of selective brain
cooling relative to mild systemic hypothermia and a control group. Decreased ICP levels and
increased superoxide dismutase levels post injury were found in both hypothermia groups
compared to the control group. Patients Results of the GOS scale, 2 years post intervention,
indicated selective brain cooling was likely to improve patient outcomes compared to other
groups. While the study was limited by the small sample number, the authors recommended
selective brain cooling as a safe and effective treatment post TBI (Liu et al., 2006).
In an RCT conducted by Clifton et al. (2001) participants were randomly assigned to either the
hypothermia group (treatment group) or the normothermia group (control group). Cooling for
both groups began approximately 4 hours post injury. Outcomes of the treatments revealed no
differences in the number of individuals who died in each group (28% for the hypothermia
group compared 27% for the normothermia group)and there were not significant differences in
the results of the neurobehavioral and neuropsychological tests conducted 6 months post
treatment. Further more medical complications were reported in those over 45 years of age in
the hypothermia group only. Despite these negative findings, the authors reported that
hypothermia was associated with significant reductions in elevated ICP during the first 96 hours
of treatment (Clifton et al., 2001). The study authors concluded that hypothermia treatment
was “not effective in improving outcomes in those with a severe TBI”.
Jiang et al. (2000) investigated the effects of prolonged mild hypothermia (33-35ºC) for up to 14
days in severe TBI patients. The authors reported that compared with patients randomized to
normothermia (control group), the hypothermia group exhibited significantly lower mortality
rates, and considerably more patients with favorable outcomes on the GOS 1 year post-injury.
Moreover, the study authors reported that after 7 days of treatment, hypothermia was
successful in significantly reducing elevated ICP compared with patients in the control group.
Contrary to common clinical beliefs, the authors found that 1 year post-injury, patients who
underwent prolonged hypothermia did not experience a higher number of complications:
including pneumonia, arrhythmias, hypotension, than those in the normothermia group (Jiang
et al., 2000).
Marion et al. (1993) reported on the results of early short-duration (<24 hours) mild (32-33ºC)
hypothermia. Their findings indicated during the cooling period, patients randomized to the
hypothermia group demonstrated significant declines in ICP, CBF and cerebral metabolism
compared with patients in the normothermia group. Although there was a trend favoring the
hypothermia group, there were no significant differences in favourable outcomes on the GOS
and the DRS 3 months post-injury. In a subsequent study by the same researchers using the
same treatment paradigm and general methodology, the previous findings were further
elucidated (Marion et al., 1997). The authors reported that patients with more severe injuries
(initial GCS of 3-4) did not benefit from hypothermia. However, among patients with an initial
GCS of 5-7, significantly more patients in the hypothermia group had a favorable GOS outcome
at 6 months compared to those in the normothermia group.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
In a related study of mild short duration hypothermia, Resnick et al. (1994) aimed to investigate
the incidence of potential adverse coagulation effects associated with hypothermia. They
reported that hypothermia did not increase the incidence of coagulopathy as measured by
differences in coagulation parameters 24 hours post-treatment (Resnick et al., 1994). However,
their findings did not exclude the possibility that hypothermia could lead to increased
hemorrhagic complications in the long-term.
Shiozaki et al. (1993) conducted an RCT to investigate the efficacy of mild hypothermia (33.5 –
34.5 ºC for 2 days) for the treatment of uncontrollable elevation in ICP. They reported that
hypothermia significantly reduced ICP while at the same time increasing cerebral perfusion
pressure. Furthermore, treatment with hypothermia was associated with a significantly lower
level of mortality and a higher incidence of favourable GOS outcome than normothermia at 6
months.
Each of the remaining studies utilized non-randomized research designs. All of these studies
indicated that hypothermia markedly reduced elevation in ICP. Furthermore, this reduction
was associated with a concomitant increase in CPP, which may explain the lower mortality rates
in those treated with hypothermia. The findings of Chen et al. (2001) also suggested that the
favorable effects of hypothermia could arise from reductions in mediators of secondary brain
injury while at the same time increasing the levels of some free radical scavengers.
When looking at the pediatric literature results were less favourable. In a large clinical trial,
Hutchison et al. (2008) reported that a period of hyporthermia for 24hrs post injury resulted in
no difference in favourable outcomes at 6 months on the GOS among children aged 1-17. In
fact, they noted several trends towards worse outcomes among patients who received
hypothermia. In another much smaller trial, Bayir et al. (2009) noted that among infants and
children, hypothermia helped to attenuate oxidative stress when compared to normothermia.
Similarly, Li et al. (2008) noted that in children 6-108 months, selective brain cooling with a
cooling cap reluted in decreased ICP and improved CSF biochemical markers. Unfortunately,
neither of these two trials evaluated differences in clinical outcomes.
A 2004 Cochrane review noted that there was no evidence that hypothermia is beneficial in the
treatment of TBI (Alderson et al., 2004). They reported that although early studies on the
subject suggested that hypothermia may be beneficial, there have been no larger trials which
have repeated these results. They also point out that the increased risks of pneumonia and
other potentially harmful side-effects make its use inappropriate unless clear benefits are
suspected. Another RCT identified by our review reported similar findings to those of this metaanalysis (Clifton et al., 2001). However, several studies have been released since these reviews
were published and research continues into the benefits of hypothermia in ABI management.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Extensive research has been conducted looking at the impact of hypothermia on clinical
outcomes post ABI in both the adult and pediatric populations. Despite the number of studies
conducted and the methods used, more research appears to be needed.
Meta-Analysis Results
As part of this current update we decided conduct a meta-analysis looking at the efficacy of
hypothermia treatment post-acute ABI with an adult population. We looked at the impact on
mortality rates, intracranial pressure and Glasgow Outcome Scores. Results from the metaanalysis indicate there were no significant differences between the hypothermia groups and
the normothermia groups when looking at the mortality rate post treatment (see Figure 1).
However results from the meta-analysis looking at the long term outcomes as measured by the
GOS, those receiving hypothermia were found to have better outcomes (see Figure 2).
Figure 1: Hypothermia Treatment and Mortality post ABI
Study name
Statistics for each study
Odds ratio and 95% CI
Odds Lower Upper
ratio
limit
limit Z-Value p-Value
Wu et al., 2006
Qiu et al., 2009
Jiang et al., 2000
Clifton et al., 2001
Harris et al., 2009
Qiu et al., 2007
Resnick et al., 1994
Clifton et al., 2011
Marion et al., 1993
Lee et al., 2010
0.325
0.282
0.413
1.048
2.250
0.603
0.917
1.472
0.180
0.519
0.750
0.112 0.947
0.107 0.748
0.167 1.021
0.663 1.657
0.439 11.522
0.223 1.630
0.315 2.671
0.578 3.747
0.008 4.009
0.066 4.083
0.560 1.006
-2.059
-2.545
-1.915
0.199
0.973
-0.997
-0.159
0.811
-1.082
-0.624
-1.921
0.039
0.011
0.055
0.842
0.330
0.319
0.873
0.417
0.279
0.533
0.055
0.5
1
hypothermia
2
normothermia
Meta Analysis
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Figure 2: Hypothermia Treatment and Glasgow Outcome Scores Post-Acute ABI.
Study name
Statistics for each study
Odds ratio and 95% CI
Odds Lower Upper
ratio
limit
limit Z-Value p-Value
Yan et al., 2010
Qiu et al., 2007
Lee et al., 2010
Qui et al., 2007
Jiang et al., 2000
Marion et al., 1993
Liu et al., 2006
1.171
2.579
0.550
2.579
2.319
2.250
3.500
1.914
0.605 2.267
1.030 6.457
0.106 2.860
1.030 6.457
0.948 5.669
0.635 7.973
1.209 10.131
1.338 2.738
0.469
2.023
-0.711
2.023
1.844
1.256
2.310
3.552
0.639
0.043
0.477
0.043
0.065
0.209
0.021
0.000
0.1 0.2 0.5 1
hypothermia
2
5 10
normothermia
Meta Analysis
Conclusions
There is Level 2 evidence to suggest hypothermia treatment helps to improve long term
outcomes post ABI.
There is conflicting evidence regarding hypothermia’s effect on mortality.
There is Level 1b evidence that systemic hypothermia is associated with an increased
incidence of pneumonia.
Although hypothermia has been shown to reduce elevated ICP by some researchers
there is no solid evidence to support its effectiveness post ABI. More research needs to
be done.
Systemic hypothermia increases the risk of pneumonia post ABI.
16.1.1.3 Hyperventilation
Controlled hyperventilation to achieve a PaCO2 of 30-35 mmHg during the first few days
following an ABI has been reported to improve outcomes. Hyperventilation causes cerebral
vasoconstriction, thus leading to a decrease in cerebral blood flow and cerebral blood volume
and hence leading to a decrease in ICP (Muizelaar et al., 1991). During mild hyperventilation,
increased oxygen extraction mechanisms allow compensation for decreases in blood flow and
volume allowing normal cellular metabolism to continue (Diringer et al., 2000). There is concern
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
that intense or prolonged hyperventilation may increase metabolic acidosis, which is common
following brain injury. Depletion of oxygen supplies forces the injured brain to turn to anaerobic
metabolism and an increase in lactic acid which has been correlated with poor outcomes
(DeSalles et al., 1986; DeSalles et al., 1987). Since hyperventilation decreases cerebral CO2, this
leads to an increase in pH diminishing the detrimental effects of acidosis. However, this
process depends on the availability of bicarbonate in the cerebral spinal fluid. Thus, prolonged
hyperventilation may not be an appropriate therapeutic measure since this may deplete
bicarbonate levels favoring ischemia and leading to poorer outcomes.
Several studies have also discussed concerns related to pre-hospital intubation leading to
inappropriate hyperventilation (Warner et al., 2007; Lal et al., 2003). Targeted hyperventilation
to within 30-45 mmHg have been associated with decreased mortality rates but both precise
regulation and proper training are recommended.
The AANS make Level II recommendations that prophylactic hyperventilation not be used. They
make Level III recommendations that hyperventilation be used as a temporizing measure for
reduction of elevated ICP but that it should be avoided within the first 24 hours after injury
when CBF is often critically low. They recommend that jugular venous oxygen saturation or
brain tissue oxygen tension measurements be performed when hyperventilation is used
(Bratton et al., 2007f) The EBIC recommend hyperventilation to manage high ICP and CPP in
association with sedation and analgesia. They recommend mild to moderate hyperventilation
initially to a PaCO2 of 30-35 mmHg. If this fails to control ICP along with osmotic therapy and
CSF drainage, then intensive hyperventilation to <30 mmHg is recommended with jugular
oxymetry monitoring of cerebral oxygenation to detect ischemia (Maas et al., 1997).
Individual Studies
Table 16.3 Hyperventilation for the Treatment of Elevated ICP Post ABI
Author/Year/
Country/Study
design/PEDro
Scores
Muizelaar et al.,
(1991)
USA
RCT
PEDro = 20
Methods
N=113 Patients with a severe brain
injury (GCS < 6) were randomly assigned
into 3 groups. Groups were given the
following: 1) those in the control group
received normal ventilation PaCO2 35 ±
2 mmHg; 2) those in the
hyperventilation (HV) group received
PaCO2 25 ± 2 mm Hg; or 3) those in the
hyperventilation plus buffer
tromethamine (THAM) received PaCO2
25 ± 2 mm Hg + THAM (0.3 M IV
solution. This was initially administered
as bolus and then as continuous infusion
Outcome
At 3 and 6 months after injury, the number of
patient with favorable outcome (good or
moderately disabled) on the GOS was
significantly lower (p < 0.03 and p < 0.05
respectively) in the HV group compared with
control and HV + THAM group. This occurred
only in patients with a motor score of 4-5 but
not 3 or less. The interaction effect of
treatment group and motor score was found
to be significant (p < 0.02) indicating that the
detrimental effect of hyperventilation was
limited to patient with better prognosis at
admission. At 12 months post-injury, this
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Scores
Methods
Outcome
for 5 days to achieve an arterial pH of
difference in outcome between groups was
7.6. The GOS at 3, 6 and 12 months was no longer significant (p = 0.13). There were no
used to assess outcomes.
significant differences in GOS outcome at any
of the 3 time points between the HV + THAM
treated group and the control group (no p
value provided).
Coles et al.,
(2002)
UK
Case Control
N=33 TBI patients (GCS 3-13) underwent
PET imaging of cerebral blood flow at
baseline and after reduction to 29±1
torr PaCO2. Jugular venous saturation
(Sjv O2) and arteriovenous oxygen
content differences (AVDO2) were
monitored in 25 patients and related to
PET variables.
The volume of critically hypoperfused
(hypoBV) and hyperperfused (hyperBV) brain
were calculated based on thresholds of 10
and 55 mL/100g/min respectively. Only
hypoBV values were significantly higher in
hyperventilated patients compared to
controls (p<0.05). Hyperventilation decreased
ICP (p<0.001) and increased CPP (p<0.0001).
Despite this improvement, hyperventilation
decreased global cerebral blood flow (31±1 to
23±1 mL.100g/min; p<0.0001) and increased
hypoBV (p<0.0001). Hyperventilation induced
increases in hypoBV were non-linear with a
threshold between 34 and 38 torr.
Diringer et al.,
(2000)
US
Case Control
N=9 Severe brain injury patients (GCS <
9) were studied a mean of 11.2±1.6 hrs
after TBI and compared to 10 healthy
normocapnic controls. Patients were
hyperventilated to 30±2 mmHg PaCO2
for 30 mins. Measurement of CBF, CBV,
CMRO2, oxygen extraction factor (OEF),
and cerebral venous oxygen content
(CvO2) were taken before and after 30
mins of hyperventilation.
Global CBF, CBV and CvO2 did not differ
between groups but in the TBI patients,
CMRO2 and OEF were reduced (1.59±0.44
ml/100g/min (p<0.01) and 0.31±0.06
(p<0.0001) respectively). During
hyperventilation, global CBF decreased to
25.5±8.7 ml/100g/min (p<0.0009), CBV fell to
2.8±0.56 ml/100g (p<0.001), OEF rose to
0.45±0.13 (p<0.02), and CvO2 fell to 8.3±3
vol% (p<0.02). CMRO2 remained the same.
Thiagarajan et al.,
(1998)
India
Case Series
N=18 Severe brain injury patients (GCS ≤
9) undergoing hyperventilation (PaCO2
25 mmHg) for the management of ICP
received concurrent hyperoxia (PaO2
200-250 mmHg) to determine if this
prevents the deleterious effects of
hyperventilation. Jugular venous bulb
oxygen saturation (SJvO2) and
arteriovenous oxygen content
difference (AVDO2) which provide
indirect evidence of the adequacy of
cerebral blood flow and global
oxygenation were evaluated.
SJvO2 decreased significantly during
hyperventilation (PaCO2 25 mmHg) and
returned to baseline when PaCO2 was
restored to 30 mmHg (p<0.0001) or when
hyperoxia (PaO2 200-250 mmHg) was induced.
Similarly AVDO2 increased significantly when
hyperventilation was induced (PaCO2 25
mmHg at a PaO2 of 100-150 mmHg) and
recovered to baseline values when the PaCO 2
was restored to 30 mm Hg (p<0.001) or when
PaO2 was increased to 200-250 mmHg
(P<0.001).
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Discussion
Muizelaar et al. (1991) conducted an RCT in which patients received prolonged hyperventilation
for 5 days. As expected, the authors reported that prolonged hyperventilation alone lead to
poor clinical outcomes. This is likely due to a depletion of cerebral bicarbonate supplies. Study
authors found that the combination of hyperventilation and tromethamine, a weak base and
buffer that crosses the blood brain barrier, resulted in significantly better outcomes than
hyperventilation alone (Muizelaar et al., 1991). This suggests that the deleterious effects of
prolonged hyperventilation may be overcome by the addition of a buffer system capable of
taking over once cerebral bicarbonate levels are depleted. Muizelaar et al. (1991) suggest that
“the use of THAM seem to counteract the deleterious effect of prolonged hyperventilation and,
therefore, its use would be beneficial when sustained hyperventilation is required for ICP
control.”
A UK study performed in 2002 found that despite the benefits of hyperventilation on ICP and
CPP, the volume of severely hypoperfused brain tissue increased (Coles et al., 2002). The
authors used PET scans to assess cerebral blood flow during hyperventilation and discovered
decreases in regional perfusion that are not detectable using global monitors of oxygen such as
saturation of jugular oxygen and arterial-venous differences in oxygenation. The authors
identified a threshold of 34 torr below which patients become vulnerable to regional
hypoperfusion. Future studies that more accurately assess patient outcomes are needed (Coles
et al., 2002)
A study by Diringer et al. (2000) found that early, brief, moderate hyperventilation does not
impair global cerebral metabolism in patients with severe TBI and, thus, is unlikely to cause
further neurological injury .The authors call for the assessment of more severe hyperventilation
and the effects of hyperventilation in the setting of increased ICP.
The findings of Thiagarajan et al. (1998) suggest that increasing the partial pressure of oxygen
above normal (hyperoxia) may also offset the deleterious effects of hyperventilation in head
injured patients. However, this study used a single group intervention design, and thus the
strength of such conclusions is limited until further studies using controlled randomized designs
are performed to corroborate these findings.
Conclusions
There is Level 2 evidence that the use of tromethamine, a weak base and buffer that crosses
the blood brain barrier, can offset the deleterious effects of prolonged hyperventilation and
lead to better outcomes than hyperventilation alone.
There is Level 4 evidence that hyperoxia can counteract the deleterious effects of
hyperventilation for the control of ICP following brain injury.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
There is Level 4 evidence that hyperventilation below 34 torr arterial CO 2 can cause an
increase in regionally hypoperfused tissue.
Tromethamine counteracts the detrimental effects of prolonged hyperventilation for
the control of ICP leading to better outcomes post-ABI.
Hyperoxia may counteract the adverse effects of prolonged hyperventiliation for the
control of ICP post-ABI.
Hyperventilation below 34 torr PaCO2 may cause an increase in hypoperfused brain
tissue.
16.1.1.4 Cerebrospinal Fluid Drainage
In an attempt to control ICP, ventricular CSF drainage is a frequently used neurosurgical
technique. Catheters are generally inserted in to the anterior horn of a lateral ventricle and
attached to an external strain gauged transducer (March 2005; Bracke et al., 1978). This allows
for concurrent pressure monitoring and fluid drainage. Generally, a few milliliters of fluid are
drained from the ventricle at a time resulting in an immediate decrease in ICP (Kerr et al.,
2000). However, ventricular space is often compressed due to associated brain swelling, which
limits the potential for drainage as a stand alone therapy for ICP control (James 1979).
When ventricular drainage is not possible, lumbar CSF drainage has been proposed as another
method for reducing elevated ICP. Standard practice has been to avoid lumbar drainage for fear
of transtentorial or tonsillar herniation, however, technological improvements have renewed
interest in its potential for reducing ICP in patients refractory to other treatments (Tuettenberg
et al., 2009).
Cerebrospinal fluid drainage use is predicated on the belief that increased ICP may result in
worse outcomes for ABI patients. However, empirical evidence regarding its direct effect on
improved outcomes is limited. It has been generally accepted that complications arising from
increased ICP are deleterious and that any intervention that improves ICP levels is therefore
worth exploring. Criticisms of ventricular drainage generally surround the intrusiveness of the
procedure and the complication of potential infections (Hoefnagel et al., 2008; Zabramski et al.,
2003).
In the most recent AANS guidelines no mention was made regarding indications for use of CSF
drainage. However if used, prophylactic antibiotic use and routine catheter exchange are not
recommended for reduction of nosocomial infection rates (Bratton et al., 2007d). CSF drainage
is listed as an acceptable treatment for ICP reduction according to the EBIC (Maas et al., 1997).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.4 Cerebrospinal Drainage for Treatment of Elevated ICP Post ABI
Author/Year/
Country/Study
design/PEDro
Scores
Methods
Outcome
Kerr et al.,
(2001)
USA
RCT
PEDro = 7
N=58 Severe TBI patients (GCS ≤ 8) were
randomly assigned to one of three
drainage protocols (1ml, 2ml, or 3ml
volume of CSF drained) each time an
intervention was required (ICP > 20
mmHg). Patients were monitored for ICP,
CPP, cerebral blood flow velocity, nearinfrared-spectroscopy-determined
regional cerebral oxygenation.
Physiological variables were time averaged
in 1-minute blocks from base-line to 10
minutes after drainage cessation.
Significant dose-time interactions were
seen in all three drainage protocols with
relation to decreases in ICP (p=0.0001).
There was a significant difference in CPP
depending on the amount of CSF
drained (p=0.04). A 3ml withdrawal of
CSF resulted in a 10.1% decrease in ICP
and a 2.2% increase in CPP that were
sustained for 10 min.
Murad et al.,
(2012)
USA
Pre-Post
N=15 All participants underwent lumbar
drainage when their ICP levels reached or
exceeded >20 mm Hg and medical
management of symptoms was complete.
Lumbar drains were placed approximately
3.5 days post admission.
Results indicate ICP was reduced
significantly following the placement of
the lumbar drain (p<0.01), CPP
increased, although not significantly,
from 76.7 mm Hg to 81.2 mm Hg. MAP
levels decreased from 96.8 mm Hg to
91.4 mmHg, again this decrease was not
found to be significant. When looking at
the administration of sedates, mannitol,
hypertonic saline, or paralysis, post
treatment only one individual required
additional boluses. Prior to treatment 12
of 15 required additional boluses.
Llompart-Pou et
al.,
(2011)
Spain
Case Series
(Retrospective
Review)
N=30 External lumbar drainage (ELD) was
used to reduce elevated ICP levels. ELD
placement was completed in those with
elevated ICP and ICP was not responding
to other measures or due to the risk of
cardiorespiratory failure other measures
were not an option. Once implanted, the
CSF was drained continuously when ICP
was 20 mm HG or higher. When ICP levels
dropped to 10 mm Hg the system was
closed and only reopened if ICP levels
began to increase and were once again
above 15 mm Hg.
ICP significantly decreased following ELD
(p<0.0001). Results of the GOS found 30
% of participants had a good recovery or
moderate disability following treatment
upon ICU discharge. Further long term
follow up noted 62% had a good
recovery or moderate disability. Study
authors reported few ELD complications.
Tuettenberg et al.,
(2009)
Germany
Pre-Post
N=100 Patients were prospectively
evaluated for changes in refractory ICP
after initiation of lumber CSF drainage.
Patient outcomes were also assessed 6
Initiation of lumbar CSF drainage led to
significant reductions in ICP (32.7 ± 10.9
to 13.4 ± 5.9 mmHg, p<0.05) and
increases in CPP (70.6 ± 18.2 to 86.2 ±
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Scores
Methods
Outcome
months after treatment.
15.4 mmHg, p<0.05). Cerebral
herniation with lethal outcome occurred
in 6 patients. Thirty-six had favorable
outcomes (on GOS), 12 were severely
disabled, 7 remained in a persistent
vegetative state, and 45 died.
Timofeev et al.,
(2008b)
UK
Pre-Post
N=24 TBI patients requiring mechanical
ventilation, neuromonitoring, and
ventriculostomy due to uncontrollable ICP
were monitored for ICP and related
parameters post drainage. Free drainage
was allowed and limited only by the height
of the collecting reservoir.
Drainage led to a decrease in ICP in all
patients with sustained reduction past
24 hrs in 13 of 24 patients. When ICP
reduction remained stable, significant
improvements in craniospacial
compensation, CPP, and PbtO2 were also
seen.
Murad et al.,
(2008)
USA
Case Series
N=8 Patients with brain injury and high ICP
refractory to medical management
underwent controlled lumbar drainage.
Patients were monitored for reductions in
ICP,
ICP levels were significantly reduced
(27±7.8 to 9±6.3 mmHg, p<0.05) after
drainage. In the 24 hrs post-drainage,
reductions were seen in the need for
hypertonic saline, mannitol, and
sedation. No complications were noted.
Kerr et al.,
(2000)
USA
Case Series
N=31 Severe TBI patients (GCS ≤ 8)
underwent cerebrospinal fluid drainage
where 6 ml of fluid was removed. Patients
were monitored for CPP, cerebral blood
flow velocity, and regional cerebral
oximetry before, during and after
drainage.
There was a significant change in ICP
immediately after drainage which
remained significant up until 10 minutes
post treatment (p=0.0001). A significant
increase in CPP was also seen
immediately after drainage but was not
maintained (p=0.0001).
Fortune et al.,
(1995)
USA
Case Series
N=22 Patients with a severe brain injury
(GCS < 8) were treated with
hyperventilation, intravenous mannitol, or
CSF drainage based on the attending
physicians discretion when ICP became >
15 mHg. Patients were continuously
monitored for SjvO2, ICP, BP, arterial O2
saturation, and end tidal CO2 during
treatment.
After drainage from ventriculostomy ICP
fell in 90% of the observations by
8.6±0.7 mmHg. In patients where ICP
dropped, SjvO2,only increased by
0.39±0.4% saturation
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In an RCT conducted by Kerr et al. (2001) a group of patients were randomized to one of 3
protocols. Those in the first had 1 ml of CSF drained, while those in the second had 2 ml of CSF
drained and those in the third had 3 ml drained. All underwent CSF drainage when their ICP was
found to be >20mm Hg. Continuous monitoring of ICP, CPP, cerebral blood flow velocity, and
near-infrared-spectroscopy-determined regional cerebral oxygenation was performed and
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
analyzed to assess the dose response to CSF drainage. Results indicate all experienced
significant decreases in ICP and increases in CPP regardless of the amount of fluid drained;
however, ICP levels began to increase toward within 10 minutes of treatment. Despite this
reduction in ICP, there were no improvements in cerebral blood flow velocity or regional
oxygenation. Also, the short duration of follow up limits the accurate measurement of the
effect of CSF drainage on ICP or any long term outcome measure.
In a recent study, Murad and his fellow researchers (2011) placed a lumbar drain in 15 patients,
following the medical management of all symptoms. In this study patients had sustained either
a TBI (n=10) and SAH (n=5). Following the placement of the lumbar drain, patients were found
to have a significant decrease in the ICP and a non-significant increase in their CPP. Post
treatment there was a significant decrease in the number of patients who required additional
boluses treatments (p<0.05). The need for sedative and paralytics was also decreased following
the insertion of the lumbar drain (p<0.05). Further no CSF infections were noted. Overall the
study found lumbar CSF drainage to be effective in lowering ICP post injury.
In a retrospective review of patient charts, Llompart-Pou et al. (2011) found that draining CSF
through external lumbar drainage was successful in improving ICP and long-term patient
outcomes. ICP decreased from 33.7 mm Hg pre treatment to 21.2 mm Hg post treatment. Initial
results from the GOS found 9 participants scored a 4 or 5 when discharged from the ICU. This
increased to 18 at the long-term evaluation.
In three case series, the effects of ventricular CSF drainage on ICP were also assessed. Timofeev
et al. (2008) found reductions in ICP were maintained at the end of the initial 24 hour period for
13 of the 24 particiapants, resulting in a reduction in the need for alternative treatments. Kerr
et al. (2000) treated 36 participants with CSF drainage after ICP levels reached >20 mmHg. They
noted a significant increase in CPP immediately after treatment and a decrease in ICP levels. Of
note those diagnosed with a subdural hematoma (SDH) ICP levels were not as likely to respond
to CSF drainage.
In the study conducted by Fortune and colleagues (1995), individuals who had sustained a
closed head injury were treated with mannitol, CSF drainage or hyperventilation with their ICP
levels exceeded 15mmHg. Physiological changes were documented continuously. Change noted
20 minutes post interventions was compared to the initial ICP, jugular venous saturation and
MAP levels. Although ICP levels decreased following treatment, fluid accumulation allowed ICP
to increase steadily once treatment was stopped. They also noted only slight improvements in
oxygenation based on jugular vein O2 saturation associated with the drop in ICP. These results
suggested only minimal improvements in cerebral blood flow (Fortune et al., 1995).
Two studies were located that evaluated the effects of lumber drainage on refractory ICP
(Murad et al., 2008; Tuettenberg et al., 2009). In the study conducted by Tuettenberg and
colleagues the safety and efficacy of lumbar drainage in patients with TBI (n=45) or SAH (n=55)
was evaluated. Study authors noted a significant reduction in ICP and a significant increase in
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
CPP (p<0.05) following lumbar drainage. Twelve patients experienced cerebral herniation with
unilateral mydriasis. Six of these 12 experienced lethal herniation. At the 6th month follow up,
results from the GOS indicated 36 patients had favorable outcomes, 12 were severely disabled,
7 remained in a persistent vegetative state and 45 had died (Tuettenberg et al., 2009). Study
authors concluded their findings with the suggestion that lumbar drainage should be
considered in instances when ICP is refractory to other treatments and basal cisterns are clearly
discernible.
In the previous year, Murad et al. (2008) also conducted a study investigating the effectiveness
of performing lumbar CSF drainage on a group of eight individuals who had sustained a severe
TBI. Study results indicate ICP was reduced in all participants and the use of hypertonic saline or
mannitol boluses was not needed to control ICP 24 hours post intervention. Based on these two
studies, lumber drainage appears to be a potential option when ICP remains refractory to other
interventions, ventricular drainage is not possible, and basal cisterns are clearly discernible.
Conclusions
Results from one RCT suggest there is Level 1b evidence that CSF drainage decreases
intracranial pressure in the short term.
There is Level 4 evidence from several studies that suggest CSF drainage does decrease ICP in
individuals post ABI.
CSF drainage has been found to reduce ICP and increase CPP in those who have sustained an
ABI.
16.1.1.5 Decompressive Craniotomy
Removal of skull sections has been suggested as a drastic measure for the management of
elevated ICP unresponsive to other therapies. It is thought that surgical decompression could
improve the damage caused by secondary injury (delayed brain damage) such as high ICP and
reduced oxygenation of the brain. In a recent meta-analysis by Sahuquillo and Arikan (2006),
the authors identified two types of surgical decompression: prophylactic or primary
decompression and therapeutic or secondary decompressive craniectomy. The former involves
performing the surgical procedure as a preventive measure against expected increases in ICP
while the latter is performed to control high ICP “refractory to maximal medical therapy”
(Sahuquillo & Arikan, 2006). However, debate regarding if and when to perform these
surgeries continues. Factors such as age and initial GCS score have been proposed as potential
prognostic factors (Guerra et al., 1999; Munch et al., 2000). Of course, any surgical procedure is
associated with inherent risks. The majority of decompressive techniques are therefore
precipitated by evacuation of a mass lesion (Compagnone et al., 2005). Once decompression is
decided upon, resection of a larger bone fragment is generally recommended to allow for
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
greater dural expansion with less risk of herniation (Compagnone et al., 2005; Skoglund et
al.,2006; Csokay et al., 2001) Therapeutic decompressive craniectomy is only performed after
other therapeutic measures (CSF drainage, moderate hypocapnia, mannitol, barbiturates,
hyperventilation, hypothermia etc.) have failed to control ICP (Morgalla et al., 2008).
The AANS make no recommendations regarding decompressive craniectomy in their most
recent recommendations. The EBIC suggest that decompressive craniectomy be considered in
“exceptional situations” (Maas et al., 1997).
Individual Studies
Table 16.5 Decompressive Craniectomy to Control Refractory Elevated ICP Post ABI
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
Adults and Decompressive Craniectomy Post TBI
Cooper et al.,
(2011)
Australia
RCT
PEDro=9
N=155 In this RCT participants were
randomly assigned to receive either
standard care (n=82) or decompressive
craniectomy (n=73) within the first 72 hours
of hospitalization. Participants received
treatment for intracranial hypertension if
intracranial pressure exceeded >20mmHg.
