Protocol

International Journal of Stroke

2016, Vol. 11(3) 368–379
Early tracheostomy in ventilated stroke ! 2016 World Stroke Organization
Reprints and permissions:
patients: Study protocol of the sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1747493015616638

international multicentre randomized wso.sagepub.com

trial SETPOINT2 (Stroke-related Early

Tracheostomy vs. Prolonged Orotracheal
Intubation in Neurocritical care Trial 2)

Silvia Schönenberger1, Wolf-Dirk Niesen2, Hannah Fuhrer2,

Colleen Bauza3, Christina Klose4, Meinhard Kieser4,
José I Suarez5, David B Seder6, Julian Bösel1; on behalf of the
SETPOINT2-study group and the IGNITE-study group

Abstract
Background: Tracheostomy is a common procedure in long-term ventilated critical care patients and frequently
necessary in those with severe stroke. The optimal timing for tracheostomy is still unknown, and it is controversial
whether early tracheostomy impacts upon functional outcome.
Method: The Stroke-related Early Tracheostomy vs. Prolonged Orotracheal Intubation in Neurocritical care Trial 2
(SETPOINT2) is a multicentre, prospective, randomized, open-blinded endpoint (PROBE-design) trial. Patients with
acute ischemic stroke, intracerebral hemorrhage or subarachnoid hemorrhage who are so severely affected that two
weeks of ventilation are presumed necessary based on a prediction score are eligible. It is intended to enroll 190 patients
per group (n ¼ 380). Patients are randomized to either percutaneous tracheostomy within the first five days after
intubation or to ongoing orotracheal intubation with consecutive weaning and extubation and, if the latter failed, to
percutaneous tracheostomy from day 10 after intubation. The primary endpoint is functional outcome defined by the
modified Rankin Scale (mRS, 0–4 (favorable) vs. 5 þ 6 (unfavorable)) after six months; secondary endpoints are mortality
and cause of mortality during intensive care unit-stay and within six months from admission, intensive care unit-length of
stay, duration of sedation, duration of ventilation and weaning, timing and reasons for withdrawal of life support meas-
ures, relevant intracranial pressure rises before and after tracheostomy.
Conclusion: The necessity and optimal timing of tracheostomy in ventilated stroke patients need to be identified.
SETPOINT2 should clarify whether benefits in functional outcome can be achieved by early tracheostomy in these
patients.

Keywords
Ischemic stroke, clinical trial, intracerebral hemorrhage, tracheostomy, neurocritical care, weaning

Received: 6 July 2015; accepted: 3 September 2015

1
Department of Neurology, University of Heidelberg, Heidelberg,
Germany
2
Department of Neurology, University of Freiburg, Freiburg im Breisgau,
Germany 6
3
Department of Public Health Sciences, Medical University of South Department of Critical Care Services, Maine Medical Center, Portland,
Carolina, Charleston, South Carolina, USA Maine, USA
4
Institute of Medical Biometry and Informatics, University of Heidelberg, Corresponding author:
Heidelberg, Germany Julian Bösel, Department of Neurology, University of Heidelberg, Im
5
Division of Vascular Neurology and Neurocritical Care, Department of Neuenheimer Feld 400, D-69120 Heidelberg, Germany.
Neurology, Baylor College of Medicine, Houston, Texas, USA Email: julian.boesel@med.uni-heidelberg.de

International Journal of Stroke, 11(3)