When second tier options were needed,
participants received mild hypothermia,
barbiturates or both. For those receiving
standard care, decompressive craniectomy
was used post >72 hours if needed. GOS
was used to assess participants 6 months
post intervention.
Although there was no significant
difference between the groups, those in
the decompressive craniectomy group
had both shorter duration of mechanical
ventilation and stay in ICU. Medical
complications and hydrocephalus were
noted more often in those in the
decompressive craniectomy group
compared to the standard care group.
Results from the GOS showed those in
the standard care group had better long
term outcomes then those in the
decomprssive craniectomy group. Of
note unfavourable outcomes were
noted in 70% in the decompressive
group. Only 51% had unfavourable
outcomes in the standard care group.
Qiu et al.,
(2009)
China
RCT
PEDro = 7
N=74 Severely brain injured patients (GCS ≤
8) with midline shift > 5 mm were
randomized to receive either unilateral
decompressive craniectomy (DC) or
unilateral routine temporoparietal
craniectomy (control).
ICP levels were significantly lower in the
DC group at 24, 48, 72 and 96 hours
post-operation in comparison to
controls. Mortality rates were reduced
from 57% to 27% in DC group with
respect to controls (p=0.01) and GOS
score 4 or 5 was 56.8% vs. 32.4% among
controls.
Jiang et al.,
(2005)
China
RCT
PEDro = 5
N=486 Adult patients with severe TBI (GCS
≤ 8) with refractory intracranial
hypertension were randomized to receive
standard trauma craniectomy (STC) with a
unilateral frontotemporoparietal bone flap
At 6 months follow up, more patients in
the STC group showed favourable (good
recovery/moderate disability) GOS
outcome compared with the LC group
(p<0.05). In addition, the incidence of
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
(12 x 15 cm) or limited craniectomy (LC)
with a routine temporoparietal bone flap (6
x 8 cm). In both groups, bone flaps were
removed and cranioplasty was performed
3-6 months after injury. GOS outcome at 6
months was compared between groups.
some secondary complications (delayed
intracranial hematoma, incisional
hernia, and CSF fistula) was lower in the
STC group (p<0.05). ICP fell more
rapidly and to a lower level in the STC
group than in the LC group (p<0.05).
Flint et al.,
(2008)
USA
Pre-Post
N=40 Computed Tomography scans for
patients with non-penetrating severe TBI
who underwent decompressive
hemicraniectomy were analyzed. Size of
contusions were measured on initial, last
pre-operative, and first post-operative
scans. Mortality and 6 month GOS scores
were recorded.
New or expanded hemorrhagic
contusions of ≥ 5cc were observed after
hemicraniectomy in 58% of patients.
The mean volume of increase
hemorrhage in these patients was 37.1
± 36.3 cc. Rotterdam CT score was
associated with total volume of
expanded contusion. Contusions
expanded >20cc post hemicraniectomy
were strongly associated with mortality
and poor outcome on the GOS at 6
months even after controlling for age
and initial GCS.
Howard et al.,
(2008)
USA
Database Review
N=40 A retrospective review of patient data
for patients with severe TBI managed with
decompressive craniectomy was
performed. Outcomes were measured
using the GOS-E.
DC effectively lowered ICP (p=0.005).
Twenty two patients died in hospital.
Initial GCS score and pupil reactivity
were associated with outcomes while
age and ISS score did not. Of the
survivors, 12 of 18 had good outcomes
on the GOSE.
Ho et al.,
(2008)
Singapore
Case series
N=16 Patients with isolated TBI and
elevated ICP refractory to maximal medical
therapy underwent decompressive
craniectomy. Patients were evaluated for
clinical outcomes using the GOS and were
then retrospectively divided into favourable
and poor outcome groups and compared
for ICP, CPP, pressure reactivity, cerebral
oxygenation, and cerebral microdialysis.
Only 5 patients had a favourable
outcome at 6-month follow up and one
made a good recovery. Significant
reductions in ICP and pressure reactivity
were seen in both groups. Patients with
favourable outcomes saw significant
improvements in oxygenation and a
reduction in cerebral ischemia
compared to no improvements in
biochemical indices for patients with
poor outcomes.
Aarabi et al.,
(2006)
USA
Retrospective
Chart Review
N=50 Patients with diffuse brain swelling
secondary to TBI who underwent
decompressive craniectomy without
removal of clots or contusion were
retrospectively evaluated for ICP reductions
and GOS > 3 months later.
Craniectomy reduced ICP below 20
mmHg in 85% of cases. Forty percent of
patients experience a good outcome on
discharge and 17/50 after 3 months.
Outcomes were independent of
abnormal papillary response, timing of
DC, brain shift, and patient age.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
Huang et al.,
(2008)
Taiwan
Retrospective
Chart Review
N=54 Patients with GCS ≤ 8, a frontal or
temporal hemorrhagic contusion greater
3
than 20 cm , and a midline shift > 4mm or
cisternal compression were studied. 16
underwent standard craniotomy with
hematoma evacuation and 38 underwent
craniectomy as the primary surgical
treatment. Mortality, reoperation rate,
GOS-E scores, and LOS were compared.
Reoperation rates (7.9% vs. 37.5%,
p<0.05) and GOS-E scores at 6 months
(5.55 vs. 3.56, p<0.01) improved in the
craniectomy group whereas LOS and
mortality was similar between groups.
Yang et al.,
(2008)
China
Retrospective
Chart Review
N=108 A retrospective review of patients
who underwent decompress craniotomy
was performed to establish the incidence of
secondary complications and associated
risk factors.
Twenty-five patients died within the
first month. Lower GCS was associated
with poorer outcomes. Complications
secondary to surgery occurred in 50% of
patients, 28% of which developed more
than one complication. Initially,
herniation through the cranial defect
was the most common complication
followed by subdural effusion.
“Syndrome of the trephined” and
hydrocephalus were common after 1
month. Older patients and more severe
injuries were associated with more
complications.
Li et al.,
(2008)
China
Non-RCT
N=263 Data from patients with severe TBI
(GCS≤8) was retrospectively reviewed to
compare large decompressive craniectomy
to routine craniectomy.
71.1% of cases in the large craniectomy
group obtained satisfactory outcomes
(GOS 3-5) compared to 58.6% of routine
craniectomy patients (p<0.05). Large
craniectomy was even more efficient in
treating very severe TBI (p<0.01). Large
craniectomy was also associated with a
lesser need for recurrent surgery and
fewer complications.
Salvatore et al.,
(2008)
USA
Chart Review
N=80 Patients with severe closed head
injury (GCS <8) who underwent selective
uncoparahippocamparectomy and tentorial
edge incision with wide decompressive
craniectomy were reviewed. GOS was
measured on follow-up (mean 30 months).
75% of patients had a favourable
outcome. Younger age, and earlier
operations were associated with better
outcomes. Preoperative GCS and
papillary reactivity had no effect on
outcome.
Olivecrona et al.,
(2007)
Sweden
Chart review
N=93 All patients treated for severe TBI
during 1998-2001 who had a GCS<9 at
intubation and sedation, first recorded CPP
>10mmHg, arrival within 24 h of trauma,
and in need of intensive care for >72 h
were included. Craniectomy was performed
Craniectomy patients showed a
decrease in average ICP from 36.4
mmHg to 12.6 mmHg directly after the
procedure. There was an increase in ICP
to 20 mmHg 8-12 h after surgery
leveling off at around 25 mmHg within
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
when the ICP could not be controlled by
evacuation of hematomas, sedation,
ventriculostomy, or low-dose pentothal
infusion. Twenty-one patients underwent
craniectomy and were compared to the
non-craneiectomy group for ICP, CPP and
GOS scores.
72 h. The GOS scores did not
significantly differ from the noncraniectomy patients.
Skoglund et al.,
(2006)
Sweden
Chart review
N=19 Charts for all patients receiving
decompressive craniectomy between 1997
and 2002 were drawn. Patients were
assessed for ICP decrease, survival rates,
and GOS scores. The size of craniectomy
was also assessed for its relation to ICP
decrease.
ICP was reduced from 29.2±3.5 to
11.1±6.0 mmHg after decompression
and 13.9±9.7 mmHg 24h after surgery
(both p<0.01). Sixty eight percent of
patients had favorable outcomes at
least one year post surgery. A significant
correlation was seen between the size
of craniectomy and decrease in ICP.
Ucar et al.,
(2005)
Turkey
Chart review
N=100 Patients with severe brain injury
(GCS<9) who received decompressive
craniectomy were retrospectively divided
into two groups; groups I (GCS 4-5,N=60)
and group II (GCS 6-8,N=40). Prognosis was
evaluated based on GOS scores at 6 months
and a backwards regression analysis was
used to assess age, GCS, timing of surgery,
and the presence of mass lesions as
indicators.
After regression analysis, only age
(p=0.046) and belonging to GCS group II
(p<0.05) were significantly related to
having a “favorable” final outcome (GOS
4-5). A paired t-test showed a significant
decrease in ICP after decompression
from 29.8±5.2 to 23.9±4.9 mmHg
(p<0.001).
Morgalla et al.,
(2008)
Germany
Case Series
N=33 Patients with severe TBI (Grades III
and IV) and subsequent swelling underwent
decompressive craniectomy. Patients were
assessed three years post surgery using the
Barthel Index.
Twenty percent of patients died and
20% remained in a vegetative state.
Thirteen of the surviving patients made
a full recovery (BI 90-100). Five patients
returned to a previous job and 4 found
new work.
Williams et al.,
(2009)
USA
Retrospective
Chart Review
N=171 Patients with severe TBI (AIS 4-5)
treated with decompressive craniectomy
were identified from the Trauma Registry.
Patients were assessed for GOS-E scores
and mortality rates on follow-up. Multiple
regression was used to identify factors
associated with good outcomes.
32% of patients died in hospital. Of the
survivors, 82% achieved good outcomes
(5-8 on the GOS-E) Patients who
experienced good outcomes were
younger (26 vs. 43, p=0.0028) and
experienced a greater reduction in ICP
post-surgery (23mmHg vs. 10 mmHg,
p<0.0001). Immediate
predecompression GCS in survivors was
higher (7 vs. 5, p<0.0001).
Polin et al.,
(1997)
USA
N=35 Patients undergoing decompressive
craniectomy were each matched to 4
patients from the Traumatic Coma Data
Postoperative ICP was significantly
lower in craniectomy patients compared
to preoperative levels (p=0.0003). A
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
Cohort control
Bank as a control based on sex, age, GCS
scores and maximum preoperative ICP.
Mortality rates and discharge GOS scores
were compared between groups.
significant increase in favorable
outcomes was seen in patients who
underwent craniectomy compared to
the matched controls (p=0.014).
Patients who did not exhibit ICP >
40Torr and underwent surgery less than
48 h after injury revealed a 60%
favorable outcome rate and did better
than control patients (p=0.0001).
Meier et al.,
(2008)
Germany
Chart Review
N=131 Clinical records for patients treated
with decompressive craniectomy were
reviewed. Outcomes were compared to
patient and radiographic factors to identify
trends.
Outcomes were correlated with initial
GCS as well as the patients age,
condition of the basal cisterns, the
degree of midline shift, pupil reactivity
on admission, established clotting
disorders, post-traumatic hydrocephalus
internus, hyperglycemia and initial
acidosis.
Bao et al.,
(2010)
China
Retrospective
cohort
N=37 Patients who had received a bilateral
decompressive craniectomy (BDC) were
retrospectively reviewed. Patients were
between the ages of 18-69 and 67.6% were
male. Pre and post surgical ICP and CPP
were compared and 6-month GOS was
assessed.
Mean ICP was reduced from 37.7 +/- 7.5
mm Hg pre surgery to 27.4 +/- 7.3 mm
Hg (p<.05) after bone removal and 11.2
+/- 7.1 mm Hg (p<.05) after dura mater
opening and enlargement. Mean ICP
was 16.3 +/- 5.9 mm Hg 1 day post
surgery, 17.4+/- 6.3 mm Hg 3 days post
surgery and 15.5 +/- 4.6 mm Hg at 7
days post surgery. Mean CPP was
increased from 57.6 +/- 7.5 mm Hg pre
surgery to 63.3 +/- 8.4 mm Hg (p<.05)
after bone removal and 77.8 +/- 8.3 mm
Hg (p<.05) after durameter opening and
enlargement. At 6-months, 54.1% of
patients made moderate (GOS=4,
32.5%) or good (GOS=5, 21.6%)
recoveries on the GOS.
Daboussi et al.,
(2009)
France
Prospective
Observational
N=26 Patients undergoing decompressive
craniectomy were prospectively tracked.
Neurological outcomes and mortality rates
were evaluated.
ICP was reduced from 37 +/- 17 to 20 +/13 mm HG (P=.0003) and mean cerebral
perfusion pressure was increased from
61 +/- 22 to 79 +/- 19 (P<.05)
immediately post surgery and remained
significant 48 hours post. Middle
cerebral artery blood flow velocity was
also significantly increased as measured
by Transcranial Doppler
Ultrasonoghraphy. Mortality rate was
27% and among those that survived 53%
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
had favourable neurologic outcomes.
Aarabi et al.,
(2009)
USA
Retrospective
Cohort
N=54 All patients undergoing DC along with
removal of a mass lesion were
retrospectively divided into 2 groups for
comparison; patients who had DC
performed without ICP monitoring (n=27,
Group A) and patients underwent ICP
monitering from day 1 to 14 before DC
(n=27, Group B).
No difference was noted between
groups for survival or outcome. Twelve
patients died in group A and 10 in group
B, while 11 had good recovery in group
A and 8 in group B.
Adamo et al.,
(2009)
USA
Case Series
N=7 Charts were reviewed of infants (2-24
months) with a severe TBI who underwent
a decompressive craniectomy.
All patients developed epidural and
subural empyemas that required
debridement and surgical drainage.
KOSCHI scores at 1-year follow up
ranged from 3b to 4b.
Timofeev et al.,
(2008a)
UK
Retrospective
chart review
N=27 Charts of moderately to severely
brain injured patients who underwent
decompressive craniectomy were reviewed.
Reductions in ICP immediately post surgery
as well as pressure reactivity were
evaluated.
Mean ICP levels were reduced from 21.2
mm Hg pre-operation to 15.7 mm Hg
post-operation (P=0.01). ICP exceeded
25 mm Hg 28.6% of the time preoperation and only 2.2% of the time
post-operation. Pressure reactivity
post-surgery was significantly associated
with favourable outcome (GOS 4 or 5) 6months post (p=0.02) but pre-surgical
reactivity was not (p=0.462)
Otani et al.,
(2010)
Japan
Cohort
N=80 Patients with acute epidural
hematomas were treated with either
hematoma evacuation (HE) or HE as well as
an external decompression (ED).
Significant differences in patient
characteristics were noted between
groups. A favorable outcome was noted
in 78.2% of patients in the HE group,
while only in 55.8% of the HE plus ED
group.
Children and Decompressive Craniectomy Post ABI
Taylor et al.,
(2001)
Australia
RCT
PEDro = 6
N=27 Children >12 who had sustained a
TBI and had a functioning intraventricular
catheter, and sustained intracranial
hypertension during the first day after
admission (ICP 20-24 mmHg for 30 min,
25-29 mmHg for 10 min, 30 mmHg or
more for 1 min) or had evidence of
herniation (dilation of one pupil or the
presence of bradycardia) were
randomized to conventional medical
management (control group) or
decompressive craniectomy plus
Significantly lower ICP in the
decompression group following
craniectomy compared with control group
(p=0.057). There was a trend towards
shorter time in intensive care in the
decompression group than in the control
group (p=0.12). The median stay in
hospital was 26.8 days (range 13.8-73.3) in
the decompression group and 47.7 days
(range 21.9-73.1) in the control group
(p=0.33) More children in the
decompression group obtained a
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/
Study design/
PEDro Scores
Methods
Outcome
conventional medical management
favourable outcome 6 months after the
(decompression group). A decompressive injury on the GOS and HSU compared with
bitemporal craniectomy was performed the control group (p=0.046)
at a median of 19.2 hr (range 7.3-29.3 hr)
from the time of injury. Control of ICP
and GOS and the Health State Utility
(HSU) index (Mark 1) 6 months after
injury were assessed.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In a recent RCT, Cooper et al. (2011) assigned a group of individuals to receive either
decompressive craniectomy or standard care (n=73) or standard care alone (n=82). Six months
after recovery all were evaluated using the GOS. Those in the decompressive group had a
shorter stay in ICU and did not require mechanical ventilation as long as those receiving
standard care. However despite this according to the results of the GOS, those in the
decompressive group had worse outcomes. Fifty-one of the 73 participants had unfavorable
outcomes compared to 42 of the 82 in the standard care group. Overall study authors did not
find, as has been previously reported, decompressive craniectomy was effective in reducing
poor outcomes in participants.
In two separate adult RCTs, Jiang et al. (2005), and Qiu et al. (2009) assessed randomly assigned
patients with refractory intracranial hypertension to receive standard trauma craniectomy with
a unilateral frontotemporoparietal bone flap (12 x 15 cm) or limited craniectomy with a routine
temporoparietal bone flap (6 x 8 cm). In both studies, the authors reported that significantly
more patients in the standard (larger) craniectomy group showed favourable GOS outcomes
than those who received limited craniectomy at 6-months (Jiang et al., 2005) and 1-year post
surgery (Qiu, 2009). Moreover, ICP fell more rapidly and to a lower level in the standard
craniectomy group than in the limited craniectomy group and one month mortality was also
reduced (Qiu, 2009).
In the first of three non-randomized control trials, Polin et al. (1997) used patients from the
Traumatic Coma Data Bank who had not received decompressive craniectomy as a matched
control for 35 patients who had been surgically decompressed. Averages of four control
patients (matched for age, sex, preoperative GCS and maximum preoperative ICP) were used as
controls for each decompressed patient. Operative patients showed decreased ICP post-surgery
and improved ICP compared to late stage control measures. The rate of favorable outcome in
the decompressed group was significantly higher than the control group. Patients who received
decompression before ICP levels reached 40 Torr and within the first 48 hours post trauma
showed the most significant improvement compared to controls. Polin et al., (1997) suggest
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
that decompressive craniectomy be considered as early as possible when routine ICP
management measures fail.
In the second non-RCT, Li et al. (2008) retrospectively reviewed clinical data for 263 patients
treated with either large decompressive craniectomy (n=135) or routine craniectomy (n=128).
They report that patients who underwent larger resections showed significantly higher rates of
satisfactory outcomes on the GOS. Similar results were seen for severe TBI patients. Patients
who underwent larger craniectomies were also less likely to require further operations and
presented fewer complications.
In the third non-RCT, Aarabi et al. (2009) retrospectively assessed differences in outcome on
the GOS between patients who had ICP monitoring 1-14 days before DC and those who did not.
They found no difference between the two groups.
In all, 18 chart reviews were located. Of these reviews, nine explicitly reported decreases in ICP
associated with decompressive craniectomy (Howard et al., 2008; Aarabi et al., 2006;
Olivecrona et al., 2007; Skoglund et al., 2006; Ucar et al., 2005; Tuettenberg et al., 2009; Ho et
al., 2008; Bao et al., 2010; Daboussi et al., 2009; Timofeev et al., 2008). Of these nine, only 1
reported an association between decreased ICP and improved outcomes (Williams et al., 2009).
Other factors that were reported to correlate with long-term outcomes included: initial GCS;
(Meier et al., 2008; Ucar et al., 2005; Yang et al., 2008; Howard et al., 2008) age; (Salvatore et
al., 2008; Ucar et al., 2005; Williams et al., 2009; Meier et al., 2008) pupil reactivity; (Howard et
al., 2008; Meier et al., 2008) and earlier operations (Salvatore et al., 2008). Larger
decompressions were also associated with better reduction in ICP (Skoglund et al., 2006) and
decompressive craniectomy was reported to be more effective for improving outcomes after
removal of hemorrhagic contusions than craniotomy (Huang et al., 2008).
In a more recent study Salvatore et al. (2008) performed uncoparahippocampectomy with
tentorial edge incision followed by decompressive craniectomy on 80 patients and reported
75% favorable outcomes. They also suggested that younger age and earlier operations were
associated with better outcomes.
In a well designed observational study, Flint et al. (2008) evaluated the effects of post-operative
expansion of hemorrhagic contusions on patient outcomes. They report that new or expanded
contusions >5cc were found in 58% of patients after decompressive hemicraniectomy and that
contusions greater than 20cc were significantly associated with poorer outcomes and mortality.
This represents a serious concern that must be monitored in patients undergoing
decompressive procedures.
A 2006 Cochrane review found no evidence to recommend routine use of decompressive
craniectomy to reduce unfavorable outcomes in adults with uncontrolled ICP (Sahuquillo &
Arikan, 2006). However, they do recommend that decompressive craniectomy may be a useful
option in the pediatric population when maximal medical treatment has failed to control ICP.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
In a study of pediatric TBI patients Taylor et al. (2001), children with intracranial hypertension
were randomized to receive decompressive craniectomy or to continue conventional medical
management (control group). The authors reported that children who received craniectomy
showed significantly lower ICP and more achieved a favourable outcome on the GOS and the
Health State Utility index at 6 months
Conclusions
There is Level 1b evidence that in adults, standard trauma craniectomy is more effective than
limited craniectomy in lowering elevated ICP and leading to better GOS outcomes at 6
months.
There is conflicting evidence supporting the use of decompressive craniectomies in adults post
TBI.
There is Level 3 evidence that resection of a larger bone flap results in greater decreases in
ICP reduction after craniectomy, better patient outcomes and leads to fewer post-surgical
complications.
There is Level 1b evidence that in children, decompressive craniectomy reduces elevated ICP.
There is Level 4 evidence from several studies does reduce ICP in children post severe TBI.
In adults standard trauma craniectomy leads to better control of ICP and better clinical
outcomes at 6 months when compared with limited craniectomy.
Resection of a larger bone flap during craniectomy may lead to a greater reduction in
ICP, better patient outcomes and fewer post-surgical complications.
Although decompressive cranectomy does reduce ICP in children more research needs
to be conducted investigating its impat on the long term clinical outcomes.
16.1.1.6 Continuous Rotational Therapy and Prone Positioning
The concept of continuous rotational therapy has been used for the prevention of secondary
complications resulting from immobilization. These include bedsores and ulcers, pneumonia,
atelectasis, deep vein thrombosis, pulmonary emboli, muscle atrophy, contracture and others.
There are some indications that continuous rotational therapy may be useful in managing
elevations in intracranial pressure. We identified a single study that evaluated the efficacy of
this intervention post ABI.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Use of the prone position has been shown to be an effective treatment for patients with acute
respiratory insufficiency in the ICU (Pelosi et al., 2002). However, many studies have excluded
patients with ABI due to fears of increasing ICP during the turning process and in the prone
position itself (Johannigman et al., 2000). We found two studies that showed increased
oxygenation in the prone position in patients with TBI and reduced intracranial compliance.
Neither the EBIC nor the AANS make any recommendations regarding continuous rotational
therapy or prone positioning.
Individual Studies
Table 16.6 Continuous Rotational Therapy and Prone Positioning in Acute Care Management Post ABI
Author/Year/
Country/Study
design
Methods
Outcome
Tillet et al.,
(1993)
USA
Case Series
N=58 Severe brain injury patients (GCS ≤
9) received continuous rotational
therapy within 24 hours of injury onset
with side-to-side rotation maintained at
40º (rotation of body to prevent
pressure ulcers and other
complications). Differences in ICP during
rotation and non-rotation periods were
compared.
No significant differences in the ICP
during rotation compared to nonrotation periods during any of the time
periods examined (day 2, 3, 4 and 5
post-admission). Highest ICP reported
in patients with unilateral injuries when
rotated to the side of their lesions.
These findings showed a statistically
significant difference on ICP when
patients were rotated to the side of the
lesion (p = 0.025).
Nekludov et al.,
(2006)
Sweden
Case Series
N=8 Patients with TBI (5 with GCS ≤ 8) or
Sub-arachnoid Hemorrhage were
studied in the Supine and Prone
positions. Hemodynamics, arterial
oxygenation, respiratory mechanics, ICP
and CPP were measured.
A significant improvement in arterial
oxygenation was observed in the prone
position (P=0.02). Both ICP (P=0.03) and
MAP (P=0.05) increased in the prone
position as well. MAP however,
increased to a greater extent resulting in
improved CPP (P=0.03).
Thelandersson et al.,
(2006)
Sweden
Case Series
N=11 Patients admitted to the neuro
intensive care unit due to TBI or
intracerebral hemorrhage. Patients
were monitored for ICP, CPP, HR, mean
arterial blood pressure (MABP), arterial
partial pressure of oxygen (PaO2 )and
carbon dioxide, and respiratory system
compliance. Measurements were taken
before, three times during and two
times after being placed in the prone
position.
No significant changes were
demonstrated in ICP, CPP or MABP.
PaO2 and SaO2 were significantly
increased in the prone position and after
10 minutes in the supine post-prone
position (P<0.05). Respiratory system
compliance was increased 1h in the
supine post-prone position (P<0.05).
Discussion
We identified a single study examining the effects of continuous rotational therapy on
intracranial pressure. This study utilized a single group pre-post intervention comparative
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
design lacking a true control group (Tillett et al.,1993). This study failed to find any direct
benefit of continuous rotational therapy for the management of elevated ICP. However,
continuous rotational therapy did not worsen ICP. Study authors suggest that care should be
taken not to rotate patients with unilateral brain injuries towards the side of the lesion to avoid
further increments in ICP.
Two studies, Thelandersson et al., (2006) and Nekludov et al., (2006), discussed the effects of
the prone position on oxygenation rates in ABI patients. Both studies were case series’ with no
control group using a prospective pre-post design. Both studies showed increased pO2 levels in
the prone position. However, only one study showed increased oxygenation after return to the
supine position (Thelandersson et al., 2006).The other study showed increased CPP associated
with increased MAP while in the prone position (Nekludov et al., 2006). Of note, in this study,
ICP also significantly increased in the prone position. Due to the small numbers associated with
both papers, the authors recommend further studies be performed to verify the efficacy of
prone positioning.
Conclusions
There is Level 4 evidence that continuous rotational therapy does not worsen intracranial
pressure in severe brain injury patients.
There is level 4 evidence that the prone position may increase oxygenation and CPP in ABI
patients with acute respiratory insufficiency.
Continuous rotational therapy may not worsen intracranial pressure in severe brain
injury patients
Prone position may increase oxygenation and cerebral perfusion pressure in patients
with acute respiratory insufficiency.
16.1.2 Pharmacological Treatments
16.1.2.1 Osmolar Therapies
Osmolar therapy is a major treatment approach in controlling intracranial hypertension and
edema following an ABI. Although mannitol is the drug most widely used in this regard,
hypertonic saline has gained popularity and some studies have called for examination of
hypertonic saline as a primary measure for ICP control (Horn et al., 1999; Ware et al., 2005).
16.1.2.1.1 Hypertonic Saline
Hypertonic saline exerts its effect mainly by increasing serum sodium and osmolarity thereby
establishing an osmotic gradient. This allows water to passively diffuse from the cerebral
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
intracellular and interstitial space into blood capillaries causing a reduction in water content
and a subsequent reduction in ICP (Khanna et al., 2000). Although mannitol works similarly,
sodium chloride has a better reflection coefficient (1.0) than mannitol (0.9) (Suarez 2004)
meaning that the blood-brain barrier is better able to keep out sodium chloride, making it a
more ideal osmotic agent. It has also been proposed that hypertonic saline normalizes resting
membrane potential and cell volume by restoring normal intracellular electrolyte balance in
injured cells (Khanna et al., 2000).
The AANS or EBIC made no recommendations for the use of hypertonic saline.
Individual Studies
Table 16.7 Hypertonic Saline for the Management of ICP Hypertension Post ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Adult Studies
Rhind et al.,
(2010)
Canada
RCT
Pedro = 7
N=65 TBI patients were randomized to
receive either 250mL of 0.9% normal
saline (NS) or an equal dose of 7.5%
hypertonic saline mixed with 6% dextran70 (HSD) pre-hospital and then compared
to each other and normal controls for
immunologic and coagulation markers.
The authors note a number of
differences between NS and HSD, &
note that in many cases, HSD is
beneficial in modulating inflammatory
response, and may help ameliorate
secondary brain injury.
Baker et al.,
(2009)
Canada
RCT
Pedro = 10
N=64 Severe head injured patients were
randomized to receive either a 250 mL
intravenous infusion of 7.5% hypertonic
saline in 6% dextran 70 (HSD) or an
equivalent dose of 0.9% isotonic normal
saline (NS) and compared to control
patients. Variation in levels of several
serum biomarkers was conducted.
No differences in patient survival or
neurocognitive outcomes were noted
between groups. However, peak levels
of biomarkers were significantly
correlated with unfavorable outcomes
measured by the GOS and GOSE.
Myburgh et al.,
(2007)
NZ/Aus
RCT
PEDro = 10
N=460 Severely brain injured patient
(GCS3-12) information was accessed from
a previous study of saline vs albumin in a
heterogeneous population of ICU patients.
Mortality and 24 month GOS-E scores
were used as outcome measures.
Overall, there was a significant increase
in mortality rates in the albumin group
(p=0.003) after 24 months. Also, there
was a significant increase in mortality
rates in patients with severe TBI (GCS<9)
treated with albumin (p<0.001). No
differences were seen between groups
on GOS-E scores.
Battison et al.,
(2005)
UK
RCT
PEDro = 5
N=9 Brain injury patients who required an
ICP monitor as part of their management
due to ICP > 20 mm Hg for > 5 min which
was not related to a transient external
noxious stimulus or systemic derangement
Both mannitol and HSD were effective in
reducing ICP, however HSD caused a
significantly greater decrease in ICP than
mannitol (p=0.044). HSD had a longer
duration of effect than mannitol
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
+
Outcome
(low serum Na , low PaO2, increased
PaCO2) received equimolar rapid
intravenous infusions of either 200 mL of
20% mannitol or 100 mL of 7.5% saline
and 6% dextran-70 solution (HSD) over 5
min in a cross over fashion. The order of
the treatments was randomized.
(p=0.044).
Cooper et al.,
(2004)
Australia
RCT
PEDro = 9
N=229 Severe brain injury patients (GCS <
9) were randomized to receive 250 mL
intravenous infusion of either 7.5%
hypertonic saline (HTS) or 250 mL of
Ringer’s lactate solution (controls) in
addition to standard intravenous
resuscitation fluids. GOS score, FIM and
Rancho Los Amigos score at 3 and 6
months post-injury were compared
between groups.
Proportion of patients surviving was
similar in both groups at hospital
discharge (55% in HTS group and 50% in
control, p=0.32) and at 6 months postinjury (55% in HTS group and 47% in
control, p=0.23). There were no
significant group differences at 6
months in median (interquartile range)
GOS score (HTS group 5 (3-6) vs. control
group 5 (5-6), p=0.45). There was no
significant difference between groups in
favorable outcomes (GOS 5-8, p=0.96)
or in any other measure of post-injury
neurological function.
Vialet et al.,
(2003)
France
RCT
PEDro = 7
N=20 Severe head injury patients (GCS < 8)
who experienced persistent coma
requiring ICP monitoring and infusion of
an osmotic agent to correct refractory
episodes of ICP that were resistant to
standard modes of therapy were
randomized to either 20% mannitol (1160
mOsm/kg/H20) or 7.5% hypertonic saline
(2400 mOsm/kg/H2O). Infused volume
was the same for both medications (2
ml/kg body weight in 20 min). Number
and duration of episodes of ICP
hypertension per day, failure rate of each
treatment (defined as the persistence of
ICP hypertension despite two successive
infusion of the same osmotic agent),
mortality rate and GOS score at 90 days
were compared.
The mean number (6.9 ± 5.6 vs. 13.3 ±
14.6 episodes) of ICP hypertension
episodes per day and the daily duration
(67 ± 85 vs. 131 ± 123 min) of ICP
hypertension episodes were significantly
lower in the hypertonic saline group
(p<0.01). The rate of clinical failure was
also significantly lower in the hypertonic
saline solution group (1 of 10 patients
vs. 7 of 10 patients, p<0.01). Mortality
rate and GOS outcome did not differ
between the two groups.
Shackford et al.,
(1998)
USA
RCT
PEDro = 5
N=41 ABI patients (GCS < 13) who
required ICP monitoring were randomly
assigned to receive 1.6% hypertonic Saline
(HTS) or Lactated Ringer’s Solution (LRS).
Differences in ICP, number of medical
Interventions to lower ICP, and GOS at
Treatment lowered ICP in both groups,
and there were no significant
differences between groups in ICP at
any time after entry. The average total
number of interventions to control
elevations in ICP during the entire study
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
discharge were compared between
groups. Groups were similar in age and
Injury Severity Score. HTS patients had a
lower admission GCS score (p=0.057), a
higher initial ICP (p=0.06), and a higher
initial mean maximum ICP (p <0.01)
compared with the LRS group.
was significantly greater in the HTS
group than in the LRS group (p < 0.01).
During the study, the change in
maximum ICP was positive in the LRS
group but negative in the HTS group
(p<0.05) There were no significant
differences between groups in the mean
GOS at discharge from hospital.
Eskandari et al.,
(2013)
USA
Prospective
Cohort
N=11 Participants were administered a
14.6% hypertonic saline (HTS) bolus over
15 minutes through a cent At the time of
infusion of the HTS, ICP, CPP HR and SBP
were recorded. Each bolus administration
was considered an individual exposure.
Repeated boluses were administered if,
after 60 minutes the patient was found to
have refractory ICP once again.
Following the administration of bolus
HTS participants ICP levels were reduced
from 40 ± 12 mm Hg to 33 ± 10 mm Hg
(p<0.05). Tem minutes post
administration there was a further
reduction to 28 ± 9 mm Hg (p<0.05).
Similar trends were noted with looking
at cerebral perfusion pressure
measurements. Again these trends were
found to be significant (p<0.05) Mean
heart rates and systolic blood pressure
were not affected by the HTS infusions.
Kerwin et al.,
(2009)
USA
Cohort
N=22 Severely brain injured patients were
administered 23.4% hypertonic saline or
mannitol to treat acute intracranial
hypertension. Patient charts were
retrospectively reviewed for reductions in
ICP immediately after administration.
Mean reduction of ICP by hypertonic
saline was 9.3 ± 7.37 mm Hg and 6.4 ±
6.57 mm Hg by mannitol. Patients
receiving mannitol were more likely to
have an ICP reduction of <5 mm Hg
whereas hypertonic saline treated
patients were more likely to have a
decrease of greater than 10 mm Hg.
However, patients receiving hypertonic
saline had statistically significantly
greater ICP immediately before
administration.
Oddo et al.,
(2009)
USA
Chart review
N=12 Patients treated with mannitol (25%,
0.75 g/kg) for episodes of elevated ICP or
hypertonic saline if not controlled by
mannitol were retrospectively evaluated.
PbtO2, ICP, MAP, CPP, central venous
pressure, and cardiac output were
monitored.
Forty-two episodes were analysed.
Compared with mannitol, hypertonic
saline was associated with lower ICP,
and higher CPP and cardiac output.
Hypertonic saline also resulted in
significant improvements in PbtO2
(p<0.01) while mannitol did not.
Schatzmann et al.,
(1998)
Germany
Case Series
N=6 Severe ABI patients (GCS ≤ 7) received
hypertonic saline (100 ml NaCl
intravenously over 5 min) after standard
agents (mannitol, sorbitol, THAM) failed in
reducing ICP. Changes in ICP were
Following hypertonic saline infusions
relative ICP decreased by an average of
43%. The corresponding pressure drop
was 18 mmHg (15-17mmHg).
Relaxation in ICP lasted for 93 min (64-
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
assessed.
126 min) and relative ICP minimum was
reached 26 min (12-33 min) after
infusion. No p values provided.
Qureshi et al.,
(1998)
USA
Chart Review
N=10 Head injury patients (GCS = 7.1 ±
0.8) received 3% hypertonic saline/acetate
solution composed of sodium chloride and
sodium acetate (50:50) at a variable rate
(75 – 150 mL/hr through a central venous
catheter). Changes in ICP, GCS and GOS
score at 1 month were evaluated.
A favorable trend toward ICP reduction
correlating with increasing serum
sodium concentration was observed (r2
= 0.91, p = 0.03). Mean ICP was reduced
from 14.2 ± 4.2 to 7.3 ± 1.8 mmHg (no p
value provided). There was reduction in
lateral displacement of the brain (2.8 ±
1.4 to 1.1 ± 0.9 mm). GCS showed a
slight improvement to 7.4 ± 1.3
compared with 7.1 ± 0.8 before
treatment (no p value). GOS 1 month
post treatment was 3.6 ± 0.4.
Pascual et al.,
(2008)
USA
Chart Review
N=12 Severe hypotensive TBI patients
(GCS ≤ 8) in ICU received hypertonic saline
(HTS) as a method for resuscitation under
a guideline regulated protocol over a
course of 3.5 years. Changes in ICP, CPP,
hemodynamics, and PbtO2 were
monitored.
HTS was associated with a significant
trend in decreased ICP and increases in
CPP and Pbt O2. Patients were only
administered HTS as an adjunct to
standard protocol as a means of
resuscitation. Only 6 patients survived
to rehabilitation.
Lescot et al.,
(2006)
France
Case Control
N=14 TBI patients (GCS 4-14) were
administered a 20 min infusion of 40 mL of
20% Saline. A CT scan was performed
before and after to assess the volume,
weight and specific gravity of contused
and non-contused brain tissue.
HTS significantly increased natremia
from 143 ± 5 mmol/L to 146 ± 5 mmol/L
and decreased ICP from 23 ± 3 to 17 ± 5
mmHg. The Volume of noncontused
hemispheric areas decreased by 13 ± 8
mL whereas the specific gravity
increased by 0.029 ± 0.027%. The
volume of contused hemispheric tissue
increased by 5 ± 5 mL without any
concomitant change in density. There
was wide individual variability in the
response of the noncontused
hemispheric tissue with changes in
specific gravity varying between 0.0124% and 0.0998%.
Ware et al.,
(2005)
USA
Chart Review
N=13 Brain injured patients who had been
administered 23.4% hypertonic saline in
conjunction with mannitol to control ICP.
Charts were reviewed for ICP, CPP, MAP,
serum sodium values, and serum
osmolarity after treatment with 23.4%
sodium chloride and mannitol.
Hypertonic saline was shown to
significantly reduce ICP (P<0.001) and
there was no significant difference
between ICP reduction by hypertonic
saline and mannitol (P=0.174). The
mean duration of ICP reduction by HTS
(96 min) was significantly longer than
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
mannitol (59min) (P=0.016).
Horn et al.,
(1999)
Germany
Case control
N=10 Patients with TBI or SAH and
therapy-resistant elevation of ICP were
given 7.5%, 2 ml/Kg b.w. intravenously at
an infusion rate of 20 ml/min. Only
patients with ICP > 25mmHg and plasma
sodium concentration <150 mmol/L were
included.
Within the first hour after HSS
administration ICP decreased from 33 ±
9 mmHg to 19 ± 6 mmHg (p< 0.05) and
further to 18 ± 5 mmHg at the time of
maximum effect. CPP rose from 68 ± 11
mmHg to 79 ± 11 mmHg after one hour
(p<0.05) and 81 ± 11 mmHg at
maximum effect.
Rockswold et al.,
(2009)
USA
Prospective
N=25 Patients with severe TBI were
treated with 23.4% NaCl (30mm over 15
mins). Patients were evaluated on
intercranial pressure (ICP), mean arterial
pressure (MAP), cerebral perfusion
pressure (CPP) and brain tissue oxygen
tension (PbtO2). For each evaluation,
patients were divided into either a low,
medium or high risk group.
Mean ICP decreased by 8.3 mm Hg
(P<0.0001). ICP level was positively
effected for all 3 groups (P<0.05) The
higher the initial ICP value, the more
effective NaCl was at reducing ICP
(P<0.05). The best MAP response was in
the groups with the lowest initial MAP
values. Most patients MAP levels were
not affected by hypertonic saline.
Patients above 70 mm Hg initial CPP
level were unaffected by treatment,
whereas patients under 70 mm Hg were
shown to have significant improvement
after treatment (P<0.05). PbtO2 was
significantly improved at 1,3, and 4
hours post-treatment, and nearly
reached significance 2 and 5 hours
(p=0.06).
Pediatric Studies
Simma et al.,
(1998)
Switzerland
RCT
PEDro = 6
N=32 Children (< 16 yrs of age) with
severe traumatic brain injuries (GCS < 8)
were randomly assigned to Ringer’s
Lactate (sodium 131 mmol/L, 277
mOsm/L) or hypertonic saline (sodium
268 mmol/L, 598 mOsm/L) treatment over
72 hours. There was no difference
between groups with respect to age,
gender ratios, or initial GCS score.
Changes in ICP and CPP were compared.
In both groups there was a positive
correlation between higher serum
sodium concentrations and lower ICP
and higher CPP. There was a
significantly lower frequency of acute
respiratory distress syndrome, shorter
ICU stays and lower occurrence of more
than two complications in the children
receiving hypertonic saline during the
first three days post injury.
Khanna et al.,
(2000)
USA
Case Series
N=10 Severe brain injury pediatric patients
(4 months – 12 years of age, GCS ≤ 8) who
failed conventional ICP therapy were
treated with continuous infusion of 3%
saline (514 mEq/L) to achieve a target
serum sodium level. Changes in ICP, CPP
There was a steady increase in serum
sodium over time that was statistically
significant at 24, 48, and 72 hrs (p<0.01).
A significant decrease in ICP and ICP
spike frequency and a significant
increase in CPP at 6, 12, 24, 48 and 72
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
and GOS outcome at discharge and at 6
months after discharge were evaluated.
Outcome
hrs (p<0.01) were found. Significant
increase in serum osmolarity at 12 hrs
(p<0.05) and at 24, 48 and 72 hrs
(p<0.01) was also noted. Median 6month GOS was 4 (range 1-5). One
patient died of uncontrolled ICP
hypertension. If the nine remaining
patients, seven had a GOS score of 4 or
5 (moderate-mild disability).
Pedro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In a recent prospective cohort study, Eskandari and colleagues (2013) found the infusion of
Hypertonic Saline (HTS) was effective in significantly reducing ICP in individuals who had
sustained a severe TBI. Further this reduction was sustained for 12 hours following the infusion
of HTS. Cerebral perfusion pressure increased within in minutes of bolus initiation. Study
authors noted that once the HTS bolus was administered in combination with the various
vasopressor agents, the appropriate CPP goals were met. On note mean heart rate and systolic
blood pressure were not affected by HST bolus (Eskandari et al., 2013).
In two studies, conducted by the same group of researchers, participants were randomly
assigned to receive either hypertonic saline-dextran solution or normal saline solution (0.9%
saline). The effect on the inflammatory and serum biomarker levels was assessed (Rhind et al.,
2010; Baker et al., 2009). They report that hypertonic saline-dextran is superior to normal saline
for controlling inflammatory and serum biomarker spikes while reducing elevated ICP. Results
of these two studies suggest this may help to ameliorate the secondary brain injury after TBI
but failed to show improvements in patient outcomes between groups.
In a post-hoc RCT, Myburgh and colleagues (2007) performed an analysis of saline compared to
albumin in 460 patients. A significant increase in mortality rates were seen in the albumin group
upon 24 month follow-up (p=0.003). Patient data was then divided into sub-groups and reanalyzed. Patients with severe injury (GCS 3-8) in the albumin group were significantly more
likely to have died than those in the saline group (p<0.001). However, patients with moderate
injuries (GCS 9-12) showed no statistical differences in between group mortality rates (p=0.50).
Overall, no differences were noted in rates of favorable outcomes as assessed by the GOS-E.
In a cross-over RCT, compared the effects of mannitol and a hypertonic saline dextran solution
(Battison et al., 2005). Although both mannitol and the saline solution were effective in
reducing elevated ICP, saline caused a significantly greater decrease over a longer period of
time. In a similar RCT conducted by Vailet et al. (2003) comparing the effects of mannitol and
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
hypertonic saline, the use of the latter resulted in significantly fewer episodes of ICP
hypertension. Furthermore, treatment with hypertonic saline was associated with fewer clinical
failures (defined as the persistence of ICP hypertension despite two successive infusions of the
same osmotic agent). Despite these apparent differences, mortality rates and GOS outcome 90
days post-injury did not differ between groups. Another non-randomized trial also suggested
that hypertonic saline was superior to mannitol for ICP reduction but baseline differences in
comparison groups make their findings difficult to interpret (Kerwin et al., 2009).
Cooper et al. (2004) in another RCT, examined the clinical outcomes of patients who were
randomly assigned to treatments with hypertonic saline or Ringer’s lactate solution. They
reported that at 6 months post-injury there was no difference between groups in survival,
favorable GOS outcome, FIM or Rancho Los Amigos score. Similarly, in an RCT where patients
were randomized to receive either hypertonic saline or lactated Ringer’s solution, the authors
reported that both treatments lead to reductions in ICP without any significant differences
between groups (Shackford et al., 1998). However, they indicated that those treated with
hypertonic saline required a significantly greater number of medical interventions to lower ICP.
Although this may appear contrary to the use of hypertonic saline in brain injured patients, it
should be noted that the hypertonic saline group had a significantly greater number of more
severe brain injury patients. As such, it is plausible that with increasing severity of brain injury,
patients would require a greater number of medical interventions to control ICP. Despite this
disparate randomization of more severe brain injury patients, the study found that at discharge
from the hospital, there were no significant differences between groups in GOS scores.
Pascual et al. (2008) demonstrated that hypertonic saline improved cerebral oxygenation and
therefore may be a valuable component in resuscitation of brain injured patients. The authors
caution that due to the small sample size (n=12), further study is necessary. These finding were
confirmed by in a study conducted by Oddo et al. (2009). Here hypertonic saline was found to
significantly improve oxygenation.
Lescot et al. (2006) applied CT technology to assess the effectiveness of hypetonic saline on
volume, weight and specific gravity of contused and non-contused brain tissue.Three days after
TBI, the blood brain barrier remained semi-permeable in non-contused areas but not in
contused areas. Contused tissue was shown to increase in volume after administration of
hypertonic saline. The authors recommend that further study be done to assess the effects of
hypertonic saline on different tissue types so that contusion site and size might be
appropriately factored into clinical decisions (Lescot et al., 2006).
In a study of severely head injured children, Simma et al. (1998) compared the effects of
hypotonic resuscitation with Ringer’s Lactate versus hypertonic saline during the first three days
post head injury. There was an inverse relationship between sodium concentration and
intracranial pressure. Increased serum sodium concentration correlated with lower ICP and
higher CPP. The children treated with hypertonic saline were reported to have a significantly
lower frequency of acute respiratory distress syndrome, a lower frequency of two or more
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
complications and significantly shorter ICU stay. This study suggests that hypertonic saline is
superior to ringer’s lactate for early fluid resuscitation in children following TBI.
The remaining six studies identified on this topic (Qureshi et al., 1998; Khanna et al., 2000;
Schatzmann et al., 1998; Horn et al., 1999; Ware et al., 2005; Rockswold et al., 2009) reported
positive results indicating that treatment with hypertonic saline resulted in significant reduction
in ICP with appreciable improvements in cerebral perfusion pressure. Furthermore, two of
these studies demonstrated that the reductions in ICP were mediated by concomitant
increments in serum sodium concentrations (Qureshi et al., 1998; Khanna et al., 2000).
Conclusions
There is Level 1b evidence (from 2 RCTS) to suggest that hypertonic saline reduces ICP more
effectively than mannitol.
There is Level 1 evidence that treatment with hypertonic saline results in similar clinical
outcome and survival when compared with treatment with Ringer’s lactate solution up to 6
months post-injury.
There is Level 1b evidence that saline solution results in decreased rates of mortality
compared with albumin.
There is Level 4 evidence that treatment with hypertonic saline reduces elevated ICP
refractory to conventional ICP management measures.
There is Level 2 evidence that hypertonic saline is similar to Ringer’s lactate solution in
lowering elevated ICP.
There is Level 4 evidence that hypertonic saline may be useful as a component of a
resuscitation algorithm by increasing cerebral oxygenation.
There is Level 1b evidence that in children, use of hypertonic saline in the ICU setting results in
a lower frequency of multiple early complications and a shorter ICU stay compared with
Ringer’s lactate.
There is Level 4 evidence to suggest HTS is effective in decreasing ICP levels in children post
TBI.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Hypertonic saline reduces ICP more effectively than mannitol.
Hypertonic saline and Ringer’s lactate solution are similar in lowering elevated ICP and
result in similar clinical outcomes and survival up to 6 months post-injury.
In children, use of hypertonic saline in the ICU setting results in a lower frequency of
early complications and shorter ICU stays compared with Ringer’s lactate in children.
Saline results in decreased mortality rates compared to albumin.
Hypertonic saline may reduce elevated ICP uncontrolled by conventional
ICP management measures.
Hypertonic saline may aid in resuscitation of brain injured patients by increasing
cerebral oxygenation.
16.1.2.1.2 Mannitol
Rapid administration of mannitol is among the first-line treatments recommended for the
management of increased ICP. However, this treatment is reported to be associated with
significant diuresis and can cause acute renal failure, hyperkalemia, hypotension, and in some
cases rebound increments in ICP (Battison et al. 2005;Doyle et al. 2001). For these reasons, the
Brain Trauma Foundation recommends that mannitol should only be used if a patient has signs
of elevated ICP or deteriorating neurological status. Under such circumstances the benefits of
mannitol for the acute management of ICP outweigh any potential complications or adverse
effects (AANS1995). There is also some evidence that with prolonged dosage, mannitol may
penetrate the blood brain barrier thereby exacerbating the elevation in ICP (Wakai et al., 2005).
Despite mannitol’s effectiveness in ICP management, recent evidence points to hypertonic
saline as a potentially more effective hyperosmotic agent.
The AANS make a level II recommendation that mannitol is effective for ICP control at 0.25 g/Kg
to 1 g/Kg body weight but systolic BP < 90 mmHg should be avoided . They make a Level III
(recommendation that mannitol use prior to ICP monitoring should be restricted in patients
with signs of transtentorial herniation or progressive neurological deterioration not attributable
to extracranial causes (Bratton et al., 2007e). The EBIC recommend mannitol as the preferred
osmotic therapy. They recommend administration via repeated bolus infusions, or as indicated
by monitoring, to a serum osmolarity of ≤ 315 (Maas et al., 1997).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.8 Mannitol for the Management of ICP and Hypertension Post ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Ichai et al.,
(2009)
France
RCT
PEDro = 6
N=34 Patients with severe TBI (GCS ≤ 8)
and intracranial hypertension were
randomly allocated to receive equally
hyperosmolar and isovolumetric therapy
with mannitol or sodium lactate followed
by cross-over rescue therapy when
necessary. Outcome measures included
ICP lowering at 4h and percentage of
successfully treated episodes of ICP.
Compared to mannitol, the effect of
sodium lactate on ICP was more
pronounced (7 vs. 4 mmHG, p=0.016),
th
more prolonged (4 hour decreases of 5.9
±1 vs. 3.2 ± 0.9 mmHg, p=0.009) and more
frequently successful (90.4% vs. 70.4%,
p=0.053).
Francony et al.,
(2008)
France
RCT
PEDro= 6
N=20 Stable patients with a sustained ICP
of >20 mmHG secondary to TBI (n=17) or
stroke (n=3) were given a single
equimolar infusion of either 20% mannitol
or 100mL of 7.45% hypertonic saline
during 20 mins of administration.
A single equimolar infusion of 20%
mannitol is as effective as 7.45%
hypertonic saline in decreasing ICP in
patients with ABI. Mannitol exerts
additional effects on brain circulation
through a possible improvement in blood
rheology.
Cruz et al.,
(2004)
Brazil
RCT
PEDro = 5
N=44 Severe ABI patients with recent
clinical signs of impeding brain death were
randomized to receive high dose mannitol
(500 ml of fast intravenous “wide open”
mannitol in a total dose of approximately
1.4 g/kg) or conventional-dose mannitol
(250 ml of 0.7 g/kg mannitol).
Immediately after the mannitol infusion,
patients in the high-dose and conventional
mannitol groups received 1000 or 500 ml
respectively of intravenous normal saline
to prevent acute hypovolemia and
possible secondary arterial hypotension.
Pupillary improvement (partial or full
pupillary constriction ≥ 1 mm toward the
normal diameter at 5-10 min after
treatment) and GOS scores at 6 months
were evaluated.
Improvement in bilateral abnormal
papillary widening was significantly more
frequent in the high-dose group than in the
conventional dose group (p<0.02). At 6
months post-injury, mortality rates were
39.1% and 66.7% in the high-dose and
conventional dose mannitol groups
respectively. Clinical outcomes on the GOS
scale were significantly better for the highdose than for the conventional dose group
(p<0.02).
Cruz et al.,
(2002)
Brazil
RCT
PEDro = 5
N=141 Adult patients with traumatic nonmissile acute intraparenchymal temporal
lobe hemorrhages (GCS 1-5) and abnormal
pupillary widening (partially or fully dilated
pupil with a diameter ≥ 4 mm associated
with very sluggish or absent light
response) were randomized to receive
intravenous conventional mannitol
Improvements in abnormal bilateral
pupillary widening were significantly more
frequent in the high dose mannitol group
than in the control group (p<0.03). At 6
months after injury, mortality rates were
19.4% and 36.2% for the high dose
mannitol and conventional mannitol
groups respectively. Overall clinical
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
treatment (0.7 g/kg) or high dose mannitol
treatment (1.4 g/kg). Immediately after
the mannitol infusions both groups
received normal saline infusions to
compensate for mannitol-induced urine
losses and to prevent arterial hypotension.
Pupillary improvements (partial or full
pupillary constriction ≥ 1 mm toward the
normal diameter), and GOS at 6 months
were assessed.
outcome on the GOS were significantly
better for the high dose mannitol group
with a greater number of patients in this
group showing a favorable outcome (good
recovery/moderate disability) compared
with the conventional mannitol group
(p<0.005).
Cruz et al.,
(2001)
Brazil
RCT
PEDro = 4
N=178 Adult patients with non-missile
traumatic acute subdural hematomas
were randomly assigned to receive
intravenous conventional mannitol
treatment (0.6-0.7 g/kg) or high dose
mannitol treatment (2 separate infusions
of 0.6-0.7 g/kg). Both groups received
normal saline infusions to compensate for
mannitol-induced urine losses and to
prevent arterial hypotension. In the high
dose mannitol group, the second mannitol
infusion began 25-30 min after the first
infusion. Pupillary improvements (partial
or full pupillary constriction ≥ 1 mm
toward the normal diameter), and GOS at
6 months were assessed.
Improvement in abnormal pupillary
widening was significantly more frequent
in the high dose mannitol group than in the
conventional mannitol group (p<0.0001).
At 6 months after injury, mortality rates
were 14.3% and 25.3% for the high dose
mannitol and conventional mannitol
groups respectively. Overall clinical
outcome on the GOS were significantly
better for the high dose mannitol group
with a greater number of patients in this
group showing a favorable outcome (good
recovery of moderate disability) compared
with the conventional mannitol group
(p<0.01).
Smith et al.,
(1986)
USA
RCT
PEDro = 4
N=80 Severe head injury patients (GCS
scores ≤ 8) were randomized to receive
mannitol only after ICP elevations > 25
mmHg (group I) or empirical mannitol
therapy irrespective of ICP readings (group
II). Mortality rates and neurological
outcome were compared between groups.
No significant difference in mortality
between groups I and II (35% vs. 42.5%,
p=0.26). There were also no significant
differences in neurological outcome
between groups. Mean highest ICP in nonsurvivors form both groups was
significantly higher than that in survivors
from both groups (p=0.0002).
Sayre et al.,
(1996)
USA
RCT
PEDro = 7
N=41 Moderate to severe (GCS < 12) head
injury patients who were being
transported to a hospital’s level 1 trauma
centre by helicopter within 6 hours of
injury and who had IV access, had airway
control with an endotracheal tube and
were being hyperventilated were.
randomized to receive either saline (5
mL/kg of 0.9% saline solution; 308
mOsmol/L) or mannitol (5 mL/kg of 20%
mannitol; 1,098 mOsmol/L) within the
Systolic BP or pulse rates did not change
significantly throughout the 2 hours
observation period and there were no
significant differences between groups
indicating that mannitol did not cause
secondary hypotension. There was no
difference in study volume administered to
the groups. However, urine output
(p<0.001) was significantly greater and
serum sodium (p<0.00001) was
significantly lower in the mannitol group
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
aircraft. Changes in systolic blood pressure compared with the placebo group.
(BP) during the 2-hour observation period
were compared.
Hartl et al.,
(1997)
Germany
Pre-Post
N=11 Severe head injury patients (GCS < 9)
who were sedated, intubated and
mechanically ventilated to maintain an
arterial PO2 > 100 mmHg and a PaCO2 of
approximately 35 mmHg received
mannitol (125 ml of 20% infused over 30
min through a central vein). Changes in ICP
were evaluated.
When the initial ICP before mannitol
infusion was below 20 mmHg, neither ICP
nor any other parameter changed
significantly during or after mannitol
infusion. When the pre-infusion ICP was
above 20 mmHg there was a significant
decrease in ICP and a significant increase in
CPP; however there was no change in
cerebral white matter oxygenation, or
jugular bulb oximetry.
Sorani et al.,
(2008)
US
Case Series
N=28 Patients with ABI were continuously
monitored for ICP while in the NICU.
Patients were administered 50g or 100g
doses (or both) of mannitol for
management of elevated ICP. Patient data
was then retrospectively analyzed to
determine the dose-response relationship
of mannitol to ICP.
ICP response to mannitol proved to be
dose related. Every 0.1 g/Kg of mannitol
administered (to a maximum of 100g)
resulted in approximately 1.0 mmHG drop
in ICP.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In a recent RCT conducted in France, patients were randomized to receive either mannitol or a
sodium lactate solution for management of acute episodes of elevated ICP (Ichai et al., 2009)
The authors report that an equimolar dose of sodium lactate had a significantly more
pronounced effect on acute elevations of ICP that lasted longer that treatment and with
mannitol. Sodium lactate was also successful in reducing elevated ICP more frequently. Based
on these results, further research into the effectiveness of sodium lactate in reducing ICP is
warranted.
In another trial conducted in France, equimolar doses (255 mOsm) of mannitol and hypertonic
saline were compared (Francony et al., 2008). The authors found that both interventions were
comparable in reducing ICP in stable patients with intact autoregulation. Mannitol was shown
to improve brain circulation through possible improvements in blood rheology, but also
significantly increased urine output. The authors suggest that both treatments may be
effective, but patient pretreatment factors should be considered before selection.
Cruz and colleagues conducted 3 separate RCTs in ABI patients to investigate the effects of high
dose mannitol on clinical outcomes 6 months post-injury (Cruz et al., 2004; 2001; 2002). All 3
trials reported positive results indicating that high dose mannitol (1.4 g/kg) was superior to
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
conventional mannitol (0.7 g/kg) in improving mortality rates, and clinical outcomes. In a
retrospective case study in the US, Sorani et al. (2008) found that for every 0.1 g/Kg increase in
mannitol dosage there was a 1.0 mmHG drop in ICP which support the findings put forth by
Cruz and colleagues. Further study is still recommended.
Most reports recommend administering mannitol only when elevated ICP is proven or strongly
suspected. Some discourage the use of mannitol before volume resuscitation and stabilization
of the patient due to the potential osmotic diuresis and hypotension that could result following
mannitol administration. These adverse effects could further compromise cerebral perfusion.
However, this approach may deprive head injured patients of the potentially beneficial effects
of mannitol upon ICP. With this in mind, Sayre et al. (1996) conducted another RCT to
investigate the effects or early mannitol administration in head injured patients in an out-ofhospital emergency care setting. The authors reported that compared with patients
randomized to receive saline, early out-of-hospital administration of mannitol does not
significantly affect blood pressure.
In another RCT by Smith et al. (1986) the authors reported that compared with patients who
were randomized to receive empirical mannitol irrespective of ICP measurements, those who
received mannitol only after the onset of intracranial hypertension (> 25 mmHg) were not
significantly different in terms of mortality rates or neurological outcomes.
The findings of a single group intervention study by Hartl et al. (1997) indicate that mannitol is
only effective in diminishing ICP when the initial ICP is hypertensive (>20 mmHg) and not when
it is below such values. Thus, the use of mannitol as a prophylactic measure against potential
elevations in ICP may not be appropriate. This was corroborated in a more recent study by
Sorani et al. (2008).
A 2008 Cochrane review suggests that mannitol may have beneficial effects on mortality when
compared to pentobarbital but detrimental effects when compared to hypertonic saline. They
also report that there is insufficient data on the effectiveness of pre-hospital administration of
mannitol (Wakai et al., 2005).
Conclusions
There is Level 1 evidence that sodium lactate is more effective than mannitol for the
management of acute elevations in ICP.
There is Level 2 evidence that higher dose mannitol is superior to conventional mannitol in
improving mortality rates, and clinical outcomes.
There is Level 2 evidence that early out-of-hospital administration of mannitol does not
adversely affect blood pressure.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
There is Level 4 evidence that mannitol is effective in diminishing intracranial hypertension
only when initial ICP values are elevated.
Sodium lactate is more effective than mannitol for reducing acute elevations in ICP
High dose mannitol results in lower mortality rates and better clinical outcomes
compared with conventional mannitol.
Early out of hospital administration of mannitol does not negatively affect blood
pressure.
Mannitol may only lower ICP when initial ICP values are abnormally elevated.
16.1.2.2 Propofol
Propofol is a fast acting sedative that is absorbed quickly and metabolized equally quickly
leading to pronounced effects of short duration. Its beneficial effects occur via decreases in
peripheral vascular tension resulting in potential neuroprotective effects, which may be
beneficial in acute ABI care. Experimental results have shown positive effects on cerebral
physiology including reductions in cerebral blood flow, cerebral oxygen metabolism, EEG
activity, and ICP (Adembri et al., 2007). However, administration of high doses can result in
propofol infusion syndrome (PRSI), which has been characterized by severe metabolic acidosis,
rhabdomyolosis, cardiac dysrhythmias, and potential cardiovascular collapse (Corbett et al.,
2006). Children are especially susceptible to PRSI (Sabsovich et al., 2007).
The AANS recommend propofol use for the control of ICP but not for improvement in mortality
or 6 month outcome (Carney and Ghajar 2007). They also indicate that high-dose propofol can
produce significant morbidity (Bratton et al., 2007b) The EBIC recommend sedation as part of
the treatment course for ABI but make no specific mention of propofol (Maas et al., 1997).