Schönenberger et al.
Executive summary
Rationale
Ventilated intensive care patients with ischemic or hem-
orrhagic stroke have a poor prognosis. Early tracheos-
tomy (TT) has advantages in selected groups of
non-cerebrovascular intensive care unit (ICU) patients,
including shorter ventilation time, shorter ICU length
of stay (LOS) and reduced complications. It is unclear
whether ventilated stroke patients, a subgroup whose
pathophysiology may favor TT, benefit from early TT.
Aim and hypothesis
SETPOINT2 is a randomized, assessor-blinded clinical
trial of early TT vs. prolonged intubation and ventila-
tor weaning for respiratory support of patients with
severe ischemic and hemorrhagic stroke. The primary
aim of the trial is to test whether early TT favorably
impacts upon long-term functional outcome; secondary
aims are to test whether early intention-to-treat (ITT)
exerts any beneficial effects on the ICU course of these
patients, including cost of treatment and intracranial
pressure (ICP) management.
Design
SETPOINT2 is a multicentre, prospective, randomized,
open-blinded endpoint study (PROBE-design). Patients
are randomized to either PDT within the first five days
from intubation or to ongoing weaning/extubation
attempts or PDT from day 10 after intubation.
Study outcomes
The primary endpoint is the dichotomized functional
outcome (mRS score of 0–4 (favorable outcome) vs.
5,6 (unfavorable outcome)) at six months after admis-
sion to the ICU. Secondary endpoints include mortality
and cause of mortality during ICU-stay and within six
months from admission, ICU-LOS, duration of sed-
ation, duration of ventilation and weaning, timing
and reasons for withdrawal of life support measures,
relevant ICP rises before and after TT.
Discussion
The potential benefits of early TT have not been pro-
spectively addressed in stroke patients, other than in
the SETPOINT pilot trial. That trial demonstrated
safety and suggested substantial benefits of early TT
in this burdened population. Positive results of
SETPOINT2 could lead to a reduction in potentially
harmful ICU-treatments such as prolonged sedation
and ventilation, greater chance of survival, earlier
369
transfer to rehabilitation and higher rehabilitation
potential, better long-term outcome, and better estima-
tion of or even reduction of the costs of care.
Introduction and rationale
According to United States data from the National
Inpatient Sample, about 1.3% of 1.5 million patients
(n ¼ 20,300) hospitalized with ischemic stroke from
2007 to 2009 underwent TT – while the number of
tracheostomies performed for hemorrhagic stroke is
unknown,1 but likely to be even higher. Historically,
mechanically ventilated patients with ischemic or hem-
orrhagic stroke have had poor functional outcomes
with reported mortality rates ranging between 40%
and 80%,2–5 and care of such patients is extremely
expensive.6,7 Effective interventions to improve sur-
vival, improve functional recovery, decrease costs,
and increase cost-effectiveness are therefore of great
importance. Unlike the medical and surgical critically
ill, many patients with stroke who require prolonged
intubation do not necessarily require mechanical venti-
lation, but have a need for an airway protection. Most
often instead, these patients cannot reliably clear secre-
tions from the airway or maintain an open upper
airway due to decreased bulbar function, decreased
airway protective reflexes, muscle weakness, and a
weak cough.8,9 Accordingly, TT offers the opportunity
to disconnect such patients from mechanical ventilation
while maintaining an open airway protected from aspir-
ation, and provides the ability to directly suction secre-
tions from the lower airways. Most importantly, TT is
more comfortable, allowing for earlier discontinuation
of analgesia and sedation,10 which may be crucial in
allowing brain-injured patients to awaken, be weaned
from ventilation, and begin early active rehabilitation
regimens. Current TT practices vary widely between
hospitals and patients. PDT, a bedside procedure, is
increasingly popular compared to surgical TT, which
is more resource-intensive, with a worse safety profile
but similar long-term outcomes.11–17 Traditional indi-
cations for TT include an anticipated duration of mech-
anical ventilation greater than 14–21 days, to prevent
complications of prolonged intubation such as vocal
cord injury, tracheomalacia, and ventilator-associated
pneumonia (VAP), and to facilitate discharge from the
ICU. Extubation delay is associated with higher mor-
bidity, mortality, ICU-LOS and hospital stay, and
cost.16,18–20 Because of inconclusive trials in the general
medical surgical population, a recent review of trials on
early TT concluded that clinicians may ‘‘defer TT
placement for at least two weeks following the onset
of acute respiratory failure to insure need for ongoing
ventilator support’’.21 TT is eventually necessary in all
ICU patients who cannot be weaned from mechanical
International Journal of Stroke, 11(3)
370
ventilation. Five randomized controlled trials (RCTs)
have investigated early TT (within seven days after
admission) in populations of medical, surgical,
trauma, and burn patients (406 in total) and report
advantages of earlier TT such as less VAP, less need
for analgesia and sedation,10 shorter duration of mech-
anical ventilation and ICU-stay and – in one trial –
lower mortality.22 A meta-analysis of these trials
reported significantly reduced duration of mechanical
ventilation and shorter intensive care.23 Large multi-
centre randomized trials of early vs. delayed TT have
been conducted in mixed ICU populations with disap-
pointing results,24,25 while an even larger retrospective
analysis likewise showed only a marginal benefit of
early TT for survival, time of ventilation, and ICU-
LOS.26 The largest project in this regard, the UK
Tracman Trial in more than 1000 mixed ICU patients
demonstrated no benefit of early TT (within four days
of critical care admission) in 30-day mortality or other
important secondary outcomes.24 A very recent sys-
temic review and meta-analysis by Siempos et al.27 sug-
gested that early TT is not associated with lower
mortality in the ICU compared to late or no TT, but
may be associated with a lower incidence of VAP. On
these grounds, considering the earlier RCTs in selected
ICU populations, it could be more relevant (and
rewarding) to investigate the matter in special ICU
subgroups.
The airway-pathophysiology of traumatic brain
injury is probably similar to stroke, and several studies
favor early over delayed TT.28–30 Until recently, the
question of early TT in non-traumatic brain diseases
including stroke had not been addressed. Potential
benefits of early TT in ventilated stroke patients were
recently reviewed and included advantages regarding
ICU-LOS, duration of ventilation, and mortality.31 A
recent analysis of the timing of TT in the National
Inpatient Sample suggested that TT performed before
day 10 was associated with decreased incidence of VAP
and decreased LOS, compared to later TT.32
PDT can be performed by neurointensivists at the
bedside quickly, at low cost, and with very low compli-
cation rates.16 Transient increases in ICP can occur, but
do not differ significantly between early and late TT.33
Details of these and other studies on TT in neuro-
critical care patients can be found in Table 1.
The only prospective randomized trial on potential
benefits of early TT in ventilated stroke patients to date
is the monocentric pilot trial SETPOINT (stroke-
related early TT vs. prolonged orotracheal intubation
in neurocritical care trial).31 Patients were randomized
to early TT within three days from intubation or pro-
longed intubation and TT performed between day
7 and 14 if they had to remain intubated. In
SETPOINT’s, early TT group, there was a significant
International Journal of Stroke, 11(3)
International Journal of Stroke 11(3)
reduction in sedation use (sedated during 42% vs. 62%
of ICU days, P ¼ 0.02), and a dramatic reduction in
ICU-mortality (16% vs. 45%, P < 0.01), although it is
noteworthy that this pilot trial was not designed to
focus on the latter.40 Mortality and functional outcome
(measured by the mRS) at six months also trended
toward better in the early TT group. The trial also
showed safety of the procedure and provided data for
a sample size calculation for a larger confirmatory trial.
Based on this pilot trial and its suggested survival bene-
fit, we designed the follow-up trial SETPOINT2 directed
at functional outcome. This is the first randomized mul-
ticentre trial to investigate potential benefits with regard
to functional outcome, mortality and other parameters
of the clinical course of early TT compared with pro-
longed intubation in ventilated stroke patients.
Methods
Design
SETPOINT2 is a multicentre, prospective, rando-
mized, open-blinded endpoint study (PROBE-
design). Patients are randomized 1:1 to either PDT
within five days after intubation (‘‘early TT’’) or to
PDT from day 10 after intubation (‘‘standard’’) if
the application of a standard weaning protocol does
not first lead to successful extubation. The timing of
TT is the only intervention – otherwise, all care is
standard. Guidelines-based recommendations to cen-
ters are applied to ensure similar management deci-
sions for both groups in fields such as weaning,
ventilation, analgesia and sedation, transfusion, cere-
bral monitoring and management, etc.
Blinding is impossible for treating physicians,
patients and legal representatives as well as for most
of the investigators. However, the endpoints of long-
term outcome and cause of mortality will be assessed
by care team-independent adjudicators blinded to the
timing of TT.
Patient population – Inclusion and
exclusion criteria
Inclusion criteria
Subjects meeting all of the following criteria will be
considered for inclusion in the trial:
1. Age 18 years or older, either sex.
2. One of the following confirmed admission
diagnoses:
a. non-traumatic acute ischemic stroke (AIS),
b. non-traumatic intracerebral hemorrhage (ICH),
c. non-traumatic subarachnoid hemorrhage (SAH).
Schönenberger et al. 371