Individual Studies
Table 16.9 Propofol for the Management of Acute ABI
Author/Year/
Country/Study
design/PEDro
Score
James et al.,
(2012)
USA
Methods
N=8 Participants were randomly assigned to
receive either propofol or dexmedetomidine
for the first 6 hours of the study.
Outcome
No significant differences were noted
between the two groups on any of the
physiological measure (ICP, CPP,
bispectral index (BIS),
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
RCT
PEDro=6
Following this participants were crossed over
into the opposite treatment protocol.
brain tissue oxygenation (PbtO2), lactate
to pyruvate (L/P ration).
Kelly et al.,
(1999)
USA
RCT
PEDro = 8
N=42 Patients with a GCS 3-12 who required
mechanical ventilation were randomly
assigned to receive propofol (20mg/ml with
0.005% EDTA) or morphine. Patients were
assessed for adverse effects, physiological
response (including ICP) and 6 month GOS.
Intracranial pressure therapy in the
propofol group was less intensive than
the morphine group (less use of
neuromuscular blocking agents,
benzodiazepines, pentobarbital, and CSF
drainage) and ICP on day 3 was
significantly lower (p<0.05). Six month
GOS scores were not significantly
different between groups for mortality
or favorable outcome rates.
Stewart et al.,
(1994)
UK
Case series
N=15 Patients were sedated with either a
continuous infusion of propofol (mean 232
mg/hr, range 150-400 mg/hr) or infusions of
morphine (mean rate 2.3 mg/hr, range 04mg/hr) and midazolam (mean rate 2.8mg/hr,
range 0-5 mg/hr). Continuous collection of
AVDO2, MABP, ICP, and CPP was performed.
A fall in AVDO2 after 4 hours from 6.3 ±
2.6 ml/dl to 3.0 ± 0.6 ml/dl was noted
while on propofol. No significant
differences were seen in any of the
other measures in either group.
Farling et al.,
(1989)
Ireland
Case Series
N=10 Patients with severe head injuries that
required sedation were given intravenous
propofol infusions as a 1% solution at a rate of
2-4 mg/kg/hr. Dose was adjusted to maintain
ICP below 10 mmHg and CPP above 60 mmHg.
Patients were monitored for HR, MABP, ICP,
CPP, pupil size and PaCO2.
A mean infusion rate of 2.88 mg/kg/hr
was sufficient for sedation and recovery
was rapid. CPP was significantly
increased at 24 hrs. Significant
differences were not seen in any of the
other variables.
Smith et al.,
(2009)
USA
Case Study
N=146 Severe TBI (GCS ≤ 8) patients who had
been treated with either vasopressures or
propofol were analyzed. Patients were
monitered for developing symptoms of
propofol infusion syndrome (PRIS)
Of the 146 patients included in the
study, only 3 patients on both propofol
and vasopressors developed PRIS. Of
note, there were no patients on either
propofol or vasopressures who
developed PRIS. PRIS was not linked to
mortality (p>0.05)
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002)
Discussion
In a recent RCT, propofol and dexmedetomidine were administered to a group of 8 individuals
who had sustained an ABI (James et al., 2012). In this randomized cross over trial each
medication was given for a 6 hour period, at which time participants were administered the
opposite medications. Medication doses were individualized. No significant differences were
found between the groups. As a result of these findings, study authors recommend that the
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
“choice of sedative regimen be based on the profile of the sedative and the individual goals for a
patient (pg 955)”
In an earlier RCT, propofol sedation was compared to morphine for safety and efficacy (Kelly et
al. 1999). Here patients were randomly assigned to either a morphine group or a propofol
group where they received interventions consisting of three simultaneous injections Injection 1
contained propofol or placebo, injection 2 morphine or placebo and injection three low-dose
morphine (both groups). Physicians were allowed to administer injections as regularly as every
5 minutes as needed. This particular design allowed for the comparison of propofol dosing and
its effectiveness while preserving the blind experimental component. However, all patients
received morphine in conjunction with propofol infusion. Propofol tended to reduce ICP
generally with significance reached on day 3 (p<0.05). Patients in the propofol group also
showed less need for neuromuscular blocking agents, benzodiazepines, pentobarbital, and CSF
drainage. At 6 months, there were no significant differences in mortality rates or GOS scores.
The authors suggest that propofol is a safe, acceptable, and possibly desirable alternative to
opiate-based sedation (Kelly et al., 1999).
In the studies conduced by Stewart et al (1994) and Farling and colleagues (1989) propofol was
reported to provide satisfactory sedation with few side effects. Stewart et al. (1994) reported
that propofol provided sedation similar to a combination of midazolam and morphine with no
differences in 6 month outcomes between groups. Farling et al. (1989) also reported that
propofol provided safe and effective sedation. Both of these studies were small and received
poor methodological scores. Further study is warranted.
There is concern regarding the potential of high-dose propofol resulting in propofol infusion
syndrome. Several case studies and research synthesis articles have recently been released
warning of this potentially fatal side effect. Cremer et al. (2001) provided the most
comprehensive report after 5 patients in their ICU died of unexplained cardiac failure on days 45 of treatment. They retrospectively assessed all patients treated in their facility from 1996-99
and identified 7 patients who died of apparent propofol infusion syndrome. Upon logistic
regression, they identified a “crude” odds ratio of 1.93 (95% CI 1.12-3.32, p=0.018) for
developing propofol infusion syndrome per unit (mg/kg per h) increase in propofol dose. Smith
et al. (2009) conducted a similar retrospective review and identified 3 patients (out of 146) with
propofol infusion syndrome (PRIS) all of whom were receiving vasopressors and propofol
infusions. They report an odds ration of 29 (95% CI, 1.5-581, p<0.05) for developing PRIS while
receiving both vasopressors and propofol. The authors also noted that no patient on either
propofol or vasopressors alone developed PRIS. Otterspoor et al. (2008) also released a review
of case reports to assess the risk factors associated with propofol infusion syndrome. They
recognized large cumulative doses, young age, acute neurological injury, low carbohydrate
intake, high fat intake, catecholamine infusion, corticosteroid infusion, critical illness, and
inborn errors of mitochondrial fatty acid oxidation as potential risk factors. Otterspoor et al.
(2008) recommend that until further research has been conducted, “4mg/kg/hr infusions
should not be exceeded in patients with severe head injury and others who are in need of
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
prolonged sedation” (pg 550) .5mg/kg/hr should never be exceeded in any patients found
acceptable for sedation in the ICU.
Conclusions
There is Level 1 evidence that propofol may help to reduce ICP and the need for other ICP and
sedative interventions when used in conjunction with morphine.
Propofol may help to reduce ICP and the need for other ICP and sedative interventions
when used in conjunction with morphine
Infusions of propofol greater than 4mg/kg per hour should be undertaken with
extreme caution.
16.1.2.3 Midazolam
Midazolam is a fast-acting benzodiazepine with a short half-life and inactive metabolites
(McCollam et al., 1999). It is anxiolytic and displays anti-epileptic, sedative, and amnestic
properties. It is a protein-bound, highly lipid soluble drug which crosses the blood brain barrier
and has a rapid onset of action within 1-5 minutes in most patients (McClelland et al.,1995).
However, delayed elimination of midazolam, resulting in prolonged sedation, has been
demonstrated in some critically ill patients.
Studies in the operating room or intensive care unit have demonstrated Midazolam to be
relatively safe in euvolemic patients or in the presence of continuous hemodynamic monitoring
for early detection of hypotension (Davis et al., 2001). Midazolam has been found to reduce
cerebrospinal fluid pressure in patients without intracranial mass lesions as well as decrease
cerebral blood flow and cerebral oxygen consumption (McClelland et al., 1995). One RCT and
two non-RCTs on midazolam use in acute ABI management were reviewed.
The AANS guidelines made no evidence based recommendations regarding midazolam’s
efficacy but they do suggest a 2 mg test dose followed by a 2-4 mg/h infusion if used (Bratton et
al., 2007b). The EBIC recommend sedation but make no specific reference to midazolam (Maas
et al., 1997).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.11 Midazolam for the Management Acute ABI
Author/Year/
Country/Study
design/Pedro
Scores
Methods
Outcome
Sanchez-IzquierdoRiera
(1998)
Spain
RCT
PEDro = 5
N=100 Trauma patients were randomely
assigned to one of three groups. Group
A: was given a continuous infusion of
midazolam (0.1 mg/kg/hr to a max of 0.35
mg/kg/hr); group B was given continuous
IV infusion of propofol (1.5 mg/kg/hr to a
max of 6 mg/kg/hr); or group C was
givencontinuous infusion of midazolam
(0.1 mg/kg/hr to a max of 0.2 mg/kg/hr)
and propofol (1.5-3 mg/kg/hr) only if
further sedation was necessary. All
patients received morphine as well.
Patients were monitored for sedation,
hemodynamic and oximetric variables.
All three regimens achieved similar
sedationand incidences of adverse
effects. No differences were found in
ICP, CPP, or jugular venous oxygen
saturation in head trauma patients.
Serum triglyceride levels were
significantly higher in propofol patients
but wakeup time was shorter.
Papazian et al.,
(1993)
France
Case Series
N=12 Patients with severe head injury (GCS
≤ 6) were given bolus doses of midazolam
(0.15 mg/kg i.v.). Patients were monitored
for MAP, ICP, and CPP.
Significant reductions in MAP (89 mmHg
to 75 mmHg, p<0.0001) and in CPP (71
mmHg to 55.8 mmHg, p<0.0001) were
observed. Overall, no significant change
in ICP was noted. However, patients
with initial ICP ≤ 18 mmHg saw increases
in ICP while those with initial ICP >18
mmHg was decreases.
Davis et al.
(2001)
USA
Case Series
N=219 Patients data was retrospectively
reviewed in two different regions with
different pre-hospital midazolam dosing
protocols. North crews used 0.1 mg/kg for
every patient being intubated while south
crews used 0.1 mg/kg up to 5mg. Multiple
linear regression was used to assess the
relationship between midazolam dose and
hypotension and systolic blood pressure.
A significant relationship was seen
between midazolam dose and both
hypotension and decreased systolic
blood pressure.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
Infusions of midazolam or propofol were reported to provide similar quality sedation in patients
with severe head trauma, although propofol was associated with a high incidence of
hypertriglyceridemia (Sanchez-Izquierdo-Riera et al., 1998). In both studies evaluating
midazolam and ICP, no significant difference was seen after midazolam administration
(Sanchez-Izquierdo-Riera et al., 1998; Papazian et al., 1993). However, significant hypotension
related to increased doses of midazolam (Davis et al., 2001) and decreases in MAP resulting in
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
decreased CPP (especially in patients with initial ICP ≤ 18 mmHg) (Papazian et al., 1993) were
also reported. The study by Sanchez-Izquierdo-Riera et al. (1998) measured ICP, CPP and MAP
in all patients and reported no between group differences. However, they did not report
comparisons with baseline values making it unclear whether or not midazolam resulted in any
negative effects. Based on current evidence, hypotension should be monitored as a potential
side effect during midazolam administration.
Conclusions
There is Level 2 evidence that midazolam has no effect on ICP but conflicting evidence
regarding its effect on MAP and CPP.
Midazolam has no effect on ICP but may result in systemic hypotension.
16.1.2.4 Opioids
Opioids are substances that have morphine-like actions. They work by binding to opioid
receptors, found principally in the central nervous system and the gastrointestinal tract. Each
opioid has a distinct binding affinity to group(s) of opioid receptors that then determines its
pharmacodynamic response. Morphine has been the most commonly used opioid following
ABI, while fentanyl and its derivatives have gained popularity owing to their more rapid onset
and shorter duration of effect (Metz et al., 2000). Controversy persists regarding the effect of
opioids on ICP and CPP. It has been reported that opioids can increase cerebral blood flow
(CBF), which may lead to an increase in ICP, (Marx et al., 1989; de Nadal et al., 2000; Werner et
al., 1995; Bunegin et al., 1989) in the presence of intracranial pathology. Despite a
reexamination of the literature, nothing new was found.
Individual Studies
Table 16.11 Opioids for the Management Acute ABI
Author/Year/
Country/Study
design/Pedro
Score
de Nadal et al.,
(2000)
Spain
RCT
PEDro = 8
Methods
Outcome
N=30 Severe head injury patients were
randomly assigned to receive morphine
(0.2 mg/kg) and fentanyl (2 μg/kg)
through IV for 1 minute every 24 hours in
a crossover fashion. ICP, MAP, and CPP
were monitored for 1 hour after
administration. Cerebral blood flood was
estimated using Transcranial Doppler
sonography.
CO2 reactivity was maintained in all
patients but 18 patients showed
impaired or abolished autoregulation.
Both morphine and fentanyl caused
significant increases in ICP and
decreases in MAP and CPP. Estimated
cerebral blood flow remained the same.
No difference was seen in ICP increases
between patients with intact
autoregulation and those without.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/Pedro
Score
Methods
Outcome
Sperry et al.,
(1992)
USA
RCT
PEDro = 7
N=9 Patients with severe head trauma
(mean GCS 6 ± 1) received an IV bolus of
fentanyl (3μg/kg) or sufentanil (0.6
μg/kg) in a randomized masked fashion.
Patients then received the other opioid
24 hrs later. MAP, HR, and ICP were
recorded continuously for the first hour
after administration.
Both fentanyl and sufentanil resulted in
significant increases in ICP (8±2 mmHg
and 6±1 mmHg respectively) and
statistically significant decreases in MAP
(11±6 mmHg and 10±5 mmHg). No
changes in heart rate were noted.
Karabinis et al.,
(2004)
Greece
RCT
PEDro = 5
N=161 Patients were randomized to
receive analgesia-based sedation
(remifentanil 9 μg/kg/h and propofol
0.5mg/kg/h (days 1-3) or midazolam
0.03mg/kg/h (days 4-5)), hypnotic-based
sedation (propofol (days 1-3;midazolam
days 4-5) and fentanyl), or morphine.
Agents were titrated to receive optimal
sedation in all three cases.
Sedation with remifentanil permitted
significantly faster (p=0.001) and more
predictable awakening for neurological
assessment (p=0.024).
Lauer et al.,
(1997)
USA
RCT
PEDro = 5
N=15 Severely brain injured patients (GCS
≤8) randomly received fentanyl,
sufentanil or morphine titrated to a
maximal decrease in MAP of 10%
followed by a continuous infusion of the
same opioid for 4 hours. Patients were
monitored for ICP, MAP, and HR.
There was no increase in ICP in any
group. There was a significant decrease
in MAP in the sufentanil group at 10 min
(p<0.05) and 45 min after initial bolus.
Albanese et al,
(1999)
France
Pre-Post
N=6 Patients randomly received a 6-min
injection of either sufentanil (1μg/kg),
alfentanil (100μg/kg), or fentanyl
(10μg/kg) followed by an infusion of
0.005, 0.7, and 0.075 μg/kg/min
respectively for 1 hr. MAP, ICP, CPP, and
SjvO2 were continuously measured every
minute throughout the hour.
All three medications were associated
with significant increases in ICP peaking
before 6 minutes and returning to
baseline by 15 min. Increases in ICP
were accompanied by decreases in MAP
and thus CPP. No evidence of cerebral
ischemia was noted.
Scholz et al.,
(1994)
Germany
Pre-Post
N=10 Head injured patients (GCS<6)
received an IV bolus of sufentanil
(2μg/kg) followed at 30 min by infusion
of sufentanil (150 μg/hr) and midazolam
(median 9 mg/hr) over 48 hrs.
Pharmacokinetic and physiological
measures were recorded.
Decreases in ICP (16.1±1.7 mmHg to
10.8±1.3 mmHg, p<0.05) and MAP
(85.53.9 mmHg to 80.2±4.9 mmHg,
p<0.05) were noted. CPP remained
stable.
Albanese et al.,.
(1993)
France
Case Series
N=10 Head trauma patients sedated with
propofol received further sedation using
an IV injection of 1 μg/kg over 6 min and
infusion of 0.005 μg/kg/min. MAP, ICP
and end tidal CO2 were measured every
Sufentanil injection resulted in a
significant increase in ICP (9±7 mmHg)
that peaked after 5 min and gradually
returned to baseline after 15 minutes.
This was accompanied by a significant
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/Pedro
Score
Methods
Outcome
minute for 30 minutes.
decrease in MAP and CPP that gradually
increased but remained significant
throughout the study.
Engelhard et al.,
(2004)
Germany
Pre-Post
N=20 Head trauma patients (GCS<8)
sedated with propofol and sufentanil
received an IV bolus of remifentanil (0.5
μg/kg) followed by an infusion of 0.25
μg/kg/min for 20 min. Patients were
monitored for MAP, ICP, CBFV using
transcranial Doppler flowmetry.
No differences were noted in MAP, ICP,
or CBFV after remifentanil
administration.
Werner et al.,
(1995)
Germany/USA
Cohort
N=30 Patients with severe TBI (GCS<6)
received an IV bolus of sufentanil (3
μg/kg) and were monitored for 30 min.
Heart rate, arterial blood gases and
esophageal temperature did not
change. MAP decreased greater than
10mmHg in 12 patients. ICP was
constant in patients with maintained
MAP, but was significantly increased in
those with decreased MAP.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
Analgesic sedation with opioids is commonly used in conjunction with hypnotic agents (i.e.
midazolam, propofol) to reduce nociceptive stimulation. This makes it difficult to evaluate the
effects of opioids in isolation. Five studies reported increases in ICP after opioid administration
(Werner et al., 1995; de Nadal et al., 2000; Sperry et al., 1992; Albanese et al., 1993; Albanese
et al., 1999), while 2 found no increase (Lauer et al., 1997; Engelhard et al., 2004) and one
reported a decrease (Scholz et al., 1994). However, mode of administration has been suggested
as a determining factor for increases in ICP (Albanese et al. 1999 & 1993). In those studies
where patients received only bolus injections of opioids, significant increases in ICP were seen
(Werner et al., 1995; de Nadal et al., 2000; Sperry et al., 1992).
Fentanyl and its derivatives have been suggested as more ideally suited for sedation in patients
with brain injury due to their rapid onset and short duration (Metz et al., 2000). In our review,
one study found remifentanil resulted in significantly faster arousal compared to propofol or
midazolam (Karabinis et al., 2004). The authors suggested that this allowed for prompt
neurological assessment. However, patients in the treatment group received remifentanyl as
the primary sedative agent and then a hypnotic agent, while patients in the control groups
received fentanyl or morphine in conjunction with a hypnotic agent. Therefore, remifentanil’s
efficacy can be compared to hypnotic based sedation but not fentanyl or morphine.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Conclusions
There was Level 1 evidence that bolus opioid administration resulted in increased ICP.
There was conflicting evidence regarding the effects of opioid infusion on ICP levels.
There was Level 2 evidence that remifentanil results in faster arousal compared to hypnotic
based sedation.
Bolus opioid administration results in increased ICP.
There is conflicting evidence regarding the effects of opioid infusion on ICP.
Remifentanil results in faster arousal compared to hypnotic based sedation.
16.2.2.5 Barbiturates
It has long been proposed that barbiturates may be useful in the control of ICP. Barbiturates
are thought to reduce ICP by suppressing cerebral metabolism to reduce metabolic demands
and cerebral blood volume (Roberts 2000). Early reports indicated that barbiturates reduced
ICP even in patients reported to be unresponsive to rigorous treatments with conventional ICP
management techniques including mannitol and hyperventilation (Marshall et al., 1979).
Further studies supported the therapeutic potential of barbiturates and suggested that failure
to control ICP can lead to death (Rea and Rockswold 1983; Rockoff et al., 1979). However, most
of these early investigations provided only anecdotal or poor evidence as they were conducted
in very small cohorts of patients lacking control comparisons. More recent studies have
explored the negative side effects associated with barbiturate coma such as adrenal
insufficiency (Llompart-Pou et al., 2007) and bone marrow suppression (Stover and Stocker
1998).
The AANS make Level II recommendations that prophylactic administration of barbiturates to
induce EEG burst suppression should not be performed. They also make Level II
recommendations that high-dose barbiturate administration can be used to control elevated
ICP that is refractory to maximum standard medical and surgical treatment (Bratton et al.,
1007). The EBIC guidelines recommend barbiturate use to increase sedation only after sedation,
analgesia, hyperventilation, osmotic therapy, and CSF drainage have failed to control ICP (Maas
et al., 1997).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.13 Barbiturates for the Management of Elevated Intracranial Pressure Post ABI
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
Eisenberg et al.,
(1988)
USA
RCT
PEDro = 4
N=73 Severe head injury patient (GCS ≤
7) were randomized to receive either
high-dose pentobarbital (loading dose
10 mg/kg over 30 min, 5 mg/kg q 1 hr x
3. Maintenance dose of 1mg/kg/hr or
adjusted to achieve serum levels of 34mg%) or conventional therapy
(elevation of the head, hyperventilation,
morphine, pancuronium, mannitol,
ventricular drainage) for the reduction
of ICP). Changes in ICP and survival at 30
days and 6 months were assessed.
The chance of ICP control in patients
with ICP refractory to conventional
management was nearly double (ratio
1.94, p=0.12) for patients in the
barbiturate group compared with
controls. After declaration of treatment
failure, 26 of the patients randomly
assigned to conventional therapy were
crossed over to receive barbiturates.
The likelihood of survival at 1 month
was 92% for those who responded to
barbiturates while 83% of the nonresponders died. 80% of all deaths in
each of the groups were due to
uncontrolled ICP. Last follow up
examination (a median of 6 months
post-injury) showed that 36% of the
responders and 90% of the nonresponders were vegetative or had died.
Perez-Barcena et al.,
(2005)
Spain
RCT
PEDro = 5
N=20 Severe traumatic brain injury
patients (GCS ≤ 8) who presented with
intracranial hypertension (ICP > 20
mmHg) refractory to conventional first
level measures were randomized to
receive either thiopental (initial bolus of
2 mg/kg over 20 seconds to reduce ICP
below 20 mm Hg. If ICP did not
decrease, a second bolus injection at a
dose of 3 mg/kg was administered
which was subsequently increased to 5
mg/kg if ICP remained elevated. Once
ICP was reduced, patients received a
continuous infusion at a rate of 3
mg/kg/hr) or pentobarbital (loading
dose of 10mg/kg for 30 min followed by
a continuous infusion at a rate of 5
mg/kg/hr for 3 hours, and then a dose
of 1 mg/kg/hr for the last hour) to
control refractory elevations in ICP.
Changes in ICP and mortality at
discharge and 6 months later were
compared between groups. Successful
ICP control was defined as a drop in ICP
below 20 mmHg for at least 48 hours.
Thiopental was able to control ICP in
50% of patients while pentobarbital was
only able to control ICP in 20% (p=0.16).
50% of patients in the thiopental group
died at discharge while this figure rose
to 80% in the pentobarbital group
(p=0.16).
Perez-Barcena et al.,
(2008)
Spain
RCT
N=44 Severe traumatic brain injury
patients (GCS ≤ 8) who presented with
intracranial hypertension (ICP > 20
mmHg) refractory to conventional first
Multivariate logistic regression resulted
in an odds ratio of 5.1 (95%CI, 1.2 to
21.9, p=0.027) in favour of thiopental
for control of refractory ICP compared
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
PEDro = 4
level measures (BTF guidelines) were
randomized to receive either thiopental
(initial bolus of 2 mg/kg over 20
seconds; if ICP remained refractory (ie.
>20 mm Hg) a second bolus of 3 mg/kg
was administered which was
subsequently increased to 5 mg/kg if
necessary, followed by continuous
infusion of 3 mg/kg/hr once ICP was
controlled) or pentobarbital (loading
dose of 10mg/kg for 30 min followed by
a continuous infusion at a rate of 5
mg/kg/hr for 3 hours, and then a dose
of 1 mg/kg/hr for the last hour) to
control refractory elevations in ICP.
to pentobarbital. The relative risk for
good control of ICP was 2.26 for
thiopental vs. pentobarbital in patients
with focal lesions and 3.52 in patients
with difuse lesions.
Schwartz et al.,
(1984)
Canada
RCT
PEDro = 5
N=59 Severe brain injury patients (GCS ≤
7) who had elevated intracranial
pressure (ICP > 25 torr for more than 15
minutes) were randomized to receive
mannitol (20 % with an initial dose of 1
gm/kg) or pentobarbital (initial
intravenous bolus of up to 10 mg/kg,
followed by continuous infusion at 0.5-3
mg/kg/hr provided that cerebral
perfusion pressure remained > 50 torr)
initially followed by the second drug as
required by further elevation of ICP
(defined as failure to control ICP by the
current treatment). Subjects were
stratified at the outset into two groups,
those with intracranial hematomas and
those without. ICP and survival at 3
months were compared between
groups.
No significant difference in mortality of
patients with evacuated hematomas in
the pentobarbital or mannitol groups
(40% and 43% respectively); however, in
those with evacuated hematomas twice
as many patients in the pentobarbital
group required the mannitol to control
raised ICP than did patients starting with
mannitol indicating that pentobarbital is
not better than mannitol for the control
of ICP (p=0.04). Patients with no
hematoma treated with pentobarbital
as initial therapy had 77% mortality
compared to 41% mortality in those
treated initially with mannitol. In these
patients, there was a higher rate of
failure to control ICP in the
pentobarbital group than in the
mannitol group indicating that
pentobarbital is not better than
mannitol to control ICP (p<0.001).
Ward et al.,
(1985)
USA
RCT
PEDro = 6
N=53 Head injury patients (GCS < 8)
were randomly assigned for placement
into a control group (conventional ICP
management measures) or a
barbiturate-treated group who received
intravenous pentobarbital (loading dose
at 5-10 mg/kg or enough to achieve
burst suppression on the
electroencephalogram. After loading
dose, pentobarbital was given hourly,
initially as a bolus and then as a
Groups were similar in terms of age, sex
distribution, cause of injury,
neurological status, intracranial lesions,
and initial ICP. Clinical outcome on the
GOS at 1 year did not differ between
groups (both groups had equal number
of deaths, patients with good outcome,
moderate or severe disability). During
the first 4 days there was no significant
difference in hourly levels of ICP levels,
the number of patients dying from
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
continuous infusion to achieve a
uncontrolled ICP hypertension, the
maintenance dose of 1-3 mg/kg
duration of ICP elevation.
adjusted to maintain a serum level of
25-45 mg%). Pentobarbital was started
as soon as possible after the injury
regardless of the ICP and continued for
at least 72 hours and was then slowly
discontinued. Changes in ICP and
outcome on the Glasgow Outcome scale
at 1 year were compared.
Fried et al.,
(1989)
USA
Non-RCT
N=7 ABI patients (GCS 4-8) were
involved in this study in which one
group received a bolus injection of
intravenous pentobarbital followed by
continuous infusions to achieve a serum
pentobarbital concentration of 20-40
mg/L (n=4) or received conventional
therapy (n=3). Measured energy
expenditure (% of predicted), 24-hour
nitrogen excretion and urinary 3methylhistadine excretions were
assessed.
Measured energy expenditure, and 24
hours nitrogen excretion were
significantly lower in the pentobarbital
group compared with control (p <0.01 in
both cases). There was no significant
difference in urinary 3-methylhistidine
excretions between groups.
Llompart-Pou et al.,
(2007)
Spain
Case control
N=40 Patients were prospectively
studied with moderate to severe TBI.
Seventeen patients were treated with
barbiturate coma and 23 had their ICP
controlled through tier I measures and
were used as a control. Adrenal function
was assessed using the high-dose
corticotrophin stimulation test within
24h after brain injury and after
barbiturate coma induction.
Within 24h, adrenal function was similar
in both groups. After barbiturate coma,
patients in group A (barbiturates)
presented higher insufficiency vs.
control (53% vs 22%, p=0.03). Patients
treated with barbiturates who
developed insufficiency required higher
levels of Norepinephrine to maintain
CPP than the barbiturate treated
individuals without insufficiency.
Stover & Stocker
(1998)
Germany
Case control
N=52 Patients with sever head injury
were investigated. Twenty three
patients did not respond to ICP
treatment and were administered
Thiopental at 5-11 mg/Kg b.w. as a bolus
followed by continuous infusion of 4-6
mg/Kg/h to maintain a burst
suppression pattern of 4-6 bursts/min.
Patients were monitored for white
blood cells while their tracheobronchial
secretions and urine were sampled for
bacterial growth.
Barbiturate coma was shown to induce
reversible leukopenia and
granulocytopenia as well as an
increased infection rate. Several
patients showed suppressed bone
marrow production on histological
examination.
Nordby and
N=38 Severe brain injury patients (GCS ≤ Better GOS outcome in the thiopental
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
Nesbakken
(1984)
Norway
Non-RCT
6, all younger than 40 years of age)
experiencing a progressive rise in ICP to
40 mmHg for 25 min despite intensive
therapy with hyperventilation, steroids
and mannitol were assigned to receive a
continuous infusion of thiopental
(loading dose of 10-20 mg/kg and a
maintenance dose of 3-5 mg/kg/hour)
and hypothermia (32-35 ºC) or to
continue conventional intensive care.
GOS outcome was assessed at 9-12
months post-injury.
group compared with the conventional
therapy group (p=0.03). Therapy with
thiopental resulted in 6 patients with
good/moderate outcome, 3 severe and
7 dead/vegetative. In contrast
conventional therapy resulted in 2
patients with good/moderate outcomes
and 13 dead/vegetative.
Thorat et al.,
(2008)
Singapore
Case Series
N=12 Patients with severe TBI were
managed with barbiturate coma if
medical therapy failed to control
elevated ICP. Patients were
continuously monitored for ICP,
pressure reactivity and PTiO2.
No significant reductions in mean ICP,
MAP, CPP, PTiO2, or PRx were reported.
Eight of the patients experienced
reductions in ICP but only 4 below 20
mmHg. Improved oxygenation was seen
in 6 of the 8 patients with PTiO2 levels
greater than 10 mmHg prior to coma.
Schalen et al.,
(1992)
Sweden
Case Series
N=38 Patients with severe TBI who
despite conventional management
developed a dangerous increase in ICP
were treated with high dose intravenous
thiopentone (5-11 mg/kg, followed by a
continuous infusion at 4-8 mg/kg/hr to
achieve and maintain a burst
suppression pattern on the
electroencephalogram (EEG).
Treatment continued until ICP
decreased and remained stable below
20 mmHg for at least 12 h, or until
treatment was considered to be
ineffective. Changes in ICP and CPP
were assessed.
After induction of treatment, a fall in
mean arterial blood pressure (MABP)
was seen in 31 patients, in 4 patients no
change occurred and a small increase
was seen 3 patients. There was a
simultaneous decrease in ICP in 26
patients, no change in 3 patients, and a
small increase in 2 patients. CPP
decreased in 18 patients, increased in
10 patients and remained unchanged in
3 patients.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
The findings of the RCT conducted by Eisenberg et al. (1988) suggest that the use of high dose
pentobarbital is an effective adjunctive therapy for the management of elevated ICP refractory
to conventional therapeutic measures. However, this study only supported the use of this high
dose barbiturate for a small subgroup of severe ABI patients (GCS ≤ 7). In contrast, the findings
of another RCT conducted by Ward et al. (1985) suggest that pentobarbital is no better than
conventional ICP management measures, which was corroborated by Thorat et al. (2008) in a
smaller case series. Results of the study by Ward et al. (1985) contradict those of Eisenberg et
al. (1988). Schwartz et al. (1984) compared pentobarbital and mannitol for the control of ICP in
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
another RCT. Their findings support Ward et al. suggesting that pentobarbital is not better than
mannitol in the treatment of ICP.
Furthermore, the latter study also reported that more than half of those treated with
pentobarbital developed arterial hypotension, an adverse effect that could worsen the
condition of patients with severe ABI. Similarly, Schalen et al. (1992) noted that although
pentobarbital may decrease elevated ICP, it may also decrease cerebral perfusion pressure due
to a decrease in arterial pressure.
Results of one meta-analysis suggest that although barbiturates reduced elevated ICP, there
was little evidence to link this with reductions in mortality or disability. Furthermore,
barbiturate therapy was associated with substantial hypotension that may have offset any ICPlowering effect (Roberts, 2000).
In accordance with recommendations made by the Brain Trauma Foundation, Perez-Barcena et
al. (2008) compared the efficacy of pentobarbital and thiopental on the management of
refractory ICP unmanageable by conventional measures. In two linked trials (the second an
extension of the first), they reported that thiopental is superior to pentobarbital in controlling
refractory ICP (Perez-Barcena et al., 2005 & 2008). In the first report, thiopental was shown to
help reduce refractory ICP in a greater number of patients although these differences were not
statistically different Perez-Barcena et al., 2005). In a follow-up report, the authors found
statistically significant results in favour of thiopental using multivariate logistic regression
(Perez-Barcena et al., 2008). Even after randomization, initial CT findings were different
between groups, however, the results held for patients with both focal and difuse lesions on
initial CT. Similarly, Llampart-Pou et al. (2007) found thiopental less likely to induce adrenal
insufficiency when compared to pentobarbitol further supporting its use when barbiturate
coma is indicated. It should be noted that in a much earlier study, Stover et al. (1998), reported
that use of thiopental significantly reduced white blood cell production and could induce
reversible leukopenia and granulocytopenia relative to TBI patients who did not require
barbiturate sedation. The authors also noticed interactions with bone marrow suppressing
antibiotics (specifically, tazobactum/piperacillin) which further exacerbated problem.
Therefore, in instances where barbiturate coma is indicated, monitoring of immunological
response is recommended.
Fried et al. (1989) conducted a study to compare the energy expenditure and nitrogen
excretion in patients treated with pentobarbital and those who received conventional ICP
therapy without pentobarbital. They reported that pentobarbital essentially lowered the energy
expenditure and nitrogen excretion and suggest that this in turn would better enable the brain
to achieve energy and nitrogen equilibrium during metabolic support of acute head-injured
patients.
There is some evidence that barbiturate therapy may contribute to improvements in long-term
clinical outcomes. In a prospective controlled trial conducted by Nordby and Nesbakken (1984),
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
the authors reported that thiopental combined with mild hypothermia resulted in better clinical
outcomes as per the Glasgow Outcome scale 1 year post-injury when compared with
conventional ICP management measures (including hyperventilation, steroids and mannitol).
However, since this study used a combination of thiopental and hypothermia, it is not possible
to attribute the better clinical outcomes to thiopental alone.
Roberts (2000), in his meta-analysis, noted that further randomized trials are needed to
determine the effects of barbiturates on clinical outcomes such as mortality and disability
following severe ABI. Similarly, a 1999 Cochrane review stated that there was no evidence that
barbiturate use in TBI patients improved outcomes and were reported to decrease blood
pressure in one of four patients, which will offset the effect of ICP reduction on CPP (Roberts
1999). Therefore, based on current evidence, barbiturate coma should be avoided until all
other measures for controlling elevated ICP are exhausted.
Conclusions
There is conflicting evidence regarding the efficacy of pentobarbital over conventional ICP
management measures.
There is Level 2 evidence that thiopental is more effective than pentobarbital for controlling
unmanageable refratory ICP.
There is Level 2 evidence that pentobarbital is no better than mannitol for the control of
elevated ICP.
There is Level 4 evidence that barbiturate therapy may cause reversible leukopenia,
granulocytopenia, and systemic hypotension.
Based on a single study, there is Level 4 evidence that a combination barbiturate therapy and
hypothermia may result in improved clinical outcomes up to 1 year post-injury.
There are conflicting reports regarding the efficacy of pentorbarbital for the control of
elevated ICP
Thiopental is beter than pentobarbital for controlling unmanageable refractory ICP.
Pentobarbital is not better than mannitol for the control of elevated ICP.
Barbiturate therapy plus hypothermia may improve clinical outcomes.
Patients undergoing barbiturate therapy should have their immunological response
and systemic blood pressure monitored.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.1.2.6 Cannabinoids
Dexanabinol (HU-211) is a synthetic, non-psychotropic cannabinoid (Mechoulam et al., 1988),
thought to act as a non-competitive N-methyl-D-aspartate receptor antagonist (Feigenbaum et
al., 1989) to decrease glutamate excitotoxicity. This drug is also believed to possess antioxidant
properties (Eshhar et al., 1995). Dexanabinol has shown very encouraging neuroprotective
effects in animal models of TBI (Shohami et al., 1995).
The AANS and the EBIC make no recommendations regarding cannabinoids.
Individual Studies
Table 16.13Cannabinoids as an Acute Therapeutic Strategy Post ABI
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
Knoller et al.,
(2002)
Israel
RCT
PEDro = 22
N=67 Severe head injury patients (GCS
4-8) were randomized to receive
intravenous dexanabinol (50 mg or 150
mg dexanabinol/1 mL of Cremophorethanol was diluted in 100 mL of saline)
or placebo (vehicle) by fast infusion
(over 15 min). ICP, cardiovascular
function (HR, mean arterial BP, CPP and
electrocardiogram) were compared.
Galveston Orientation and Amnesia Test
(GOAT), GOS and DRS scores up to 6
months post-injury were used as
secondary outcome measures. There
were no significant differences between
drug and vehicle groups in the
distribution of sex, age, injury cause,
injury type (single or multiple), mean
time to treatment, and GCS scores.
ICP in drug treated group decreased
significantly on day 2 and 3 (p< 0.02 and
p<0.005 respectively). ICP control
achieved without lowering systemic BP.
Significant reduction percentage of time
CPP was < 50 mmHg in the drug treated
group on days 2 & 3 (p < 0.05). No
significant differences in mortality rates
between groups (p = 0.54). On the GOS,
drug treated patients improved faster
than controls and a significantly higher
percentage achieved good recovery
compared with controls at 1 month
(p=0.04), with a similar trend at 3
months (p=0.1) post-injury. Similarly on
the DRS, a higher proportion of drug
treated patients achieved no disability
compared with controls. Trend toward
better scores on the GOAT in the drug
treated group compared to the placebo
group throughout the 6 months follow
up period, although the % of patients
with a GOAT score of 70 (maximum =
100) plateaued at 3 months.
Maas et al.,
(2006)
The Netherlands
RCT
PEDro = 10
N=861 Severe brain injury patients (GCS
motor score 2-5) were randomly
assigned to a single intravenous
injection of 150 mg dexanabinol or
placebo given within 6 hours of injury.
The extended GOS scores at 6 months
did not differ between groups (p=0.78):
50% of patients in the dexanabinol
group and 51% of those in the placebo
group had an unfavourable outcome
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro Score
Methods
Outcome
Trial drug was given by infusion over 15
min. ICP and CPP were measured hourly
for the first 72 hours. The primary
outcome was the extended GOS at 6
months. The Barthel Index (BI) and
measures of quality of life (SF36 and the
community reintegration questionnaire
(CIQ)) were also assessed at 6 months.
(odds ratio for a favourable outcome
was 1.07; 95% CI 0.83 – 1.39). There
were no differences in mortality,
occurrence of neuroworsening, or in
events related to recovery between
groups. No beneficial effects of
dexanabinol for improving control of ICP
and CPP, BI or on quality of life
measures (SF36, CIQ).
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
Knoller et al. (2002) randomly assigned 67 severe brain injury patients to receive dexanabinol
(50 or 150 mg) or placebo. Their findings were encouraging in that the active drug group
showed significant improvements in ICP and CPP. However, despite showing initially significant
improvements on the GOS and DRS at 1 month post-treatment, these benefits progressively
lost significance over the 6-month follow-up (Knoller et al., 2002).
Recently, Maas et al. (2006) conducted a large-scale multi-centre RCT to conclusively establish
the efficacy of dexanabinol in the treatment of ABI. In this study, 861 severe brain injury
patients admitted to 86 different centres from 15 countries were randomized to receive
dexanabinol or placebo within 6 hours of injury. The authors reported that compared with
placebo treatment, dexanabinol did not significantly improve outcomes on the extended GOS,
mortality rates, Barthel index, or quality of life measures (SF36, CIQ) at 6-months. Moreover,
dexanabinol failed to provide any acute control of derangements in ICP or CPP (Maas et al.,
2006). These strongly negative findings suggest that the initial benefits reported by Knoller et
al. (2002) could have simply been due to the small sample size in this earlier study. Overall, this
suggests that overt generalizations made from the findings of smaller RCTs should be
undertaken with caution.
Conclusions
Based on the findings of one large-scale multi-centre RCT, there is Level 1 evidence that
treatment with dexanabinol does not provide acute improvements in ICP or long-term clinical
benefits post-ABI.
Dexanabinol is not effective in controlling ICP or in improving clinical outcomes postABI.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.2.2.7 Corticosteroids
Numerous corticosteroids have been used in brain injury care including dexamethasone,
methylprednisolone, prednisolone, betamethasone, cortisone, hydrocortisone, prednisone and
triamcinolone (Alderson and Roberts 2005). Using such a broad spectrum of agents within
diverse patient groups has made understanding corticosteroid efficacy difficult. Adding to this
difficulty is a lack of understanding regarding the mode of steroid action. Grumme et al. (1995)
report that laboratory studies have associated reductions in wet brain weight, facilitation of
synaptic transmission, reduction of lipid peroxidation, enhanced blood flow, preservation of
electrolyte distribution, and membrane stabilization with corticosteroid use (Grumme et al.,
1995). Although it had been thought that benefits could arise from reductions in ICP as well as
neuro-protective activity, several studies have suggested some limitations in the use of
corticosteroid. Focal lesions seem to respond well to corticosteroid therapy while diffuse
intracerebral lesions and hematomas are less responsive (Grumme et al., 1995; Cooper et al.,
1979).
Questions regarding the safety of corticosteroid administration have been brought to light in
the wake of several large scale trials. Alderson and Roberts (1997) conducted a systematic
review of corticosteroid literature and concluded that there was a 1.8% improvement in
mortality associated with corticosteroid use. However, their 95% confidence interval ranged
from a 7.5% reduction to a 0.7% increase in deaths. This only added to the uncertainty around
corticosteroid safety and prompted a large multi-center trial. Roberts et al. (2004) studied
corticosteroid use in acute brain injury with the goal of recruiting 20, 000 TBI patients; after
10,008 patients were recruited it became clear that corticosteroid use caused significant
increases in mortality and the trial was halted.
The AANS stated that steroid use was not recommended for improving outcomes or reducing
ICP and that high-dose methylprednisolone was associated with increased mortality and was
contraindicated (Bratton et al., 2007c). The EBIC state that there was no established indication
for the use of steroids in acute head injury management (Maas et al., 1997).
Individual Studies
Table 16.14 Corticosteroids for the Management of Elevated Intracranial Pressure and Neuro-protection Post
ABI
Author/Year/
Country/Study
design/PEDro
Score
Roberts et al.,
(2004)
International
RCT
PEDro = 10
Methods
N=10,008 Patients with head injury
(GCS≤14) whose physician was uncertain
about administering methylprednisolone
were randomized into a treatment group
(48h administration) and a control group.
Patients were then monitored for death
Outcome
Compared with the placebo, the risk of
death was higher in the corticosteroid
group (relative risk 1.18, p=0.0001). The
relative increase in deaths due to
corticosteroids did not differ by injury
severity (p=0.22) or time since injury
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
within 2 weeks and death or disability at 6 (p=0.05).
months.
Grumme et al.,
(1995)
Germany/
Austria
RCT
PEDro = 9
N=396 Patients diagnosed with head injury
were randomized to a treatment group
(200mg triamcinolone acetonide within 4h
of trauma, then 3x40mg/day IV for 4 days,
and 3X20 mg/day for 4 days) or placebo.
Outcomes were measured using the GOS
at discharge and 1 year after trauma.
No significant difference was seen
between groups although a trend
towards improved outcomes in the
treatment groups was noted. A significant
difference was seen in the subset of
patients with GCS<8 and focal lesions
compared to placebo (p=0.0145) when
good outcomes were compared.
Dearden et al.,
(1986)
UK
RCT
PEDro = 4
N=130 Severely head injured patients
were randomly allocated dexamethasone
(50mg on admission, 100mg on days 1,2,3,
50mg on day 4 and 25mg on day 5), or
placebo. ICP and 6 month GOS scored
were measured.
Patients in the placebo group with ICP >
20mmHg showed significantly poorer
outcomes compared to similar patients in
the placebo group (p=0.0377). No other
differences were noted.
Gianotta et al.,
(1984)
USA
RCT
PEDro = 7
N=88 Patients with a GCS ≤ 8 6 hours after
nonpenetrating head trauma were given
either high dose methylprednisolone
sodium succinate (30mg/kg q6h x 2, then
25o mg q6h x 6, then tapering over 8
days), low dose methylprednisolone
(1.5mg/kg q6h x 2, then 25 mg q6h x 6,
then tapering over 8 days) or placebo.
Follow-up was performed on all surviving
patients at 6 months and were graded
according to the GOS.
At six months, no significant differences
in mortality was seen between groups.
There were also no significant differences
in morbidity between groups.
Braakman et al.,
(1983)
Netherlands
RCT
PEDro = 4
N=161 Comatose patients admitted after a No significant differences were seen in 1
non-missile related head injury were
month survival rates or 6 month GOS
randomized to receive high-dose
scores between groups.
dexamethasone or placebo. Survival at
one month and 6 month GOS scores were
used as assessments of effectiveness.
Saul et al.,
(1981)
USA
RCT
PEDro = 4
N=100 Severely brain injured patients
(GCS<8) were given methylprednisolone
(5mg/kg/day) or no drug. GCS was
measured daily while in hospital and GOS
at 6-months was used as the ultimate
outcome.
No significant difference was seen in
proportion of “good” and “disabled”
outcomes compared to “vegetative” and
“dead” outcomes between groups
(p=0.22).
Kaktis et al.,
(1980)
USA
RCT
N=76 Head injured adults who were
comatose on admission were randomly
allocated to a high-dose groups
(dexamethasone 24mg/day), low-dose
No significant differences in ICP levels
were seen between groups at any point
within the first 72 hours after injury.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
PEDro = 4
group (dexamethasone 16mg/day), or
placebo. Patients were monitored for ICP
levels in 6 hour increments up until 72
hours after injury.
Cooper et al.,
(1979)
USA
RCT
PEDro = 8
N=76 Patients with Grady Coma Grade 3-5
were stratified for severity of injury and
then divided into high dose
dexamethasone (96mg/day), low dose
dexamethasone (16mg/day) or placebo for
6 days. Outcome was was assessed at 6
months post treatment with a GOS scoring
system.
No significant improvement in outcome
was seen between groups for good
outcomes at 6 months, ICP patterns, or
serial neurological examinations in
hospital.
Watson et al.,
(2004)
USA
Cohort
N=404 Patients were included if one of the
following criteria was met: a cortical
contusion visible on CT; subdural, epidural,
or intracerebral hematoma; depressed
skull fracture; penetrating head wound;
seizure within 24h of injury; or a GCS ≤ 10
(n=125). After controlling for seizure risk,
patients treated with glucocorticoids were
compared for odds of developing first and
second late posttraumatic seizures with
those receiving no glucocorticoids.
Patients dosed with glucocorticoids
within 1 day of their TBI were more likely
to develop first late seizures than were
those without (p=0.04,hazard ratio =
1.74) Those receiving glucocorticoids ≥ 2
days post injury had no similar
associations (p=0.66, HR = 0.77).
Glucocorticoid administration was not
associated with second late seizure
development in any group.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In light of a series of inconclusive studies into the effectiveness and safety of corticosteroid use,
a very large multinational randomized collaboration for assessment of early
methylprednisolone administration was initiated in 1999 (Roberts et al., 2004). In order to
achieve 90% power, recruitment of 20,000 patients in the Corticosteroid Randomization after
Severe Head Injury (CRASH) trial was the goal. After the random allocation of 10,008 patients,
the experiment was halted. Of 4,985 patients allocated corticosteroids, 1052 died within 2
weeks compared to 893 of 4979 patients in the placebo group. This indicated a relative risk of
death equal to 1.8 in the steroid group (p=0.0001). Further analysis showed no differences in
outcomes between 8 CT subgroups or between patients with major extracranial injury
compared to those without. The authors also conducted a systematic review and meta-analysis
of existing trials using corticosteroids for head injury. Before the CRASH trial, a 0.96 relative risk
of death was seen in the corticosteroid group. Once the patients from the CRASH trial were
added the relative risk changed to 1.12. The authors suggest that based on this large
multinational trial, corticosteroids should not be used in head injury care no matter what the
severity of injury.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Two other studies assessed methylprednisolone in ABI management. Giannotta et al. (1984)
conducted an RCT of patients with GCS ≤ 8 treated with methylprednisolone. Patients were
divided into one of three groups: a high dose, low dose or placebo group, then assessed at 6
months based on the GOS grading system. They reported no differences in mortality rates
between groups. The authors then compressed the low dose and placebo groups and
performed further analyses. They found that patients less than 40 years old in the high dose
group showed significant decreases in mortality when compared to the low dose/ placebo
group. However, they also found no significant differences between these groups in beneficial
outcomes. Even if the decreases in death are taken into account, the authors point out that
decreasing mortality without decreasing morbidity may not be valuable. Saul et al. (1981)
conducted another RCT where patients received methylprednisolone or no drug at all. They
noted that there were no differences between the two groups for 6 month GOS scores.
Grumme et al. (1995) conducted an RCT in Germany and Austria in which GOS scores were
assessed 1 year after injury in patients treated with the synthetic corticosteroid triamcinolone.
While no overall effect between groups was found, further analysis was performed on subsets
of patients. A significant increase in beneficial outcomes was seen in patients who had both a
GCS<8 and a focal lesion. The authors suggest that in light of this evidence, patients with both
GCS<8 and a focal lesion would benefit from steroid administration immediately after injury.
Four randomized trials were found that assessed dexamethasone in ABI. Dearden et al. (1986)
assessed consecutively admitted head injured patients treated with dexamethasone. They
noted that patients experiencing ICP levels > 20mmHg showed significantly poorer outcomes on
the 6 month GOS scores. Braakman et al. (1983) found no differences between patients treated
with dexamethasone compared to placebo in 1 month survival rates or 6 month GOS scores.
Similarly, Cooper et al. (1979) performed a double blind randomized controlled study of the
effects of dexamethasone on outcomes in severe head injuries. Patients were divided into three
groups and no significant differences were seen in outcomes. The authors performed several
post-mortem examinations and indicate that often, patients initially diagnosed with focal
lesions were in fact suffering from diffuse injuries which are not amenable to corticosteroid
treatment. Finally, Kaktis & Pitts (1980) assessed the effects of low-dose (16mg/day) and highdose (14mg/day) dexamethasone on ICP levels in brain injured patients. They noted no
differences in ICP at any point during the 72 hour follow-up period.
In a cohort study conduced by Watson et al. (2004) patients receiving any form of
glucocorticoid therapy (dexamethasone 98%, prednisone 2.4%, methylprednisone 1.6%, or
hydrocortisone 1.6%) were compared too patients treated without corticosteroids for risk of
development of post-traumatic seizures. Their inclusion criteria allowed for patients with only
one of a list of complications to be included resulting in a diverse group of TBI patients. They
noted that patients receiving glucocorticoid treatment on the first day post injury were at
increased risk of developing first late seizures compared to patients receiving no treatment.
They also saw no improvement in patients receiving glucocorticoids after the first day. The
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
authors suggest that this ads further strength to the argument against routine corticosteroid
use in TBI (Watson et al., 2004).
Conclusions
There is Level 1 evidence that methylprednisolone increases mortality rates in ABI patients
and should not be used.
There is Level 2 evidence that triamcinolone may improve outcomes in patients with a GCS<8
and a focal lesion.
There is Level 1b evidence that dexamethasone does not improve ICP levels and may worsen
outcomes in patients with ICP > 20mmHg.
There is Level 3 evidence that glucocorticoid administration may increase the risk of
developing first late seizures.
Methylprednisolone increases mortality rates in ABI patients and should not be used
Triamcinolone may improve outcomes in patients with a GCS<8 and a focal lesion
Dexamethasone does not improve ICP levels and may worsen outcomes in patients with ICP >
20mmHg
Glucocorticoid administration may increase the risk of developing first late seizures
16.1.2.8 Progesterone
Progesterone has drawn interest as a potential neuroprotective agent. Animal studies suggest
that progesterone modulates excito-toxicity, reconstitutes the blood brain barrier, reduces
cerebral edema, regulates inflammation, and decreases apoptosis (Stein, 2008). In the human
population, Groswasser et al. (1998) observed that female TBI patients seemingly recovered
better than male patients and progesterone was suggested as a possible cause of this disparity
Trials are now being undertaken to accurately assess the effects of progesterone in the ABI
population.
The AANS and the EBIC made no recommendations regarding progesterone use in ABI
treatment.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.15Progesterone for Treatment of Acute ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Wright et al.
(2007)
USA
RCT
PEDro=10
N=100 Adult TBI patients (GCS 4-12) who
arrived within 11 h after injury were
randomized 4:1 to IV progesterone or
placebo. Treatment patients received 0.71
mg/kg progesterone at 14mL/h for 1 h, then
0.5mg/kg at 10 mL/h for 11 h, and then
10mL/h maintenance infusions every 12h to
a total of 3 days treatment. Patients were
assessed for adverse event rates, 30-day
mortality, and 30 day GOS-E scores.
Adverse event rates were similar
between groups and no serious adverse
events were associated with
progesterone. Patients in the
progesterone group had lower 30-day
mortality rates (RR 0.43; 95%CI 0.18 –
0.99). Moderately severe patients (GCS
9-12) in the progesterone group were
more likely to have a moderate to good
recovery on GOS-E (p=0.0202).
Xiao et al.
(2008)
China
RCT
PEDro=7
N=159 Patients with severe TBI (GCS3-8)
were prospectively randomized to receive
progesterone (1.0 mg/kg via intramuscular
injection b.i.d.) or placebo. Neurological
outcome was measured using the GOS.
Modified FIM and mortality rates were also
evaluated.
Patients receiving progesterone showed
more favourable outcomes on the GOS
at 3 months (p=0.034) and 6 months
(p=0.048). Progesterone patients also
had higher mFIM scores (p<0.05 and
p<0.01) and lower mortality rates
(p<0.05) at 3 and 6 month follow-ups.
No instances of complications were
found after progesterone
administration.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
Two RCTs were identified that assessed progesterone use for the treatment of acute ABI.
Wright et al. (2007) conducted a phase II clinical trial of progesterone for care of moderate and
severe ABI patients (GCS 4-12) in response to positive clinical observations and animal trials. As
a phase II trial, the initial goal was to assess the safety of progesterone administration. For this
purpose, patients were allocated 4:1 to the progesterone group compared to the placebo
group. Patients were monitored for any complications so that inter-group comparisons could be
made. Patients in the progesterone group showed no increase in complication rates and a
decreased 30-day mortality rate. Moderately severe patients in this group also showed
significantly greater rates of moderate to good GOS-E scores. The authors point to limitations in
sample size and group distribution as cautioning factors but feel the results are encouraging
and warrant a larger, more thorough clinical trial.
More recently, Xiao et al. (2008) conducted a placebo controlled RCT of progesterone use in TBI
patients in China. Patients received a 5-day course of progesterone during acute management.
They reported significant improvements in GOS scores, mFIM scores at 3 and 6-month followModule 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
up and decreases in mortality rates at 6 months. They also reported no complications
associated with progesterone administration.
These two studies suggest that progesterone is safe and effective improving patient outcomes
after ABI. Further study should be performed to verify these results and identify specific
indications for its use.
Conclusions
There is Level 1 evidence that progesterone improves GOS and modified FIM scores, and
decreases mortality rates in ABI patients.
Progesterone decreases 30-day mortality rates.
Progesterone improves GOS and modified FIM scores at 3 and 6 months post-injury.
16.1.2.9 Bradykinin Antagonists
Pharmacological interventions for the treatment of TBI generally neglect the acute phase of the
injury, which is marked by an acute inflammationi that contributes to the pathology of TBI.
Research has shown that any type of tissue injury or death act as strong triggers for the
initiation of an inflammatory response. The kinin-kallikrein pathway is one of the components
of this acute inflammatory cascade following brain injury (Marmarou et al., 1999; Narotam et
al., 1998). The generation of bradykinin from this pathway leads to a detrimental cascade of
events ultimately ending in altered vascular permeability and tissue edema (Francel, 1992). Upregulation of kinins following concussive brain injury in rats and blunt trauma in humans has
been reported, emphasizing their importance in the pathophysiology of brain injury. Recent
animal research using BK2 receptor knockout mice has demonstrated direct involvement of this
receptor in the development of the inflammatory induced secondary damage and subsequent
neurological deficits resulting from diffuse TBI (Hellal et al., 2003). These findings strongly
suggest that specific inhibition of the BK2 receptor could prove to be an effective therapeutic
strategy following brain injury.
Bradycor is a bradykinin antagonist which acts primarily at the BK2 receptor (Marmarou et al.,
1999; Narotam et al., 1998) making it attractive for the management of inflammation post-ABI.
Anatibant is another BK2 receptor antagonist that is believed to more strongly bind to these
receptors (Marmarou et al., 2005). Animal research has suggested that anatibant dampens
acute inflammation, reduces brain edema, and improves long-term neurological function (Hellal
et al., 2003; Kaplanski et al., 2002; Pruneau et al., 1999; Stover et al., 2000).
The AANS and EBIC make no recommendations regarding bradykinin antagonists.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Individual Studies
Table 16.16Bradykinin Antagonist as an Acute Therapeutic Strategy Post ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Shakur et al.,
(2009)
UK
RCT
PEDro = 9
N=219 Subjects with traumatic brain injury
(GCS ≤12) patients aged 16-65 who had
received their injury less than 8 hours
prior to examination. Patients were
randomly administered a placebo or
anatibant in a high (30mg loading dose
and 15mg/day), medium (20mg loading
dose and 10mg/day) or low dose (10mg
loading dose and 5mg/day). For 4 days a
maintenance doses was administered.
Patients were followed-up for serious
adverse events (SAE), mortality and
morbidity (GCS, DRS, modified Oxford
Handicap Scale score) 15 days post event.
Due to patients’ safety issues, the tril
ended early. In patients (26.3%) treated
with anatibant serious adverse side
effects were noted compared to those
treated with placebo (19.3%). Analyses
suggested that anatibant was not
significantly related to causing a serious
adverse side effect within 2 weeks
(p=.19). Difference in GCS and Disability
Rating Scale outcomes were not
statistically significant (p>0.05).
Marmarou et al.,
(2005)
USA
RCT
PEDro = 4
N=25 Severe head injury patients (GCS 38) were randomized to receive a single
dose of anatibant (LF16-0687Ms at 3.75 or
22.5 mg), a selective and potent
antagonist of the BK B2 receptor or
placebo. Injections were administered
within 8 hr of injury or 12 hr if surgery was
required. GOS at 1, 3 and 6 months postinjury, maximum ICP during the first 5
days, and CPP were compared.
Given small sample size and the fact
that groups were not comparable at
entry, no conclusion could be drawn
with respect to the efficacy of anatibant
in preventing brain edema, increased
ICP or decreased CPP. More patients
showed improvement and favorable
outcomes (good outcome/moderate
disability) on the GOS at 3 and 6 months
among those treated with anatibant
22.5 mg compared with the other
groups (anatibant 3.75 mg or placebos).
Marmarou et al.,
(1999)
USA
RCT
PEDro = 8
N=139 Patients with severe TBI (GCS 3-8)
were randomized to receive either
Bradycor (deltibant, CP-1027 at 3
µg/kg/min) or placebo as a continuous
intravenous infusion for 5 days with the
infusion beginning within 12 hrs of the
injury. ICP and Therapeutic Intensity Levels
(TIC – the need for therapeutic
interventions to control ICP) changes were
compared. Long-term outcome assessed
at 3 and 6 months after injury using the
GOS.
Percentage of time ICP of > 15 mmHg on
days 4 and 5 was significantly lower in
the Bradycor group compared with
placebo (p=0.035). There were fewer
deaths in the Bradycor group (20% vs.
27% in placebo). Patients in the
Bradycor group showed a 10.3% and
12% improvement in GOS at 3 and 6
months respectively (p=0.26). Patients
treated with Bradycor required less
therapy for the control of ICP (mannitol,
pressors, barbiturates, ventricular
drainage, hyperventilation) although
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
this effect was not statistically
significant.
Narotam et al.,
(1998)
South Africa
RCT
PEDro = 6
N=20 Patients with focal cerebral
contusions presenting within 24-96 hours
of closed head injury (GCS 9-14) were
randomized to receive a 7-day infusion of
TM
CP-0127 (Bradycor 3.0 µg/kg/min) or
placebo (lactated Ringer’s solution).
Therapy Intensity Level (TIC – the need for
therapeutic intervention to control ICP),
changes in ICP, and neurological function
as measured by the GCS scores were
compared between groups.
No differences in age, baseline GCS,
initial ICP between groups was found.
However, the CP-0127 group had a
longer interval from time of injury to
initiation of drug infusion (p=0.027).The
mean rise in peak ICP from baseline was
greater in the placebo group than with
CP-0127 (p=0.018). The mean
deterioration in GCS score in the
placebo group was significantly greater
than in the CP-0127 group (p=0.002).
CP-0127 had a significant effect in
preventing elevation of ICP (10/11
patients in contrast to the placebo
group where elevated ICP occurred in
7/9 patients, p=0.005). Clinically
significant neurological deterioration
(higher TIC) occurred more in the
placebo group than in the CP-0127
group (77% vs. 9%, p=0.005).
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
We identified two studies that evaluated the efficacy of Bradycor in the acute treatment of ABI.
In the first of these studies, Narotam et al. (1998) randomly assigned a small number of
patients to receive bradycor or placebo for 7 days. They reported that treatment with bradycor
resulted in a significant reduction in ICP elevations. Moreover, compared with the bradycor
group, patients in the placebo group experienced a greater deterioration in GCS scores over the
course of the study. Furthermore, the need for other therapeutic interventions to control ICP
was markedly lower in those who received bradycor (Narotam et al., 1998).
A larger RCT conducted by Marmarou et al. (1999) partially confirmed the therapeutic efficacy
of bradycor for ABI. Marmarou et al. (1999) reported that compared with those assigned to a
placebo group, patients in the bradycor group experienced a significant reduction in intracranial
hypertension (ICP > 15 mmHg). However, despite trends favouring the bradycor treatment,
there were no significant differences from those treated with the placebo in mortality rates,
improvements in GOS scores at 3 and 6 months, or the intensity of therapeutic interventions
needed to control ICP (Marmarou et al., 1999).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
In a more recent study by Marmarou and colleagues (2005), severe head injury patients were
randomized to receive anatibant (3.75 or 22.5 mg) or placebo. Anatibant is believed to be a
more potent bradykinin antagonist, and this pilot trial was likely inspired by the mixed results
obtained for the therapeutic efficacy of the less potent bradycor. However, due to a small
sample size and a lack of baseline comparability between groups in this study, the authors were
unable to draw any conclusion to support or refute the efficacy of anatibant in preventing brain
edema, or deteriorations in ICP and CPP (Marmarou et al., 2005). The authors suggested that a
larger trial of anatibant is likely warranted as a greater number of patients who received the
higher dose of this drug showed favourable outcomes on the GOS at 3 and 6 months compared
with the other groups (Anatibant 3.75 mg or placebos) (Marmarou et al., 2005).
A large-scale multi-center trial of anatibant was therefore conducted by Shakur et al. (2009).
Four-hundred patients within 8 hours of experiencing a severe brain injury were to be
randomized to receive one of three doses of anatibant or placebo. However, the trial was
halted after randomization of 228 patients due to an elevated risk of serious adverse events
among patients receiving anatibant. This study appears to have been extremely problematic. A
number of treatment protocols were breached and a dispute between the sponsors and the
clinical trial team were ultimately dealt with in the High Court of Justice (Shakur et al., 2009).
Analysis of results from those patients who were included suggested higher levels of serious
adverse events in patients receiving anatibant with no improvement in mortality or morbidity.
However, the authors report that there were no SAEs suspected to be related to the study drug
as judged by the investigators and call for a larger trial to be conducted.
Conclusions
Based on the findings of two RCTs, there is Level 1 evidence that Bradycor (a bradykinin
antagonist) is effective preventing acute elevations in ICP post-ABI.
There is conflicting evidence to support the use of bradykinin antagonists to improve
functional clinical outcomes such as the GOS.
Some bradykinin antagonists prevent acute elevations in ICP but their effects on longterm clinical outcomes are uncertain.
16.1.2.10 Dimethyl Sulfoxide
Dimethyl sulfoxide (DMSO) has been suggested for the treatment of elevated ICP post-ABI. It
causes a strong diuresis, protects cells from mechanical damage, reduces edema in tissue
through its ability to stabilize cell membranes, and is believed to act as a free radical scavenger
(Kulah et al., 1990). DMSO is also believed to increase tissue perfusion thereby improving cell
oxygenation, neutralizing metabolic acidosis and decreasing intracellular fluid retention (Kulah
et al., 1990).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
The AANS and the EBIC make no recommendations regarding dimethyl sulphoxide.
Individual Studies
Table 16.17 DMSO as an Acute Therapeutic Strategy Post ABI
Author/Year/
Country/Study
design
Methods
Outcome
Kulah et al.,
(1990)
Turkey
Case Series
N=10 Severe closed head injury patients
(GCS ≤ 6) who presented to hospital within 6
hours of injury received an intravenous bolus
infusion of DMSO (50cc DMSO in 5%
dextrose) when ICP was ≥ 25 mmHg. ICP,
CPP, and arterial pressure were assessed pre
and post-treatment.
3 of the 10 patients died due to
uncontrolled ICP. In most cases DMSO
reduced raised ICP within 10 minutes
after the onset of infusion with a
parallel increase CPP. DMSO had no
effect on systemic blood pressure.
There were no cardiac or circulatory
disturbances following the injection of
DMSO. However, DMSO caused only a
temporary decrease in ICP as
continuous infusions of DMSO (up to 7
days) did not prevent the ICP from
returning to elevated baseline levels.
Marshall et al.,
(1984)
USA
Collection of
case studies
No Score
N=5 Severe head injury patients and 1
patient with a cortical venous thrombosis in
whom ICP could not be controlled using
standard methods (head elevation,
hyperventilation, ventricular drainage,
mannitol, barbiturates) received a rapid
infusion of 10% DMSO at a dose of 1 g/kg or
20% DMSO titrated against the ICP rather
than as a bolus infusion. In both cases, an
upper dose limit of 8 g/kg/day was sought.
All patients showed satisfactory control
of elevated ICP (defined as a reduction
of ICP to < 25 mmHg for more than 15
minutes) within minutes (range 2-24
minutes) after DMSO administration.
However, despite initial improvements
in ICP, over time, fluid overload, severe
electrolyte disturbances, and an
ultimate loss of ICP control occurred.
Most patients experienced significant
hypernatremia as a side effect.
Karaca et al.,
(1991)
Turkey &
Canada
Case Series
N=10 Severe head injury patients (GCS ≤ 9)
with elevated ICP received DMSO every 6
hours for 1-10 days (28% solution diluted
with physiological saline to give a final dose
of 1.2 g/kg delivered intravenously). 4 of the
10 patients also received oxygen
intermittently for the first 24 hours. Acute
changes in ICP and Neurological assessment
at 6 days and 3 months after were
evaluated.
All patients showed a reduction in ICP
after 24 hours and 7 had normal ICP
after 6 days of treatment. Although
reductions in ICP were seen within the
first 30 min after DMSO administration,
the effect was not sustained and most
patients required maintenance doses for
2-10 days to minimize fluctuations in
ICP. Neurological assessment at 6 days
showed 2 patients with severe
neurological deficits, 2 with moderate
impairments, and 6 patients with mild
to no deficit. After a 3 month follow-up,
1 patient remained severely impaired
and 7 patients showed mild to no
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design
Methods
Outcome
deficit.
Discussion
We identified three studies examining the effects of DMSO in the management of ICP and brain
swelling post-ABI. In a study using a single group intervention design conducted by Kulah et al.
(1990) 10 severe brain injury patients with elevated ICP were treated with a single bolus
injection of DMSO within 6 hours of injury. The authors reported that in the majority of cases,
DMSO was effective in controlling ICP elevations within minutes of injection (Kulah et al., 1990).
This was followed by a concomitant increase in CPP. Unfortunately, these benefits appear to be
only transient, since continuous infusions of DMSO for up to seven days failed to control
elevations in ICP.
In a similar study conducted by Karaca et al. (1991) 10 severe head injury patients were treated
with repeated injections of DMSO for up to 10 days. The authors reported that although
reductions in ICP were seen within the first 30 min after DMSO administration, the effect was
not sustained and most patients required maintenance doses for 2-10 days to minimize
fluctuations in ICP.
The findings of Marshall et al. (1984) similarly suggest that rapid infusions of DMSO are
effective in controlling elevated ICP in patients in whom ICP cannot be controlled by standard
measures.
Conclusions
There is Level 4 evidence that dimethyl sulfoxide transiently reduces ICP elevations.
Dimethyl sulfoxide may cause temporary reductions in ICP elevations post-ABI.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.2 Prompting Emergence from Coma
16.2.1 Non-Pharmacological
16.2.1.1 Sensory Stimulation
It has been reported that one in eight patients with severe closed head injury suffers from
prolonged coma and vegetative state following their injury (Levin et al., 1991). It has also been
estimated that 50% of vegetative survivors from severe brain injuries regain consciousness
within one year of their injury with up to 40% subsequently improving to a higher level on the
Glasgow Outcome Scale (Task Force 1994). The idea that sensory stimulation could enhance the
speed and degree of recovery from coma has gained popularity. Early studies employed single
stimuli to a single sense (unimodal stimulation), whereas more current studies have focused on
sensory stimulation to all the senses using various stimuli (multimodal stimulation).
Neither the AANS nor the EBIC make recommendations regarding sensory stimulation in
comatose brain injured patients.
Individual Studies
Table 16.18 Sensory Stimulation for the Management of Patients in a Coma or Vegetative State Post ABI
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Abbasi et al.,
(2009)
RCT
Iran
PEDro = 7
N=50 Comatose head injury patients (GCS
6-8) were assigned to receive a regular
family visiting program or routine care.
Family visits were 15 mins long for each of
6 days and were structured to include
affective, auditory and tactile stimulation
by family members. Patients were
evaluated using the GCS at baseline and 30
minutes after each family visit by a double
blinded nurse.
Patients receiving family visits showed
significant increases in GCS scores on
each day during the study period. After
six days, GCS scores in the intervention
group were significantly higher (8.8 vs.
6.8, p=0.0001)
Johnson et al.,
(1993)
RCT
UK
PEDro = 3
N=14 Comatose severe brain injury patients
(GCS ≤ 8) were randomized (within 24 hrs
of hospital admission) to one of 2 groups.
The experimental group received
stimulation of five senses (olfactory, visual,
auditory, gustatory, tactile) for 20 min/day
for all of their stay in the ICU (medium stay
8.1 days) while the control group received
no stimulation. GCS, state of ventilation,
spontaneous eye movements,
oculocephalic response, oculovestibular
response were assessed daily.
3-methoxy 4-hydroxyphenylglycol levels
were significantly higher in the sensory
stimulation group post-treatment
(p<0.006). No significant group
differences post-treatment were seen in
heart rate (p<0.499), or skin conductance
(p<0.092). No data was provided on
changes on the main outcome measure
(GCS).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Catecholamine, serotonin,
acytylcholinesterase, 3 methoxy 4hydroxyphenylglycol, skin conductance,
heart rate were assessed 20 min pre and
post treatment periods.
Mitchell et al.,
(1990)
UK
Non-RCT
N=24 Comatose severe head injury patients
(GCS 4-6) were assigned to 2 groups
matched for age, gender, type and location
of injury, and GCS score at admission.
Group I received Coma Arousal Procedure
(CAP) involving vigorous sensory
stimulation including auditory, tactile,
olfactory, gustatory, visual, kinesthetic
propioceptive and vestibular stimulations in
a sequential order. Each of the 5 sensory
modalities were stimulated in a cyclical
manner for about 1 hour 1-2 times/day.
Group II acted as the control, receiving no
sensory stimulation at all. Duration and
degree of coma and the GCS were used to
assess outcomes.
Duration and degree of coma and the
GCS were used to assess outcomes. The
duration of coma for the CAP group was
significantly shorter than the total
duration for the control group (p<0.05).
Davis and
Gimenez
(2003)
USA
Non-RCT
N=12 Comatose severe brain injury male
patients (GCS ≤ 8) who had suffered a their
brain injury at least 3 days earlier and who
had a Rancho Los Amigos (RLA) score
between Level I and Level III (unresponsive
to sensory stimuli or responding at a low or
inconsistent level to sensory stimuli) were
assigned to a structured auditory sensory
stimulation program including 1)
orientation and commands, 2) bells, blocks
and claps, 3) music, 4) familiar voices, and
5) television or radio or to receive no
stimulation. Participants received 5-8
stimulation sessions/day lasting 5-15 min
each depending on type of stimuli for up to
7 days. Outcomes were assessed using the
GCS, sensory stimulation assessment
measure (SSAM), RLA scale, DRS.
Mean daily GCS scores were not
statistically different between groups
(however, the treatment group’s were
lower and rose over time, while the
control group’s was higher and decreased
over time). Difference between groups
on SSAM scores before and after arousal
was statistically significant (p = .015). RLA
change score for the intervention group
was 1.2, but no change was found in the
control group. DRS improvements from
baseline to discharge were significantly
better in the intervention group
compared with the control group (p =
0.0005).
Kater
(1989)
USA
Non-RCT
N=30 Acute brain injury patients known to
have experienced impaired cognition
(cognitive functioning level < 8 in the
Rancho Los Amigos scale) and hospitalized
for a minimum of 2 weeks prior to entering
the study were assigned to receive
Patients in the experimental treatment
group had a significantly improved
cognitive level score 3 months post injury
compared with controls (6.33 vs. 4.40, p
< 0.05). Cognitive functioning level varied
inversely with coma severity (based on
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
controlled structured sensory-stimulation
(visual, auditory, olfactory, gustatory,
tactile, and kinesthetic stimulation for 45
min 2x/day and 6days/week for 1-3
months) or nursing care that did not
include planned structured sensory
stimulation. Patients in both groups were
matched one to one on the basis of sex,
age, approximate type of injury, GCS score,
and length of time since injury. The Rancho
Los Amigos Scale (RLA) was used for
outcome assessment 3 months post injury.
initial GCS score). Subjects with moderate
and deep coma severity (GCS 3-10)
seemed to benefit the most from the
sensory stimulation, while subjects with
light coma severity showed little
difference from the corresponding
control subjects. The type of pre-injury
environment was also correlated with
outcome 3 months after treatment, with
subjects coming from enriched
environments showing significantly
improved cognitive levels compared with
subjects coming from non-enriched
environments (p<0.05).
Hall et al.,
(1992)
Canada
Non-RCT
N=6 Comatose severe closed head injury
patients (GCS ≤ 8) who acted as their own
control received alternating weeks for 30
min per day of Specific Directed Stimulation
(SDS) involving multisensory input at the
subject’s level of responding and nondirected stimulation (NDS) involving
stimulation not specific to the patient’s
level of responding. Eye opening, motor
movements, vocalizations/verbalization
were rated using the Rader Scale. GCS,
Rancho Los Amigos levels, and the Western
Neuro Sensory Stimulation Profile (WNSSP)
were also rated.
Quality of responses was greater during
the SDS condition. Subjects obtained
higher Rader scores for eye movements
during the SDS condition than during the
NDS condition. General improvement on
the WNSSP over the course of the
treatment for both conditions, with
subjects improving from 20% to 80% by
the end of the treatment. General
improvement on the GCS and Rancho Los
Amigos levels under both conditions over
the course of the treatment.
Gruner and
Terhaag
(2000)
Germany
Pre-Post
N=16 Severe head trauma patients (GCS <
8) with coma for at least 48 hours received
multimodal early onset sensory stimulation
(acoustic, tactile, olfactory, gustatory, and
kinesthetic) administered daily in 2 units of
1 hour each. Vegetative parameters (e.g.
heart and respiratory frequency) were
assessed pre and post treatment.
Significant changes in heart and
respiratory frequencies werenoted, with
the most significant changes being found
following tactile and acoustic stimulation.
No statistical comparisons were
reported.
Wilson et al.,
(1996)
UK
Pre-Post
N=24 ABI patients who were in vegetative
state were subjected to multimodal (stimuli
to each of the senses in turn during each
treatment session) and unimodal (stimuli to
one sensory modality only within a
treatment session) stimulation. Familiar
personal items (i.e. favorite perfume,
favorite sound) were also used with both
multimodal and unimodal stimulation.
Frequency with which eyes were
observed opened increased significantly
following multimodal (p<0.001) and
multimodal familiar (p<0.05) stimulation;
whereas no significant changes were
seen following unimodal stimulation.
Significant increases in the frequency of
spontaneous movements with eyes
opened following multimodal (p<0.005),
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
Treatments administered in 3-week blocks,
(2 treatments/day) until subjects were no
longer in vegetative state or until they were
discharged. Behaviours pre/post treatment
that suggested increased arousal were used
to assess outcome: eyes shut and no body
movement, eyes shut and reflexive body
movement, eyes shut and spontaneous
body movements, eyes open no body
movement, eyes open and reflexive body
movement, eyes open and spontaneous
body movement, engaged in activity and
vocalization.
unimodal (p<0.05) and multimodal
familiar stimulation (p<0.05). Significant
decrease in frequency of eyes shut with
no body movement following both
multimodal (p<0.005) and multimodal
familiar stimulation (p<0.05). Significant
reduction in reflexive movements with
eyes shut following multimodal
stimulation (p<0.025). Greatest changes
in behaviour achieved through
multimodal stimulation,
Wood et al.,
(1992)
USA
Non-RCT
N=8 severe closed head injury patients (GCS
9-10) were divided into 2 groups matched
for age, gender, type of injury, time since
injury, GCS scores, and Rancho Los Amigos
scores at admission. Both groups of
patients were retrospectively recruited
from the same rehab centre before (control
group) and after (experimental group) the
implementation of the specialized sensory
regulation procedure (SSRP). The SSRP
involved sensory stimulation with low
ambient noises, regular rest intervals free
from any kind of stimulation, and
appropriate interstimulus intervals during
therapy. The control group received
standard sensory stimulation in an
unregulated manner. The GCS, Rancho Los
Amigos scale were assessed at baseline (4
days following admission) and before
discharge (4 days before discharge).
Average length of stay in the
experimental group was 88.7 days
compared with 125.7 days in the control
group. The experimental group made
greater progress on GCS and Rancho Los
Amigos scores compared with the control
group. All patients in the experimental
group progressed into an acute rehab
setting compared with only one of the
control patients (the other 3 control
patients returned to their families where
they required total care from family,
visiting nurses and care attendants). No
statistical comparisons reported.
Pierce et al.,
(1990)
USA/Australia
Case-Control
N=31 Individuals, who has sustained a
severe head injury (GCS < 6) and had a
prolonged or persistent vegetative state for
at least 2 weeks were included in the study.
Patients were treated with a coma arousal
intervention involving a sequence of
vigorous multisensory stimulation
(auditory. vestibular, visual and cutaneous)
provided by close family relatives for up to
8 hrs/day 7 days/week continuing until the
patient was accepted for conventional
rehabilitation therapy. Results were
compared with those of a historical group
Within the various time periods, the
number patients who emerged from the
coma did not differ significantly between
groups. No significant differences were
found in reasonable recovery between
coma arousal group and the historical
control group (42% vs. 31%, p>0.025).
No significant improvements were noted
in either the time to obey simple
commands (p>0.2) or in the GOS (p<0.25)
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design/PEDro
Score
Methods
Outcome
of similar patients (n=135) from a previous
publication who did not receive this
intervention. Duration of coma and the
GOS at 10-12 months post-injury were
compared.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Table 16.18a Summary of Studies of Sensory Stimulation to Promote Emergence from Coma or Vegetative
State Post ABI
Authors
Methods
Results
Abbasi et al.,
N=50 Family visits program for tactile,
+ for higher GCS scores each day and overall
(2009)
auditory and affective stimulation
15min/day for 6 days vs. routine care.
Johnson et al.,
N=14 Sensory stimulation to all five
+ for higher 3-methoxy 4(1993)
senses 20 min/day for the length of
hydroxyphenylglycol levels
stay in the ICU (median 8.3 days) vs. no
sensory stimulation.
Note: No data provided on the primary
outcome (i.e. GCS)
Mitchell et al.,
N=24 Coma Arousal Sensory
+ for reduction in duration of coma
(1990)
Stimulation Procedure (auditory,
tactile, olfactory, gustatory, visual,
kinesthetic propioceptive and
vestibular stimulations) for 1 hour 1-2
times/day for up to 4 weeks vs. no
sensory stimulation at all.
Davis and
N=12 Structured auditory sensory
ND for mean daily GCS scores
Gimenez (2003)
stimulation program vs. no stimulation + for SSAM scores
for 5-8x/day for 7 days.
+ for DRS scores
Kater
N=30 Controlled structured sensory+ for RLA cognitive levels
(1989)
stimulation to all senses for 45 min
2x/day 6days/week for 1-3 months vs.
standard nursing care.
Hall et al.,
N=6 Specific Directed Stimulation (SDS) + Rader scores for eye movements
(1992)
involving multisensory input at the
ND on the WNSSP
subject’s level of responding vs. nonND on the GCS
directed stimulation (NDS) involving
ND on the RLA
stimulation not specific to the patient’s
level of responding (control) for 30
min/day.
Grunner and
N=16 Multimodal early onset sensory
+ for heart rate and respiratory changes
Terhaag (2000)
stimulation (acoustic, tactile, olfactory,
gustatory, and kinesthetic)
Note: no statistical comparisons reported for
administered daily in 2 units of 1 hour
this study.
each.
Wilson et al.,
N=24 Multimodal (stimuli to each of
+ for frequency of eye opening with
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
(1996)
the senses in turn during each
treatment session) and unimodal
(stimuli to one sensory modality only
within a treatment session) stimulation
in 3-week blocks, 2 treatments/day
until subjects where no longer in
vegetative state or until they were
discharged.
Wood et al.,
(1992)
following multimodal stimulation
+ for frequency of spontaneous movements
following unimodal and multimodal
stimulation
+ for reduction in eyes shut with no body
movements following multimodal
stimulation
+ for reduction in reflexive movements with
eyes shut following multimodal stimulation
+ for shorter length of stay
+ for greater progress on the GCS and RLA
scale
N=8 Specialized sensory regulation
procedure (SSRP) involving sensory
stimulation with low ambient noises,
regular rest intervals free from any
kind of stimulation, and appropriate
inter-stimulus intervals during therapy
Note: no statistical comparisons reported for
vs. standard sensory stimulation in an
this study.
unregulated manner (control).
Pierce et al.,
N=31 Multisensory stimulation
ND for number of patients emerging from
(1990)
(auditory. vestibular, visual and
coma; ND for time to obey simple
cutaneous) provided by close family
commands; ND for GOS scores
relatives for up to 8 hrs/day 7
days/week vs. no stimulation at all in a
retrospective group.
ND = No difference between groups; + = Improvement compared with control; - = Impairments compared with
control
Discussion
One of the major challenges for sensory stimulation to promote emergence from coma is that
outcome assessment measures are often qualitative and more difficult to assess. With this in
mind, Rader et al. (1989) developed the Sensory Stimulation Assessment Measure (SSAM),
based on the Rancho Los Amigos Levels of cognitive functioning, in an attempt to better
quantify the efficacy of sensory stimulation. This measure was used in one of the studies
located (Hall et al., 1992).
A meta-analysis of sensory stimulation for comatose or vegetative state ABI patients found only
three clinical trials that used rigid experimental designs to meet the desired inclusion criteria
(Lombardi et al., 2002). The authors noted that these studies varied widely in terms of
outcomes measured, treatment and experimental design making it impossible to carry out a
conventional quantitative synthesis of the data. Instead, this Cochrane review provided an
unconventional qualitative analysis of these studies. The authors concluded that there was no
reliable evidence to support or refute the efficacy of sensory stimulation programs for patients
in a coma or vegetative state post ABI.
Since the release of this review, few new studies have been published in this area. However, in
a well designed trial, Abbasi et al. (2009) conducted an RCT to evaluate the effect of sensory
stimulation through structured family visits on consciousness as assessed by the GCS. Families
received training on coma; how to provide appropriate stimulation and how to remain calm.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Patients receiving family visits showed significantly greater GCS scores on each day of the
intervention and attained a mean GCS that was 2 points higher than the control group.
Although no long-term outcomes were evaluated and no follow-up was reported, these results
suggest that family provided stimulation may be an effective intervention for stimulating
recovery from coma.
In the only other RCT identified on this topic, Johnson et al. (1993) randomly assigned patients
to a group that received multimodal sensory stimulation or to a group received no purposeful
sensory stimulation at all. The primary outcome in this study was changes in the GCS posttreatment. However, Johnson et al. (1993) did not report any data on this measure and only
presented data on biochemical and physiological parameters of questionable clinical
importance. The strength of the study findings have been questioned due to the “poor”
methodological score (PEDro = 3); conclusions, were not based upon the study’s findings
Overall, the studies identified in this area generally show greater improvements in a variety of
measures following multimodal sensory stimulation. Some studies aimed to investigate if the
duration of coma could be reduced using sensory stimulation as their only objective. For
example, Mitchell et al. (1990) reported that patients subjected to multimodal sensory
stimulation experienced significant reductions in the duration of coma compared with controls.
Again, duration of coma was their only outcome and in the absence of other measures of
clinical importance, such as functional indicators (i.e. GOS or DRS scores), such results fall short
in demonstrating any clinical functional benefit of sensory stimulation. A 2002 Cochrane review
showed similar results stating that there was insufficient evidence to refute or support the use
of multisensory programs for patients in coma or vegetative state (Lombardi et al., 2002).
Sensory stimulation is usually an intervention that provided in addition to standard care and yet
there have been a paucity of RCTs. However, as demonstrated by Abbasi et al. (2009) it is
feasible to randomize patients to one of these two options. Further research is needed.
Conclusions
There is Level 1b evidence that multimodal sensory stimulation provided by family members
improves consciousness of severe ABI patients with GCS 6-8.
There is Level 2 evidence to suggest that sensory stimulation may improve clinical outcomes,
physiological parameters, and behaviours indicative of emergence from coma post ABI.
Sensory stimulation provided by family members improves consciousness for patients
with GCS 6-8.
Sensory stimulation may help to promote emergence from coma or vegetative state
post ABI.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.2.1.2 Music Therapy
Musical sounds stimulate the auditory pathway and activate emotional functions in the brain.
If the music is familiar to the patient (i.e. a favorite song) then the stimuli can become
meaningful for the patient. Anecdotaly, it has been noted that music therapy could specifically
be used to encourage arousal from coma post-ABI. We identified two studies which used music
therapy as a specific treatment to encourage emergence from coma post-ABI.
The AANS and EBIC make no recommendations regarding music therapy.
Individual Studies
Table 16.19 Music and Musicokinetic Therapy for Patients with Coma or Vegetative State Post ABI
Author/Year/
Country/Study
design
Methods
Outcome
Noda et al.
(2004)
Japan
Pre-Post
N=26 patients with persistent vegetative
state (12 head trauma, 9 subarachnoid
hemorrhage, 3 stroke and 2 cases of
encephalopathy) received musicokinetic
therapy (MKT) that involves vertical
motions on a trampoline and live music in
synchrony with the vertical motion.
Following this, the patient lay on the
trampoline and received massage therapy
for 5 minutes while listening to slow music.
Patients underwent these steps for 40
minutes 1/week for 3 months. Persistent
Vegetative State (PVS) score was
determines before and 3 months after the
sessions. The PVS scoring system had been
proposed by the Society for Treatment of
Coma (Japan).
After MKT PVS scores were significantly
better compared with pre-MKT scores
(p<0.001). Patients whose brain
damage was caused by trauma or
subarachnoid hemorrhage
demonstrated larger improvements in
their PVS scores compared with patients
who suffered stroke or anoxic
encephalopathy. Time elapsed after
brain damage was not correlated to the
pre-MKT scores (p=0.873), but it was
negatively correlated to the post-MKT
scores (p=0.029). MKT was most
effective when initiated within 6 months
of injury compared to when initiated > 6
months post-injury.
Wilson et al.
(1992)
UK
Collection of
case studies
No Score
N=4 ABI patients (age ranged from 15-29)
who were in vegetative state received
various forms of stimulation over 23
consecutive days. Patients received one
stimulation session every morning and
afternoon and they received 15 sessions of
each of three treatments: unimodal (one
sense only), multimodal (all senses
stimulated) or music therapy (well
pronounced rhythm and speed in excess of
60 beats/minute from music belonging to
the patient). Pre and post-treatment
behaviour was scored as follows: eyes shut
2/4 patients showed significant
increases in eyes opened with body
movement following music therapy. 3/4
subjects also showed significant
increases in activity following
multimodal but not unimodal
stimulation.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/Study
design
Methods
Outcome
with no body movement, eyes open with
no body movement, engaging in activity.
Discussion
No RCTs were identified in this area and the only two studies included were generally of weaker
methodological quality. Noda et al. (2004), used an innovative approach to treat patients with
what they termed “musicokinetic therapy”. This involved vertical motions on a trampoline and
listening to live music in synchrony with the vertical motion. The rationale for this approach was
that this therapy should result in the activation of multiple pathways within the brain
simultaneously to more effectively promote awareness of environmental stimuli. Outcome
assessments were based on the Persistent Vegetative State (PVS) scoring system proposed by
the Society for Treatment of Coma of Japan in 1997. The authors reported significantly better
PVS scores post-treatment (Noda et al., 2004). Treatment was most effective when initiated
within 6 months of injury compared to when initiated > 6 months post-injury (Noda et al.,
2004). Despite these positive findings, the practicality of providing this therapy to multi-trauma
patients is questionable. The inherent physical stress involved with jumping on a trampoline
could potentially exacerbate physical injuries to other organs.
The other study identified was a collection of 4 case studies of patients who received a
combination of unimodal, mulitimodal and music therapy (Wilson et al., 1992). The authors
reported that half of the patients showed an increased in the frequency of behaviours
suggestive of emergence from coma following music therapy.
Conclusions
There is Level 4 evidence that music therapy as an adjunct to other modes of sensory
stimulation may be used to promote emergence from coma post ABI.
Music therapy might be useful in promoting emergence from coma post ABI.
16.2.1.3 Electrical Stimulation
Electrical stimulation is a common therapeutic approach used in the rehabilitation of a variety
of diseases of the nervous system. Some reports have proposed that electrical stimulation may
be beneficial in severely brain injured patients. It is believed that electrical stimulation applied
peripherally may stimulate the reticular activating centre and cortical areas responsible for
consciousness and arousal (Peri et al., 2001). Stimulation of the median nerve has been shown
to cause significant increments in blood flow and improved electroencephalogram activity
(Cooper et al., 1999).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
The AANS and the EBIC make no recommendations regarding electrical stimulation.
Individual Studies
Table 16.20 Electrical Stimulation Post ABI
Author/Year/
Country/Study
design/
Score
Methods
Outcome
Cooper et al.,
(1999)
USA
RCT
PEDro = 4
N=6 Comatose TBI patients (GCS 4-8) were
randomized to receive right median nerve
stimulation (asymmetric biphasic pulses at
an amplitude of 20 mA with pulse width of
300 µs at 40 Hz for 20 sec/min) or sham
stimulation for 8-12 hrs/day for a period of 2
weeks. Stimulation started when the
patient’s medical condition stabilized and
within 1 week of admission. GCS, days spent
in the intensive care unit, and GOS 1 month
post-injury were compared between groups.
At 1 week, the treated group improved
by an average of 4.0 on the GCS
compared with an average increase of
only 0.7 in the controls. By 2 weeks, the
treated group improved by an average
of 6.4 on the GCS compared with 1.3 for
the control group. The treated group
stayed in the ICU for an average of 7.7
days compared with 17.0 days for the
control group. The GOS for the treated
group averaged III compared with II for
the control group. No statistical
comparisons were reported.
Peri et al.,
(2001)
USA
RCT
PEDro = 6
N=10 Comatose non-penetrating TBI
patients (GCS 3-8) were randomized to
receive median nerve electrical stimulation
(300 ms intermittent pulses, 20 seconds on
and 40 seconds off at 40 Hz 8 hours/day for
each day in coma for up to 14 days) or sham
stimulation to the patients’ dominant arm.
GCS was used to assess the emergence from
coma (defined as GCS ≥ 9) and was
compared between groups. The GOS and
the FIM/FAM were also assessed 3-months
post-injury.
The treatment group emerged from
coma on average 2 days earlier than the
control group, however this difference
was not significant (p=0.31). There was
no significant difference between
groups in GOS or FIM/FAM scores,
although there was a trend for more
improvement in the electrical
stimulation group.
Liu et al.
(2003)
Taiwan
Pre-Post
N=6 patients (2 brain trauma, 1 aneurysm
rupture, 1 hemorrhagic stroke, 2 hypoxic
encephalopathy) received right median
nerve stimulation (asymmetric biphasic
pulses at an amplitude of 20 mA with a pulse
width of 300 µs at 35 Hz for 20 sec on/50
sec off). Stimulation was performed for 10
hours (comatose patients) or 8 hours (once a
patient became conscious) per day during
the daytime for 3 months. Cerebral
perfusion using SPECT scan and dopamine
levels were evaluated before, and one and
three months after electrical stimulation.
Significant increase in cerebral perfusion
bilaterally in all patients following
stimulation. 4 patients regained
consciousness within 35 days after initial
stimulation. Dopamine levels were
elevated in the majority of patients
following stimulation. Young patients (<
40 years of age) had better results than
older patients.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Discussion
Three studies which investigated the efficacy of median nerve electrical stimulation in
promoting emergence from coma were identified. In the first of these studies, severe TBI
patients were randomized to receive right median nerve stimulation or sham stimulation for 812 hours per day for a period of 2 weeks (Cooper et al., 1999). This study lacked statistical
comparisons and provided more of a qualitative analysis. The authors reported that the treated
group seemed to show better improvements on the GCS, GOS and showed shorter lengths of
stay in the intensive care unit compared with sham-stimulated controls (Cooper et al., 1999).
The lack of statistical group comparisons weakens any conclusions that could be drawn from
these findings.
Peri et al. (2001) randomly assigned comatose severe TBI patients to receive radial nerve
electrical stimulation or sham stimulation on their dominant arm 8 hours per day for up to 14
days. Unlike the seemingly positive findings reported by the previous study, Peri et al. (2001)
reported that median nerve electrical stimulation did not significantly improve the duration of
coma, GOS or FIM/FAM clinical scores.
The third study employed a single group intervention design and reported that median nerve
electrical stimulation caused considerable increments in cerebral perfusion which appeared to
be coupled with elevations in dopamine levels (Liu et al., 2003). Dopamine has been involved in
the regulation of consciousness (Krimchansky et al., 2004). However, the authors failed to
demonstrate a direct correlation between dopamine levels and increased levels of
consciousness.
Conclusions
There is Level 1b evidence that median nerve electrical stimulation does not improve
emergence from coma post-ABI.
Median nerve electrical stimulation does not improve emergence from coma post-ABI.
16.2.2 Pharmacological Interventions
16.2.2.1 Amantadine
Amantadine is a Dopamine agonist that acts both pre and post-synaptically to up-regulate
Dopamine activity (Meythaler et al., 2002). Dopamine is thought to be involved in frontal lobe
stimulation and plays a role in behavior, mood, language, motor control, hypothalamic function
and arousal (Sawyer et al., 2008). Amantadine was initially developed for prophylactic use as an
antiviral agent in the prevention of influenza A, but is now in common use in the treatment of
Parkinson’s disease. Amantadine’s properties as a potential neuro-active agent were quickly
recognized (Zafonte et al., 2001) and there is now interest in its use as a potentially useful
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
treatment in the management of ABI (Schneider et al., 1999). Researchers believe that
amantadine could significantly improve arousal in comatose patients. Potential side effects,
which are easily reversible, include over-stimulation, peripheral edema, livido reticularis, and
lowering of the seizure threshold (Schneider et al., 1999). The favorable risk-benefit profile of
amantedine suggests that it may be an attractive treatment option for inducing arousal from
coma (Hughes et al., 2005).
Neither the AANS nor the EBIC have made recommendations regarding amantadines use in ABI
management.
Individual Studies
Table 16.21 Amantadine for Arousal from Post ABI Coma
Author/Year/
Country/ Study
design/ PEDro Scores
Methods
Outcome
Adults
Meythaler et al.,
(2002)
USA
RCT- cross-over
PEDro=6
N=35 Patients with severe TBI related
diffuse axonal injury (GCS <11) were
randomly assigned to a placebo
controlled crossover design trial. Patients
were administered 200mg amantadine or
placebo daily for 6 weeks and then the
opposite for the next 6 consecutive
weeks. Outcome measures included the
Disability Rating Scale, Mini Mental
Status Exam, Glascow Outcome Scale,
Galveston Orientation and Amnesia Test,
and the Functional Independence
Measure (cognitive).
In group one (amantadine first), there
was an improvement in MMSE scores
of 14.3 points (p=.0185), DRS of 9.8
points (p=0.0022), GOS of 0.8 points
(p=0.0077), and FIM-cog of 15.1
points (p=0.0033) but no
improvement in the second six weeks
on placebo (p>0.05). In group two
(placebo first), there was an
improvement of MMSE of 10.5 points,
in the DRS of 9.4 points (p=0.0006),
GOS of 0.5 points (p=0.0231), and
FIM-cog of 11.3 points (p=0.003,
Wilcoxon signed rank) spontaneously
on placebo. In the second six weeks,
group two also continues to make
significant gains in MMSE (6.3 points,
p=0.409), DRS (3.8 points, p=0.0099),
and FIM-cog (5.2 points, p=0.0173
Wilcoxon signed rank).
Hughes et al.,
(2005)
Canada
Chart Review
N=123 Some patients admitted over a 10year period who remained in coma after
becoming medically stable were
administered 100-200 mg of amantadine
twice daily. Charts for these patients
were compared with charts from those
who did not receive amantadine and
emergence from coma was reviewed.
No significant difference in the
number of patients emerging from
coma was seen between groups
(p=0.42). Somatosensory evoked
potential (SSEP) was identified as a
significant predictor of emergence
(p=0.02).
Saniova et al.,
(2004)
N=74 Patients with severe head injury
(GCS<8) were retrospectively identified
Patients treated with amantadine
showed significant improvement on
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/ Study
design/ PEDro Scores
Methods
Outcome
Slovak Rep.
Chart review
as having been treated with amantadine
or not. Groups were assessed for
differences in discharge GCS and
mortality rates.
discharge GCS scores (p<0.0001) and
decreased mortality (p<0.001).
Whyte et al.,
(2005)
USA
Cohort
N=47 In this longitudinal observational
study, comatose brain injured patients
(GCS 3-8) were retrospectively reviewed
for exposure to amantadine and assessed
for impovements in Disability Rating
Score (DRS) and time until commands
were followed.
Patients receiving amantadine
showed significant improvements in
DRS scores in the first week following
administration (p<0.01) which
remained in the second week post
treatment (p=0.06).
Peadiatric Population
Vargus-Adams et al.,
(2010)
USA
RCT
PEDro = 6
N=7 Children aged 5-18years who were in
a vegetative or minimally conscious state
were involved in a study of amantadine
to determine the effects on the level of
consciousness post-ABIT. Participants
were randomly assigned to receive either
treatment or placebo. The study
proceeded over 7 weeks: 3 weeks
intervention, 1 week for washout
followed by the same 3 week
intervention. Those receiving
amantadine, a maximum dose of 4mg/kg
was given for the first week and during
the second and third week they received
a maximum daily dose of 6mg/kg. The
Coma Near Coma Scale and Coma
Recovery Scale Revised were used 3
times a week to assess the level of
consciousness of the participants.
Results found that higher doses of
amatadine is well tolerated by ABI
children and may be associated with
improving the recovery of
consciousness in ABI children who are
in a vegetative or minimally conscious
state.
McMahon et al.,
(2009)
USA
RCT
PEDro = 7
N=7 Children with ABI were randomized
to receive either amantadine or placebo
for 3 weeks followed by a wash-out week
and three weeks of the other agent.
Patients were evaluated on the
coma/near-coma scale (CNCS) and the
coma recovery scale-revised (CRS-R)
three times per week as well as weekly
subjective evaluations by arousal and
consciousness weekly.
No significant differences were noted
in the slopes of recovery between the
two agents on either of the outcome
scales. Improvements in
consciousness were noted by the
physician during weeks when
amantadine was given.
Patrick et al.,
(2006)
USA
RCT
N=25 Children and adolescents with
severe TBI (Rancho Los Amigos Scale level
<4) who remained in a low response
state at least 1 month post-injury were
The weekly rate of change was
significantly better on all three
measures on medication than off
medication (p<0.05). Rancho Los
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Author/Year/
Country/ Study
design/ PEDro Scores
PEDro=7
Methods
Outcome
randomized to receive either amantadine Amigos Scale levels also improved
or pramipexole. Subjects were evaluated significantly on medication (p<0.05).
with the Coma Near Coma Scale, Western
NeuroSensory Stimulation Profile, and
Disability Rating Scale at baseline and
weekly.
PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002).
Discussion
In the only RCT regarding amantadine’s effectiveness for improving consciousness in adults,
Meythaler et al. (2002) randomly assigned patients to receive amantadine or placebo for 6
weeks with a crossover to the other for a second 6 week period. All patients had suffered
severe TBI related diffuse axonal injuries. Patients were assessed for several indicators of
alertness and cognitive function. They found that patients who received amantadine initially
made significant gains on all outcome measures but made no further gains when they were
switched to placebo. Patients assigned to the placebo group initially made smaller but still
significant gains when on the placebo but then went on to make further improvements on the
Disability Rating Scale, Mini Mental Status Exam, Galveston Orientation and Amnesia Test, and
the Functional Independence Measure (cognitive) in the second 6 week period after
amantadine induction. The authors note that although patients who had received placebo
made some natural recovery, patients receiving amantadine made more pronounced
improvements. Furthermore, the improvements made by patients receiving amantadine in the
second 6 week period suggests that amantadine aids in recovery no matter when it is
administered.
Three other studies were located which assessed amantadine. Hughes et al. (2005) conducted a
chart review of all comatose brain injured patients admitted over a 10-year period in which
patients who received amantadine were compared with a control group of patients who did not
receive amantadine. They noted that patients receiving amandatine were no more likely to
emerge from coma (p=0.42). The authors caution of potential confounders, such as potential
selection bias, which may have affected the results. Also, the point at which a patient was
considered to have emerged from the coma was arbitrarily assessed. Whyte et al. (2005) also
conducted a retrospective review of comatose patients who received amantadine. They
isolated patients who received amantadine in weeks 4-16 post injury to assess its potential in
improving consciousness after medical stability was reached. They noted that patients who
received amantadine showed significant improvements in DRS scores one week after
administration when compared to patients treated by other methods. They also measured the
time to first response to directions in which they saw no significant difference in amantadine
patients.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Results of a chart review found patients who were treated with amantadine as a component of
standard therapy showed significant improvements in GCS at discharge and decreased
mortality rates compared to those who did not receive amantadine (Saniova et al., 2004). While
the retrospective nature of these three studies makes it difficult to draw conclusions, all three
authors suggest that amantadine is a safe option with promising potential and that further
study is warranted.
Three randomized trials of amantadine use in children were located. In a RCT, conducted by
Vargus-Adams et al. (2010) ABI children participated in a study to determine the effectiveness
of amantadine to improve the level of consciousness. It was found that amantadine may be
beneficial to improving the recovery of consciousness in ABI children. In an earlier study,
children were administered amantadine or placebo for three weeks followed by the opposite
for three weeks (McMahon et al., 2009). Although two patients drop out of the study, the
authors reported no significant differences in coma recovery during amantadine administration.
These two drop-outs along with the small number of subjects may have masked any potential
improvements. The authors suggest that amantadine did indeed show signs of improving
consciousness and should be studied further.
In the second study, Patrick et al. (2006) conducted a randomized trial in which children and
adolescents who remained in a low-responsive state 1 month post-injury were assigned to
receive amantadine or pramipexole (both dopamine agonists). Patients in both groups made
significant improvements on the Coma Near Coma Scale, the Western NeuroSensory
Stimulation Profile, and the Disability Rating Scale weekly gains. Patients also improved on
Rancho Los Amigos Scale level. There were no significant side effects to treatment which,
combined with the positive results, suggest that dopamine agonists may be a viable option for
coma arousal in children and adolescents. However, the lack of control group and small sample
size warrant further study before conclusions are drawn.
Conclusion
There is Level 1b evidence that amantadine may improve levels of consciousness and
cognitive function in patients in various stages of coma.
There is Level 1a evidence that amantadine improves the level of consciousness in children
post ABI.
There is Level 1b evidence, from one RCT, that amantadine and pramipexole improves the
levels of consciousness in TBI children and adolescents.
Amantadine may improve consciousness and cognitive function in comatose ABI patients.
Dopamine enhancing drugs may facilitate rate of recovery post traumatic brain injury in
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
children; however, due to the small sample sizes more definitive research is needed.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.3 Summary
1. There is Level 2 evidence suggesting a 30° head elevation reduces intracranial pressure
with concomitant increments in CPP.
2. There is Level 2 evidence to suggest head elevation does reduce ICP in children post
TBI; however, it was not found to have a significant impact of CPP.
3. There is Level 2 evidence to suggest hypothermia treatment helps to improve long
term outcomes post ABI.
4. There is conflicting evidence regarding hypothermia’s effect on mortality.
5. There is Level 1b evidence that systemic hypothermia is associated with an increased
incidence of pneumonia.
6. There is Level 2 evidence that the use of tromethamine, a weak base and buffer that
crosses the blood brain barrier, can offset the deleterious effects of prolonged
hyperventilation and lead to better outcomes than hyperventilation alone.
7. There is Level 4 evidence that hyperoxia can counteract the deleterious effects of
hyperventilation for the control of ICP following brain injury.
8. There is Level 4 evidence that hyperventilation below 34 torr arterial CO 2 can cause an
increase in regionally hypoperfused tissue.
9. Results from one RCT suggest there is Level 1b evidence that CSF drainage decreases
intracranial pressure in the short term.
10. There is Level 4 evidence from several studies that suggest CSF drainage does decrease
ICP in individuals post ABI
11. There is Level 1b evidence that in adults, standard trauma craniectomy is more
effective than limited craniectomy in lowering elevated ICP and leading to better GOS
outcomes at 6 months.
12. There is conflicting evidence supporting the use of decompressive craniectomies in
adults post TBI.
13. There is Level 3 evidence that resection of a larger bone flap results in greater
decreases in ICP reduction after craniectomy, better patient outcomes and leads to
fewer post-surgical complications.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
14. There is Level 1b evidence that in children, decompressive craniectomy reduces
elevated ICP.
15. There is Level 4 evidence from several studies does reduce ICP in children post severe
TBI.
16. There is Level 4 evidence that continuous rotational therapy does not worsen
intracranial pressure in severe brain injury patients.
17. There is level 4 evidence that the prone position may increase oxygenation and CPP in
ABI patients with acute respiratory insufficiency.
18. There is Level 1b evidence (from 2 RCTS) to suggest that hypertonic saline reduces ICP
more effectively than mannitol.
19. There is Level 1 evidence that treatment with hypertonic saline results in similar
clinical outcome and survival when compared with treatment with Ringer’s lactate
solution up to 6 months post-injury.
20. There is Level 1b evidence that saline solution results in decreased rates of mortality
compared with albumin.
21. There is Level 4 evidence that treatment with hypertonic saline reduces elevated ICP
refractory to conventional ICP management measures.
22. There is Level 2 evidence that hypertonic saline is similar to Ringer’s lactate solution in
lowering elevated ICP.
23. There is Level 4 evidence that hypertonic saline may be useful as a component of a
resuscitation algorithm by increasing cerebral oxygenation.
24. There is Level 1b evidence that in children, use of hypertonic saline in the ICU setting
results in a lower frequency of multiple early complications and a shorter ICU stay
compared with Ringer’s lactate.
25. There is Level 4 evidence to suggest HTS is effective in decreasing ICP levels in children
post TBI.
26. There is Level 1 evidence that sodium lactate is more effective than mannitol for the
management of acute elevations in ICP.
27. There is Level 2 evidence that higher dose mannitol is superior to conventional
mannitol in improving mortality rates, and clinical outcomes.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
28. There is Level 2 evidence that early out-of-hospital administration of mannitol does
not adversely affect blood pressure.
29. There is Level 4 evidence that mannitol is effective in diminishing intracranial
hypertension only when initial ICP values are elevated.
30. There is Level 1 evidence that propofol may help to reduce ICP and the need for other
ICP and sedative interventions when used in conjunction with morphine.
31. There is Level 2 evidence that midazolam has no effect on ICP but conflicting evidence
regarding its effect on MAP and CPP.
32. There was Level 1 evidence that bolus opioid administration resulted in increased ICP.
33. There was conflicting evidence regarding the effects of opioid infusion on ICP levels.
34. There was Level 2 evidence that remifentanil results in faster arousal compared to
hypnotic based sedation.
35. There is conflicting evidence regarding the efficacy of pentobarbital over conventional
ICP management measures.
36. There is Level 2 evidence that thiopental is more effective than pentobarbital for
controlling unmanageable refratory ICP.
37. There is Level 2 evidence that pentobarbital is no better than mannitol for the control
of elevated ICP.
38. There is Level 4 evidence that barbiturate therapy may cause reversible leukopenia,
granulocytopenia, and systemic hypotension.
39. Based on a single study, there is Level 4 evidence that a combination barbiturate
therapy and hypothermia may result in improved clinical outcomes up to 1 year postinjury.
40. Based on the findings of one large-scale multi-centre RCT, there is Level 1 evidence
that treatment with dexanabinol does not provide acute improvements in ICP or longterm clinical benefits post-ABI.
41. There is Level 1 evidence that methylprednisolone increases mortality rates in ABI
patients and should not be used.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
42. There is Level 2 evidence that triamcinolone may improve outcomes in patients with a
GCS<8 and a focal lesion.
43. There is Level 1b evidence that dexamethasone does not improve ICP levels and may
worsen outcomes in patients with ICP > 20mmHg.
44. There is Level 3 evidence that glucocorticoid administration may increase the risk of
developing first late seizures.
45. Based on the findings of two RCTs, there is Level 1 evidence that Bradycor (a
bradykinin antagonist) is effective preventing acute elevations in ICP post-ABI.
46. There is conflicting evidence to support the use of bradykinin antagonists to improve
functional clinical outcomes such as the GOS.
47. There is Level 4 evidence that dimethyl sulfoxide transiently reduces ICP elevations.
48. There is Level 1b evidence that multimodal sensory stimulation provided by family
members improves consciousness of severe ABI patients with GCS 6-8.
49. There is Level 2 evidence to suggest that sensory stimulation may improve clinical
outcomes, physiological parameters, and behaviours indicative of emergence from
coma post ABI.
50. There is Level 4 evidence that music therapy as an adjunct to other modes of sensory
stimulation may be used to promote emergence from coma post ABI.
51. There is Level 1b evidence that median nerve electrical stimulation does not improve
emergence from coma post-ABI.
52. There is Level 1b evidence that amantadine may improve levels of consciousness and
cognitive function in patients in various stages of coma.
53. There is Level 1a evidence that amantadine improves the level of consciousness in
children post ABI.
54. There is Level 1b evidence, from one RCT, that amantadine and pramipexole improves
the levels of consciousness in TBI children and adolescents.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
16.4 References
Medical aspects of the persistent vegetative state (1). The Multi-Society Task Force on PVS. N Engl J Med
1994; 330: 1499-1508.
Aarabi B, Hesdorffer DC, Ahn ES, Aresco C, Scalea TM, Eisenberg HM. Outcome following decompressive
craniectomy for malignant swelling due to severe head injury. J Neurosurg 2006; 104: 469-479.
Aarabi B, Hesdorffer DC, Simard JM et al. Comparative study of decompressive craniectomy after mass
lesion evacuation in severe head injury. Neurosurgery 2009; 64: 927-939.
Abbasi M, Mohammadi E, Sheaykh RA. Effect of a regular family visiting program as an affective,
auditory, and tactile stimulation on the consciousness level of comatose patients with a head injury. Jpn
J Nurs Sci 2009; 6: 21-26.
Adamo MA, Drazin D, Waldman JB. Decompressive craniectomy and postoperative complication
management in infants and toddlers with severe traumatic brain injuries. J Neurosurg Pediatr 2009; 3:
334-339.
Adelson PD, Wisniewski SR, Beca J et al. Comparison of hypothermia and normothermia after severe
traumatic brain injury in children (Cool Kids): a phase 3, randomised controlled trial. Lancet Neurol 2013;
12: 546-553.
Adembri C, Venturi L, Pellegrini-Giampietro DE. Neuroprotective effects of propofol in acute cerebral
injury. CNS Drug Rev 2007; 13: 333-351.
Agbeko RS, Pearson S, Peters MJ, McNames J, Goldstein B. Intracranial pressure and cerebral perfusion
pressure responses to head elevation changes in pediatric traumatic brain injury. Pediatr Crit Care Med
2012; 13: e39-e47.
Albanese J, Durbec O, Viviand X, Potie F, Alliez B, Martin C. Sufentanil increases intracranial pressure in
patients with head trauma. Anesthesiology 1993; 79: 493-497.
Albanese J, Viviand X, Potie F, Rey M, Alliez B, Martin C. Sufentanil, fentanyl, and alfentanil in head
trauma patients: a study on cerebral hemodynamics. Crit Care Med 1999; 27: 407-411.
Alderson P, Gadkary C, Signorini DF. Therapeutic hypothermia for head injury. Cochrane Database Syst
Rev 2004; CD001048.
Alderson P, Roberts I. Corticosteroids in acute traumatic brain injury: systematic review of randomised
controlled trials. BMJ 1997; 314: 1855-1859.
Alderson P, Roberts I. Corticosteroids for acute traumatic brain injury. Cochrane Database Syst Rev
2005; CD000196.
Baker AJ, Rhind SG, Morrison LJ et al. Resuscitation with hypertonic saline-dextran reduces serum
biomarker levels and correlates with outcome in severe traumatic brain injury patients. J Neurotrauma
2009; 26: 1227-1240.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Bao YH, Liang YM, Gao GY, Pan YH, Luo QZ, Jiang JY. Bilateral decompressive craniectomy for patients
with malignant diffuse brain swelling after severe traumatic brain injury: a 37-case study. J Neurotrauma
2010; 27: 341-347.
Battison C, Andrews PJ, Graham C, Petty T. Randomized, controlled trial on the effect of a 20% mannitol
solution and a 7.5% saline/6% dextran solution on increased intracranial pressure after brain injury. Crit
Care Med 2005; 33: 196-202.
Bayir H, Adelson PD, Wisniewski SR et al. Therapeutic hypothermia preserves antioxidant defenses after
severe traumatic brain injury in infants and children. Crit Care Med 2009; 37: 689-695.
Biswas AK, Bruce DA, Sklar FH, Bokovoy JL, Sommerauer JF. Treatment of acute traumatic brain injury in
children with moderate hypothermia improves intracranial hypertension. Crit Care Med 2002; 30: 27422751.
Bracke M, Taylor AG, Kinney AB. External drainage of cerebrospinal fluid. Am J Nurs 1978; 78: 13551358.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
III. Prophylactic hypothermia. J Neurotrauma 2007a; 24 Suppl 1: S21-S25.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
XI. Anesthetics, analgesics, and sedatives. J Neurotrauma 2007b; 24 Suppl 1: S71-S76.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
XV. Steroids. J Neurotrauma 2007c; 24 Suppl 1: S91-S95.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
II. Hyperosmolar therapy. J Neurotrauma 2007d; 24 Suppl 1: S14-S20.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
VI. Indications for intracranial pressure monitoring. J Neurotrauma 2007e; 24 Suppl 1: S37-S44.
Bratton SL, Chestnut RM, Ghajar J et al. Guidelines for the management of severe traumatic brain injury.
XIV. Hyperventilation. J Neurotrauma 2007f; 24 Suppl 1: S87-S90.
Bunegin L, Albin MS, Ernst PS, Garcia C. Cerebrovascular responses to sufentanil citrate (SC) in primates
with and without intracranial hypertension. J Neurosurg Anesthesiol 1989; 1: 138-139.
Carhuapoma JR, Gupta K, Coplin WM, Muddassir SM, Meratee MM. Treatment of refractory fever in the
neurosciences critical care unit using a novel, water-circulating cooling device. A single-center pilot
experience. J Neurosurg Anesthesiol 2003; 15: 313-318.
Carney NA, Ghajar J. Guidelines for the management of severe traumatic brain injury. Introduction. J
Neurotrauma 2007; 24 Suppl 1: S1-S2.
Chen L, Piao Y, Zeng F, Lu M, Kuang Y, Ki X. Moderate hypothermia therapy for patients with severe head
injury. Chin J Traumatol 2001; 4: 164-167.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Clifton GL. Is keeping cool still hot? An update on hypothermia in brain injury. Curr Opin Crit Care 2004;
10: 116-119.
Clifton GL, Choi SC, Miller ER et al. Intercenter variance in clinical trials of head trauma--experience of
the National Acute Brain Injury Study: Hypothermia. J Neurosurg 2001; 95: 751-755.
Clifton GL, Valadka A, Zygun D et al. Very early hypothermia induction in patients with severe brain
injury (the National Acute Brain Injury Study: Hypothermia II): a randomised trial. Lancet Neurol 2011;
10: 131-139.
Coles JP, Minhas PS, Fryer TD et al. Effect of hyperventilation on cerebral blood flow in traumatic head
injury: clinical relevance and monitoring correlates. Crit Care Med 2002; 30: 1950-1959.
Compagnone C, Murray GD, Teasdale GM et al. The management of patients with intradural posttraumatic mass lesions: a multicenter survey of current approaches to surgical management in 729
patients coordinated by the European Brain Injury Consortium. Neurosurgery 2005; 57: 1183-1192.
Cooper DJ, Myles PS, McDermott FT et al. Prehospital hypertonic saline resuscitation of patients with
hypotension and severe traumatic brain injury: a randomized controlled trial. JAMA 2004; 291: 13501357.
Cooper DJ, Rosenfeld JV, Murray L et al. Decompressive craniectomy in diffuse traumatic brain injury. N
Engl J Med 2011; 364: 1493-1502.
Cooper JB, Jane JA, Alves WM, Cooper EB. Right median nerve electrical stimulation to hasten
awakening from coma. Brain Inj 1999; 13: 261-267.
Cooper PR, Moody S, Clark WK et al. Dexamethasone and severe head injury. A prospective double-blind
study. J Neurosurg 1979; 51: 307-316.
Corbett SM, Moore J, Rebuck JA, Rogers FB, Greene CM. Survival of propofol infusion syndrome in a
head-injured patient. Crit Care Med 2006; 34: 2479-2483.
Cremer OL, Moons KG, Bouman EA, Kruijswijk JE, de Smet AM, Kalkman CJ. Long-term propofol infusion
and cardiac failure in adult head-injured patients. Lancet 2001; 357: 117-118.
Cruz J, Minoja G, Okuchi K. Improving clinical outcomes from acute subdural hematomas with the
emergency preoperative administration of high doses of mannitol: a randomized trial. Neurosurgery
2001; 49: 864-871.
Cruz J, Minoja G, Okuchi K. Major clinical and physiological benefits of early high doses of mannitol for
intraparenchymal temporal lobe hemorrhages with abnormal pupillary widening: a randomized trial.
Neurosurgery 2002; 51: 628-637.
Cruz J, Minoja G, Okuchi K, Facco E. Successful use of the new high-dose mannitol treatment in patients
with Glasgow Coma Scale scores of 3 and bilateral abnormal pupillary widening: a randomized trial. J
Neurosurg 2004; 100: 376-383.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Daboussi A, Minville V, Leclerc-Foucras S et al. Cerebral hemodynamic changes in severe head injury
patients undergoing decompressive craniectomy. J Neurosurg Anesthesiol 2009; 21: 339-345.
Davis AE, Gimenez A. Cognitive-behavioral recovery in comatose patients following auditory sensory
stimulation. J Neurosci Nurs 2003; 35: 202-9, 214.
Davis DP, Kimbro TA, Vilke GM. The use of midazolam for prehospital rapid-sequence intubation may be
associated with a dose related increase in hypotension. Prehospital Emergency Care 2001; 5: 161-168.
de Nadal M, Munar F, Poca MA, Sahuquillo J, Garnacho A, Rossello J. Cerebral hemodynamic effects of
morphine and fentanyl in patients with severe head injury: absence of correlation to cerebral
autoregulation. Anesthesiology 2000; 92: 11-19.
DeSalles AA, Kontos HA, Becker DP et al. Prognostic significance of ventricular CSF lactic acidosis in
severe head injury. J Neurosurg 1986; 65: 615-624.
DeSalles AA, Muizelaar JP, Young HF. Hyperglycemia, cerebrospinal fluid lactic acidosis, and cerebral
blood flow in severely head-injured patients. Neurosurgery 1987; 21: 45-50.
Diringer MN, Yundt K, Videen TO et al. No reduction in cerebral metabolism as a result of early
moderate hyperventilation following severe traumatic brain injury. J Neurosurg 2000; 92: 7-13.
Doyle JA, Davis DP, Hoyt DB. The use of hypertonic saline in the treatment of traumatic brain injury. J
Trauma 2001; 50: 367-383.
Durward QJ, Amacher AL, Del Maestro RF, Sibbald WJ. Cerebral and cardiovascular responses to changes
in head elevation in patients with intracranial hypertension. J Neurosurg 1983; 59: 938-944.
Eisenberg HM, Frankowski RF, Contant CF, Marshall LF, Walker MD. High-dose barbiturate control of
elevated intracranial pressure in patients with severe head injury. J Neurosurg 1988; 69: 15-23.
Engelhard K, Reeker W, Kochs E, Werner C. Effect of remifentanil on intracranial pressure and cerebral
blood flow velocity in patients with head trauma. Acta Anaesthesiol Scand 2004; 48: 396-399.
Eshhar N, Striem S, Kohen R, Tirosh O, Biegon A. Neuroprotective and antioxidant activities of HU-211, a
novel NMDA receptor antagonist. Eur J Pharmacol 1995; 283: 19-29.
Eskandari R, Filtz MR, Davis GE, Hoesch RE. Effective treatment of refractory intracranial hypertension
after traumatic brain injury with repeated boluses of 14.6% hypertonic saline. J Neurosurg 2013.
Farling PA, Johnston JR, Coppel DL. Propofol infusion for sedation of patients with head injury in
intensive care. A preliminary report. Anaesthesia 1989; 44: 222-226.
Fay T. Observation on generalized refrigeration in cases of severe cerebral trauma. Res Publ Assoc Res
Nerv Ment Dis 1945; 4: 611-619.
Feigenbaum JJ, Bergmann F, Richmond SA et al. Nonpsychotropic cannabinoid acts as a functional Nmethyl-D-aspartate receptor blocker. Proc Natl Acad Sci U S A 1989; 86: 9584-9587.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Feldman Z, Kanter MJ, Robertson CS et al. Effect of head elevation on intracranial pressure, cerebral
perfusion pressure, and cerebral blood flow in head-injured patients. J Neurosurg 1992; 76: 207-211.
Flint AC, Manley GT, Gean AD, Hemphill JC, III, Rosenthal G. Post-operative expansion of hemorrhagic
contusions after unilateral decompressive hemicraniectomy in severe traumatic brain injury. J
Neurotrauma 2008; 25: 503-512.
Fortune JB, Feustel PJ, Graca L, Hasselbarth J, Kuehler DH. Effect of hyperventilation, mannitol, and
ventriculostomy drainage on cerebral blood flow after head injury. J Trauma 1995; 39: 1091-1097.
Francony G, Fauvage B, Falcon D et al. Equimolar doses of mannitol and hypertonic saline in the
treatment of increased intracranial pressure. Crit Care Med 2008; 36: 795-800.
Fried RC, Dickerson RN, Guenter PA et al. Barbiturate therapy reduces nitrogen excretion in acute head
injury. J Trauma 1989; 29: 1558-1564.
Gal R, Cundrle I, Zimova I, Smrcka M. Mild hypothermia therapy for patients with severe brain injury.
Clin Neurol Neurosurg 2002; 104: 318-321.
Globus MY, Alonso O, Dietrich WD, Busto R, Ginsberg MD. Glutamate release and free radical
production following brain injury: effects of posttraumatic hypothermia. J Neurochem 1995; 65: 17041711.
Grumme T, Baethmann A, Kolodziejczyk D et al. Treatment of patients with severe head injury by
triamcinolone: a prospective, controlled multicenter clinical trial of 396 cases. Res Exp Med (Berl) 1995;
195: 217-229.
Gruner ML, Terhaag D. Multimodal early onset stimulation (MEOS) in rehabilitation after brain injury.
Brain Inj 2000; 14: 585-594.
Guerra WK, Piek J, Gaab MR. Decompressive craniectomy to treat intracranial hypertension in head
injury patients. Intensive Care Med 1999; 25: 1327-1329.
Harris OA, Muh CR, Surles MC et al. Discrete cerebral hypothermia in the management of traumatic
brain injury: a randomized controlled trial. J Neurosurg 2009; 110: 1256-1264.
Hartl R, Bardt TF, Kiening KL, Sarrafzadeh AS, Schneider GH, Unterberg AW. Mannitol decreases ICP but
does not improve brain-tissue pO2 in severely head-injured patients with intracranial hypertension. Acta
Neurochir Suppl 1997; 70: 40-42.
Ho CL, Wang CM, Lee KK, Ng I, Ang BT. Cerebral oxygenation, vascular reactivity, and neurochemistry
following decompressive craniectomy for severe traumatic brain injury. J Neurosurg 2008; 108: 943-949.
Hoefnagel D, Dammers R, Ter Laak-Poort MP, Avezaat CJ. Risk factors for infections related to external
ventricular drainage. Acta Neurochir (Wien ) 2008; 150: 209-214.
Horn P, Munch E, Vajkoczy P et al. Hypertonic saline solution for control of elevated intracranial
pressure in patients with exhausted response to mannitol and barbiturates. Neurol Res 1999; 21: 758764.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Howard JL, Cipolle MD, Anderson M et al. Outcome after decompressive craniectomy for the treatment
of severe traumatic brain injury. J Trauma 2008; 65: 380-385.
Huang AP, Tu YK, Tsai YH et al. Decompressive craniectomy as the primary surgical intervention for
hemorrhagic contusion. J Neurotrauma 2008; 25: 1347-1354.
Hughes S, Colantonio A, Santaguida PL, Paton T. Amantadine to enhance readiness for rehabilitation
following severe traumatic brain injury. Brain Inj 2005; 19: 1197-1206.
Hull RD, Raskob GE, Pineo GF et al. Subcutaneous low-molecular-weight heparin compared with
continuous intravenous heparin in the treatment of proximal-vein thrombosis. N Engl J Med 1992; 326:
975-982.
Hutchinson PJ, Kolias AG, Czosnyka M, Kirkpatrick PJ, Pickard JD, Menon DK. Intracranial pressure
monitoring in severe traumatic brain injury. BMJ 2013; 346: f1000.
Hutchison JS, Ward RE, Lacroix J et al. Hypothermia therapy after traumatic brain injury in children. N
Engl J Med 2008; 358: 2447-2456.
Ichai C, Armando G, Orban JC et al. Sodium lactate versus mannitol in the treatment of intracranial
hypertensive episodes in severe traumatic brain-injured patients. Intensive Care Med 2009; 35: 471-479.
James HE. Analysis of therapeutic modalities for head-injured children. Childs Brain 1979; 5: 263-271.
James ML, Olson DM, Graffagnino C. A pilot study of cerebral and haemodynamic physiological changes
during sedation with dexmedetomidine or propofol in patients with acute brain injury. Anaesth
Intensive Care 2012; 40: 949-957.
Jiang J, Yu M, Zhu C. Effect of long-term mild hypothermia therapy in patients with severe traumatic
brain injury: 1-year follow-up review of 87 cases. J Neurosurg 2000; 93: 546-549.
Jiang JY, Xu W, Li WP et al. Efficacy of standard trauma craniectomy for refractory intracranial
hypertension with severe traumatic brain injury: a multicenter, prospective, randomized controlled
study. J Neurotrauma 2005; 22: 623-628.
Johannigman JA, Davis K, Jr., Miller SL et al. Prone positioning for acute respiratory distress syndrome in
the surgical intensive care unit: who, when, and how long? Surgery 2000; 128: 708-716.
Johnson DA, Roethig-Johnston K, Richards D. Biochemical and physiological parameters of recovery in
acute severe head injury: responses to multisensory stimulation. Brain Inj 1993; 7: 491-499.
Karabinis A, Mandragos K, Stergiopoulos S et al. Safety and efficacy of analgesia-based sedation with
remifentanil versus standard hypnotic-based regimens in intensive care unit patients with brain injuries:
a randomised, controlled trial (ISRCTN50308308). Crit Care 2004; 8: R268-R280.
Kater KM. Response of head-injured patients to sensory stimulation. West J Nurs Res 1989; 11: 20-33.
Kelly DF, Goodale DB, Williams J et al. Propofol in the treatment of moderate and severe head injury: a
randomized, prospective double-blinded pilot trial. J Neurosurg 1999; 90: 1042-1052.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Kerr EM, Marion D, Sereika MS et al. The effect of cerebrospinal fluid drainage on cerebral perfusion in
traumatic brain injured adults. J Neurosurg Anesthesiol 2000; 12: 324-333.
Kerr ME, Weber BB, Sereika SM, Wilberger J, Marion DW. Dose response to cerebrospinal fluid drainage
on cerebral perfusion in traumatic brain-injured adults. Neurosurg Focus 2001; 11: E1.
Kerwin AJ, Schinco MA, Tepas JJ, III, Renfro WH, Vitarbo EA, Muehlberger M. The use of 23.4%
hypertonic saline for the management of elevated intracranial pressure in patients with severe
traumatic brain injury: a pilot study. J Trauma 2009; 67: 277-282.
Khanna S, Davis D, Peterson B et al. Use of hypertonic saline in the treatment of severe refractory
posttraumatic intracranial hypertension in pediatric traumatic brain injury. Crit Care Med 2000; 28:
1144-1151.
Knoller N, Levi L, Shoshan I et al. Dexanabinol (HU-211) in the treatment of severe closed head injury: a
randomized, placebo-controlled, phase II clinical trial. Crit Care Med 2002; 30: 548-554.
Krimchansky BZ, Keren O, Sazbon L, Groswasser Z. Differential time and related appearance of signs,
indicating improvement in the state of consciousness in vegetative state traumatic brain injury (VS-TBI)
patients after initiation of dopamine treatment. Brain Inj 2004; 18: 1099-1105.
Lal D, Weiland S, Newton M, Flaten A, Schurr M. Prehospital hyperventilation after brain injury: a
prospective analysis of prehospital and early hospital hyperventilation of the brain-injured patient.
Prehosp Disaster Med 2003; 18: 20-23.
Lauer KK, Connolly LA, Schmeling WT. Opioid sedation does not alter intracranial pressure in head
injured patients. Can J Anaesth 1997; 44: 929-933.
Lavinio A, Timofeev I, Nortje J et al. Cerebrovascular reactivity during hypothermia and rewarming. Br J
Anaesth 2007; 99: 237-244.
Ledwith MB, Bloom S, Maloney-Wilensky E, Coyle B, Polomano RC, Le Roux PD. Effect of body position
on cerebral oxygenation and physiologic parameters in patients with acute neurological conditions. J
Neurosci Nurs 2010; 42: 280-287.
Lee HC, Chuang HC, Cho DY, Cheng KF, Lin PH, Chen CC. Applying cerebral hypothermia and brain oxygen
monitoring in treating severe traumatic brain injury. World Neurosurg 2010; 74: 654-660.
Lee ST. Intracranial pressure changes during positioning of patients with severe head injury. Heart Lung
1989; 18: 411-414.
Lescot T, Degos V, Zouaoui A, Preteux F, Coriat P, Puybasset L. Opposed effects of hypertonic saline on
contusions and noncontused brain tissue in patients with severe traumatic brain injury. Crit Care Med
2006; 34: 3029-3033.
Levin HS, Saydjari C, Eisenberg HM et al. Vegetative state after closed-head injury. A Traumatic Coma
Data Bank Report. Arch Neurol 1991; 48: 580-585.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Li G, Wen L, Yang XF, Zheng XJ, Zhan RY, Liu WG. Efficacy of large decompressive craniectomy in severe
traumatic brain injury. Chin J Traumatol 2008; 11: 253-256.
Li H, Lu G, Shi W, Zheng S. Protective effect of moderate hypothermia on severe traumatic brain injury in
children. J Neurotrauma 2009; 26: 1905-1909.
Liu JT, Wang CH, Chou IC, Sun SS, Koa CH, Cooper E. Regaining consciousness for prolonged comatose
patients with right median nerve stimulation. Acta Neurochir Suppl 2003; 87: 11-14.
Liu WG, Qiu WS, Zhang Y, Wang WM, Lu F, Yang XF. Effects of selective brain cooling in patients with
severe traumatic brain injury: a preliminary study. J Int Med Res 2006; 34: 58-64.
Llompart-Pou JA, Abadal JM, Perez-Barcena J et al. Long-term follow-up of patients with post-traumatic
refractory high intracranial pressure treated with lumbar drainage. Anaesth Intensive Care 2011; 39: 7983.
Llompart-Pou JA, Perez-Barcena J, Raurich JM et al. Effect of barbiturate coma on adrenal response in
patients with traumatic brain injury. J Endocrinol Invest 2007; 30: 393-398.
Maas AI, Dearden M, Teasdale GM et al. EBIC-guidelines for management of severe head injury in
adults. European Brain Injury Consortium. Acta Neurochir (Wien ) 1997; 139: 286-294.
Maas AI, Murray G, Henney H, III et al. Efficacy and safety of dexanabinol in severe traumatic brain
injury: results of a phase III randomised, placebo-controlled, clinical trial. Lancet Neurol 2006; 5: 38-45.
March K. Intracranial pressure monitoring: why monitor? AACN Clin Issues 2005; 16: 456-475.
Marion DW. Therapeutic moderate hypothermia for severe traumatic brain injury. J Intensive Care Med
1997; 12: 239-248.
Marion DW, Obrist WD, Carlier PM, Penrod LE, Darby JM. The use of moderate therapeutic hypothermia
for patients with severe head injuries: a preliminary report. J Neurosurg 1993; 79: 354-362.
Marion DW, Penrod LE, Kelsey SF et al. Treatment of traumatic brain injury with moderate hypothermia.
N Engl J Med 1997; 336: 540-546.
Marshall LF, Smith RW, Shapiro HM. The outcome with aggressive treatment in severe head injuries.
Part II: acute and chronic barbiturate administration in the management of head injury. J Neurosurg
1979; 50: 26-30.
Marx W, Shah N, Long C et al. Sufentanil, alfentanil, and fentanyl: impact on cerebrospinal fluid pressure
in patients with brain tumors. J Neurosurg Anesthesiol 1989; 1: 3-7.
McClelland M, Woster P, Sweasey T, Hoff JT. Continuous midazolam/atracurium infusions for the
management of increased intracranial pressure. J Neurosci Nurs 1995; 27: 96-101.
McCollam JS, O'Neil MG, Norcross ED, Byrne TK, Reeves ST. Continuous infusions of lorazepam,
midazolam, and propofol for sedation of the critically ill surgery trauma patient: a prospective,
randomized comparison. Crit Care Med 1999; 27: 2454-2458.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
McMahon MA, Vargus-Adams JN, Michaud LJ, Bean J. Effects of amantadine in children with impaired
consciousness caused by acquired brain injury: a pilot study. Am J Phys Med Rehabil 2009; 88: 525-532.
Mechoulam R, Feigenbaum JJ, Lander N et al. Enantiomeric cannabinoids: stereospecificity of
psychotropic activity. Experientia 1988; 44: 762-764.
Meier U, Ahmadi S, Killeen T, Al-Zain FT, Lemcke J. Long-term outcomes following decompressive
craniectomy for severe head injury. Acta Neurochir Suppl 2008; 102: 29-31.
Meixensberger J, Baunach S, Amschler J, Dings J, Roosen K. Influence of body position on tissue-pO2,
cerebral perfusion pressure and intracranial pressure in patients with acute brain injury. Neurol Res
1997; 19: 249-253.
Metz C, Gobel L, Gruber M, Hoerauf KH, Taeger K. Pharmacokinetics of human cerebral opioid
extraction: a comparative study on sufentanil, fentanyl, and alfentanil in a patient after severe head
injury. Anesthesiology 2000; 92: 1559-1567.
Meythaler JM, Brunner RC, Johnson A, Novack TA. Amantadine to improve neurorecovery in traumatic
brain injury-associated diffuse axonal injury: a pilot double-blind randomized trial. J Head Trauma
Rehabil 2002; 17: 300-313.
Mitchell S, Bradley VA, Welch JL, Britton PG. Coma arousal procedure: a therapeutic intervention in the
treatment of head injury. Brain Inj 1990; 4: 273-279.
Morgalla MH, Will BE, Roser F, Tatagiba M. Do long-term results justify decompressive craniectomy after
severe traumatic brain injury? J Neurosurg 2008; 109: 685-690.
Moseley AM, Herbert RD, Sherrington C, Maher CG. Evidence for physiotherapy practice: a survey of the
Physiotherapy Evidence Database (PEDro). Aust J Physiother 2002; 48: 43-49.
Muizelaar JP, Marmarou A, Ward JD et al. Adverse effects of prolonged hyperventilation in patients with
severe head injury: a randomized clinical trial. J Neurosurg 1991; 75: 731-739.
Munch E, Horn P, Schurer L, Piepgras A, Paul T, Schmiedek P. Management of severe traumatic brain
injury by decompressive craniectomy. Neurosurgery 2000; 47: 315-322.
Murad A, Ghostine S, Colohan AR. Controlled lumbar drainage in medically refractory increased
intracranial pressure. A safe and effective treatment. Acta Neurochir Suppl 2008; 102: 89-91.
Murad A, Ghostine S, Colohan AR. A case for further investigating the use of controlled lumbar
cerebrospinal fluid drainage for the control of intracranial pressure. World Neurosurg 2012; 77: 160-165.
Myburgh J, Cooper DJ, Finfer S et al. Saline or albumin for fluid resuscitation in patients with traumatic
brain injury. N Engl J Med 2007; 357: 874-884.
Nekludov M, Bellander BM, Mure M. Oxygenation and cerebral perfusion pressure improved in the
prone position. Acta Anaesthesiol Scand 2006; 50: 932-936.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Ng I, Lim J, Wong HB. Effects of head posture on cerebral hemodynamics: its influences on intracranial
pressure, cerebral perfusion pressure, and cerebral oxygenation. Neurosurgery 2004; 54: 593-597.
Noda R, Maeda Y, Yoshino A. Therapeutic time window for musicokinetic therapy in a persistent
vegetative state after severe brain damage. Brain Inj 2004; 18: 509-515.
Nordby HK, Nesbakken R. The effect of high dose barbiturate decompression after severe head injury. A
controlled clinical trial. Acta Neurochir (Wien ) 1984; 72: 157-166.
Oddo M, Levine JM, Frangos S et al. Effect of mannitol and hypertonic saline on cerebral oxygenation in
patients with severe traumatic brain injury and refractory intracranial hypertension. J Neurol Neurosurg
Psychiatry 2009; 80: 916-920.
Olivecrona M, Rodling-Wahlstrom M, Naredi S, Koskinen LO. Effective ICP reduction by decompressive
craniectomy in patients with severe traumatic brain injury treated by an ICP-targeted therapy. J
Neurotrauma 2007; 24: 927-935.
Otani N, Takasato Y, Masaoka H et al. Surgical outcome following a decompressive craniectomy for
acute epidural hematoma patients presenting with associated massive brain swelling. Acta Neurochir
Suppl 2010; 106: 261-264.
Otterspoor LC, Kalkman CJ, Cremer OL. Update on the propofol infusion syndrome in ICU management
of patients with head injury. Curr Opin Anaesthesiol 2008; 21: 544-551.
Papazian L, Albanese J, Thirion X, Perrin G, Durbec O, Martin C. Effect of bolus doses of midazolam on
intracranial pressure and cerebral perfusion pressure in patients with severe head injury. Br J Anaesth
1993; 71: 267-271.
Pascual JL, Maloney-Wilensky E, Reilly PM et al. Resuscitation of hypotensive head-injured patients: is
hypertonic saline the answer? Am Surg 2008; 74: 253-259.
Patrick PD, Blackman JA, Mabry JL, Buck ML, Gurka MJ, Conaway MR. Dopamine agonist therapy in lowresponse children following traumatic brain injury. Journal of Child Neurology 2006; 21: 879-885.
Pelosi P, Brazzi L, Gattinoni L. Prone position in acute respiratory distress syndrome. Eur Respir J 2002;
20: 1017-1028.
Perez-Barcena J, Barcelo B, Homar J et al. (Comparison of the effectiveness of pentobarbital and
thiopental in patients with refractory intracranial hypertension. Preliminary report of 20 patients).
Neurocirugia (Astur ) 2005; 16: 5-12.
Perez-Barcena J, Llompart-Pou JA, Homar J et al. Pentobarbital versus thiopental in the treatment of
refractory intracranial hypertension in patients with traumatic brain injury: a randomized controlled
trial. Crit Care 2008; 12: R112.
Peri CV, Shaffrey ME, Farace E et al. Pilot study of electrical stimulation on median nerve in comatose
severe brain injured patients: 3-month outcome. Brain Inj 2001; 15: 903-910.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Pierce JP, Lyle DM, Quine S, Evans NJ, Morris J, Fearnside MR. The effectiveness of coma arousal
intervention. Brain Inj 1990; 4: 191-197.
Polderman KH, Tjong Tjin JR, Peerdeman SM, Vandertop WP, Girbes AR. Effects of therapeutic
hypothermia on intracranial pressure and outcome in patients with severe head injury. Intensive Care
Med 2002; 28: 1563-1573.
Polin RS, Shaffrey ME, Bogaev CA et al. Decompressive bifrontal craniectomy in the treatment of severe
refractory posttraumatic cerebral edema. Neurosurgery 1997; 41: 84-92.
Qiu W, Guo C, Shen H et al. Effects of unilateral decompressive craniectomy on patients with unilateral
acute post-traumatic brain swelling after severe traumatic brain injury. Crit Care 2009; 13: R185.
Qiu W, Shen H, Zhang Y et al. Noninvasive selective brain cooling by head and neck cooling is protective
in severe traumatic brain injury. J Clin Neurosci 2006; 13: 995-1000.
Qiu W, Zhang Y, Sheng H et al. Effects of therapeutic mild hypothermia on patients with severe
traumatic brain injury after craniotomy. J Crit Care 2007; 22: 229-235.
Qureshi AI, Suarez JI, Bhardwaj A et al. Use of hypertonic (3%) saline/acetate infusion in the treatment
of cerebral edema: Effect on intracranial pressure and lateral displacement of the brain. Crit Care Med
1998; 26: 440-446.
Rea GL, Rockswold GL. Barbiturate therapy in uncontrolled intracranial hypertension. Neurosurgery
1983; 12: 401-404.
Resnick DK, Marion DW, Darby JM. The effect of hypothermia on the incidence of delayed traumatic
intracerebral hemorrhage. Neurosurgery 1994; 34: 252-255.
Rhind SG, Crnko NT, Baker AJ et al. Prehospital resuscitation with hypertonic saline-dextran modulates
inflammatory, coagulation and endothelial activation marker profiles in severe traumatic brain injured
patients. J Neuroinflammation 2010; 7: 5.
Roberts I. Barbiturates for acute traumatic brain injury. Cochrane Database Syst Rev 2000; CD000033.
Roberts I, Yates D, Sandercock P et al. Effect of intravenous corticosteroids on death within 14 days in
10008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled
trial. Lancet 2004; 364: 1321-1328.
Rockoff MA, Marshall LF, Shapiro HM. High-dose barbiturate therapy in humans: a clinical review of 60
patients. Ann Neurol 1979; 6: 194-199.
Rockswold GL, Solid CA, Paredes-Andrade E, Rockswold SB, Jancik JT, Quickel RR. Hypertonic saline and
its effect on intracranial pressure, cerebral perfusion pressure, and brain tissue oxygen. Neurosurgery
2009; 65: 1035-1041.
Sabsovich I, Rehman Z, Yunen J, Coritsidis G. Propofol infusion syndrome: a case of increasing morbidity
with traumatic brain injury. Am J Crit Care 2007; 16: 82-85.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Sahuquillo J, Arikan F. Decompressive craniectomy for the treatment of refractory high intracranial
pressure in traumatic brain injury. Cochrane Database Syst Rev 2006; CD003983.
Salvatore C, Marco M, Antonio R, Salvatore I, Eugenio B. Combined internal uncusectomy and
decompressive craniectomy for the treatment of severe closed head injury: experience with 80 cases. J
Neurosurg 2008; 108: 74-79.
Sanchez-Izquierdo-Riera JA, Caballero-Cubedo RE, Perez-Vela JL, mbros-Checa A, Cantalapiedra-Santiago
JA, ted-Lopez E. Propofol versus midazolam: safety and efficacy for sedating the severe trauma patient.
Anesth Analg 1998; 86: 1219-1224.
Saniova B, Drobny M, Kneslova L, Minarik M. The outcome of patients with severe head injuries treated
with amantadine sulphate. J Neural Transm 2004; 111: 511-514.
Sawyer E, Mauro LS, Ohlinger MJ. Amantadine enhancement of arousal and cognition after traumatic
brain injury. Ann Pharmacother 2008; 42: 247-252.
Sayre MR, Daily SW, Stern SA, Storer DL, van Loveren HR, Hurst JM. Out-of-hospital administration of
mannitol to head-injured patients does not change systolic blood pressure. Acad Emerg Med 1996; 3:
840-848.
Schalen W, Sonesson B, Messeter K, Nordstrom G, Nordstrom CH. Clinical outcome and cognitive
impairment in patients with severe head injuries treated with barbiturate coma. Acta Neurochir (Wien )
1992; 117: 153-159.
Schatzmann C, Heissler HE, Konig K et al. Treatment of elevated intracranial pressure by infusions of 10%
saline in severely head injured patients. Acta Neurochir Suppl 1998; 71: 31-33.
Schneider WN, Drew-Cates J, Wong TM, Dombovy ML. Cognitive and behavioural efficacy of amantadine
in acute traumatic brain injury: an initial double-blind placebo-controlled study. Brain Inj 1999; 13: 863872.
Scholz J, Bause H, Schulz M et al. Pharmacokinetics and effects on intracranial pressure of sufentanil in
head trauma patients. Br J Clin Pharmacol 1994; 38: 369-372.
Schubert A. Side effects of mild hypothermia. J Neurosurg Anesthesiol 1995; 7: 139-147.
Schulz-Stiibner S, Thiex R. Raising the head-of-bed by 30 degrees reduces ICP and improves CPP without
compromising cardiac output in euvolemic patients with traumatic brain injury and subarachnoid
haemorrhage: a practice unit. 2006. p. 173-80.
Schwartz ML, Tator CH, Rowed DW, Reid SR, Meguro K, Andrews DF. The University of Toronto head
injury treatment study: a prospective, randomized comparison of pentobarbital and mannitol. Can J
Neurol Sci 1984; 11: 434-440.
Shackford SR, Bourguignon PR, Wald SL, Rogers FB, Osler TM, Clark DE. Hypertonic saline resuscitation of
patients with head injury: a prospective, randomized clinical trial. J Trauma 1998; 44: 50-58.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Shiozaki T, Kato A, Taneda M et al. Little benefit from mild hypothermia therapy for severely head
injured patients with low intracranial pressure. J Neurosurg 1999; 91: 185-191.
Shiozaki T, Sugimoto H, Taneda M et al. Effect of mild hypothermia on uncontrollable intracranial
hypertension after severe head injury. J Neurosurg 1993; 79: 363-368.
Shohami E, Novikov M, Bass R. Long-term effect of HU-211, a novel non-competitive NMDA antagonist,
on motor and memory functions after closed head injury in the rat. Brain Res 1995; 674: 55-62.
Simma B, Burger R, Falk M, Sacher P, Fanconi S. A prospective, randomized, and controlled study of fluid
management in children with severe head injury: lactated Ringer's solution versus hypertonic saline. Crit
Care Med 1998; 26: 1265-1270.
Skoglund TS, Eriksson-Ritzen C, Jensen C, Rydenhag B. Aspects on decompressive craniectomy in
patients with traumatic head injuries. J Neurotrauma 2006; 23: 1502-1509.
Smith H, Sinson G, Varelas P. Vasopressors and propofol infusion syndrome in severe head trauma.
Neurocrit Care 2009; 10: 166-172.
Smith HP, Kelly DL, Jr., McWhorter JM et al. Comparison of mannitol regimens in patients with severe
head injury undergoing intracranial monitoring. J Neurosurg 1986; 65: 820-824.
Sorani MD, Morabito D, Rosenthal G, Giacomini KM, Manley GT. Characterizing the dose-response
relationship between mannitol and intracranial pressure in traumatic brain injury patients using a highfrequency physiological data collection system. J Neurotrauma 2008; 25: 291-298.
Sperry RJ, Bailey PL, Reichman MV, Peterson JC, Petersen PB, Pace NL. Fentanyl and sufentanil increase
intracranial pressure in head trauma patients. Anesthesiology 1992; 77: 416-420.
Stewart L, Bullock R, Rafferty C, Fitch W, Teasdale GM. Propofol sedation in severe head injury fails to
control high ICP, but reduces brain metabolism. Acta Neurochir Suppl (Wien ) 1994; 60: 544-546.
Stover JF, Stocker R. Barbiturate coma may promote reversible bone marrow suppression in patients
with severe isolated traumatic brain injury. Eur J Clin Pharmacol 1998; 54: 529-534.
Suarez JI. Hypertonic saline for cerebral edema and elevated intracranial pressure. Cleve Clin J Med
2004; 71 Suppl 1: S9-13.
Tateishi A, Soejima Y, Taira Y et al. Feasibility of the titration method of mild hypothermia in severely
head-injured patients with intracranial hypertension. Neurosurgery 1998; 42: 1065-1069.
Thelandersson A, Cider A, Nellgard B. Prone position in mechanically ventilated patients with reduced
intracranial compliance. Acta Anaesthesiol Scand 2006; 50: 937-941.
Thiagarajan A, Goverdhan PD, Chari P, Somasunderam K. The effect of hyperventilation and hyperoxia
on cerebral venous oxygen saturation in patients with traumatic brain injury. Anesth Analg 1998; 87:
850-853.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Thorat JD, Wang EC, Lee KK, Seow WT, Ng I. Barbiturate therapy for patients with refractory intracranial
hypertension following severe traumatic brain injury: its effects on tissue oxygenation, brain
temperature and autoregulation. J Clin Neurosci 2008; 15: 143-148.
Tillett JM, Marmarou A, Agnew JP, Choi SC, Ward JD. Effect of continuous rotational therapy on
intracranial pressure in the severely brain-injured patient. Crit Care Med 1993; 21: 1005-1011.
Timofeev I, Czosnyka M, Nortje J et al. Effect of decompressive craniectomy on intracranial pressure and
cerebrospinal compensation following traumatic brain injury. J Neurosurg 2008a; 108: 66-73.
Timofeev I, hyot-Fizelier C, Keong N et al. Ventriculostomy for control of raised ICP in acute traumatic
brain injury. Acta Neurochir Suppl 2008b; 102: 99-104.
Tokutomi T, Miyagi T, Takeuchi Y, Karukaya T, Katsuki H, Shigemori M. Effect of 35 degrees C
hypothermia on intracranial pressure and clinical outcome in patients with severe traumatic brain injury.
J Trauma 2009; 66: 166-173.
Tokutomi T, Morimoto K, Miyagi T, Yamaguchi S, Ishikawa K, Shigemori M. Optimal temperature for the
management of severe traumatic brain injury: effect of hypothermia on intracranial pressure, systemic
and intracranial hemodynamics, and metabolism. Neurosurgery 2003; 52: 102-111.
Tuettenberg J, Czabanka M, Horn P et al. Clinical evaluation of the safety and efficacy of lumbar
cerebrospinal fluid drainage for the treatment of refractory increased intracranial pressure. J Neurosurg
2009; 110: 1200-1208.
Ucar T, Akyuz M, Kazan S, Tuncer R. Role of decompressive surgery in the management of severe head
injuries: prognostic factors and patient selection. J Neurotrauma 2005; 22: 1311-1318.
Vargus-Adams JN, McMahon MA, Michaud LJ, Bean J, Vinks AA. Pharmacokinetics of amantadine in
children with impaired consciousness due to acquired brain injury: preliminary findings using a sparsesampling technique. PM&R 2010; 2: 37-42.
Vialet R, Albanese J, Thomachot L et al. Isovolume hypertonic solutes (sodium chloride or mannitol) in
the treatment of refractory posttraumatic intracranial hypertension: 2 mL/kg 7.5% saline is more
effective than 2 mL/kg 20% mannitol. Crit Care Med 2003; 31: 1683-1687.
Wakai A, Roberts I, Schierhout G. Mannitol for acute traumatic brain injury. Cochrane Database Syst Rev
2005; CD001049.
Ward JD, Becker DP, Miller JD et al. Failure of prophylactic barbiturate coma in the treatment of severe
head injury. J Neurosurg 1985; 62: 383-388.
Ware ML, Nemani VM, Meeker M, Lee C, Morabito DJ, Manley GT. Effects of 23.4% sodium chloride
solution in reducing intracranial pressure in patients with traumatic brain injury: a preliminary study.
Neurosurgery 2005; 57: 727-736.
Warner KJ, Cuschieri J, Copass MK, Jurkovich GJ, Bulger EM. The impact of prehospital ventilation on
outcome after severe traumatic brain injury. J Trauma 2007; 62: 1330-1336.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013
Evidence-Based Review of Moderate to Severe Acquired Brain Injury
Werner C, Kochs E, Bause H, Hoffman WE, Schulte am EJ. Effects of sufentanil on cerebral
hemodynamics and intracranial pressure in patients with brain injury. Anesthesiology 1995; 83: 721-726.
Whyte J, Katz D, Long D et al. Predictors of outcome in prolonged posttraumatic disorders of
consciousness and assessment of medication effects: A multicenter study. Arch Phys Med Rehabil 2005;
86: 453-462.
Williams RF, Magnotti LJ, Croce MA et al. Impact of decompressive craniectomy on functional outcome
after severe traumatic brain injury. J Trauma 2009; 66: 1570-1574.
Wilson SL, Cranny SM, Andrews K. The efficacy of music for stimulation in prolonged coma - four single
case experiments. Clinical Rehabilitation 1992; 6: 181-187.
Wilson SL, Powell GE, Brock D, Thwaites H. Vegetative state and responses to sensory stimulation: an
analysis of 24 cases. Brain Inj 1996; 10: 807-818.
Winkelman C. Effect of backrest position on intracranial and cerebral perfusion pressures in
traumatically brain-injured adults. Am J Crit Care 2000; 9: 373-380.
Wood RL, Winkowski TB, Miller JL, Tierney L, Goldman L. Evaluating sensory regulation as a method to
improve awareness in patients with altered states of consciousness: a pilot study. Brain Inj 1992; 6: 411418.
Yamamoto T, Mori K, Maeda M. Assessment of prognostic factors in severe traumatic brain injury
patients treated by mild therapeutic cerebral hypothermia therapy. Neurol Res 2002; 24: 789-795.
Yan Y, Tang W, Deng Z, Zhong D, Yang G. Cerebral oxygen metabolism and neuroelectrophysiology in a
clinical study of severe brain injury and mild hypothermia. J Clin Neurosci 2010; 17: 196-200.
Yang XF, Wen L, Shen F et al. Surgical complications secondary to decompressive craniectomy in patients
with a head injury: a series of 108 consecutive cases. Acta Neurochir (Wien ) 2008; 150: 1241-1247.
Zabramski JM, Whiting D, Darouiche RO et al. Efficacy of antimicrobial-impregnated external ventricular
drain catheters: a prospective, randomized, controlled trial. J Neurosurg 2003; 98: 725-730.
Zafonte RD, Lexell J, Cullen N. Possible applications for dopaminergic agents following traumatic brain
injury: part 2. J Head Trauma Rehabil 2001; 16: 112-116.
Module 16_Acute Interventions for Acquired Brain Injury_V9_2013
http://www.abiebr.com
Updated August 2013