Table 1. Relevant studies on tracheostomy in neurological ICU patients in chronological order

Study N Study pathologies Study design Main findings

34
Qureshi et al. 69 Infratentorial lesions Retrospective Reduced ICU-LOS, and ventilation
duration by early TT

Rabinstein and 97 AIS, ISH, SAH survivors after Retrospective Reduced ICU-LOS and ventilation
Wijdicks35 prolonged intubation and/or duration by early TT
TT

Van der Lely et al.36 130 Mixed, including 31 neuro- Retrospective Neurological/neurosurgical group
logical/neurosurgical fastest to wean after TT among
subgroups

Huttner et al.37 392 Supratentorial ICH Retrospective Predictors of TT need: COPD,

hematoma volume, ganglionic
location, hydrocephalus

Pinheiro et al.38 28 Mixed neurological, including Retrospective Reduced 30-day mortality by early TT
‘‘cerebrovascular accidents’’

Seder et al.16 135 Mixed neurocritical Retrospective/ Reduction of ICU-LOS, ventilation

prospective duration, ICU-charges by percu-
taneous vs. surgical TT

Szeder et al.39 150 Supratentorial ICH Retrospective Predictors of TT need: GCS, septum
pellucidum shift, thalamic location,
hydrocephalus (TRACH score)

Bösel et al.40 60 AIS, ICH, SAH Prospective Early TT is feasible, safe, and possibly
Randomized reduces sedation need and
mortality

Villwock et al.32 13.165 AIS, ICH, SAH Retrospective Early TT may reduce incidence of
VAP, LOS and lower hospital costs

Lee et al.41 95 AIS, ICH, SAH Retrospective No significant difference between

early and late TT
AIS: acute ischemic stroke; ICH: intracerebral hemorrhage; SAH: subarachnoid hemorrhage, TT: tracheostomy, ICU-LOS: intensive care unit-length of
stay; GCS: Glasgow coma scale, COPD: chronic obstructive pulmonary disease; VAP: ventilator-associated pneumonia; LOS: length of stay; ICU:
intensive care unit.

3. Anticipated need for prolonged (and at least)

assisted mechanical ventilation, based on a SET
score > 10 (Table 2), and clinical judgment of the
treating neurointensivist.
4. Informed consent by the patient and/or legal proxy.
5. Principal indication for TT (at least one of the
following):
a. ongoing demand for suctioning of tracheobron-
chial secretions,
b. central nervous system (CNS) – related respira-
tory insufficiency,
c. aspiration or danger of aspiration due to
dysphagia,
d. persistent, recurrent upper airway collapse pre-
venting adequate ventilation and oxygenation.
Exclusion criteria
Subjects presenting with any of the following criteria
will not be included in the trial:
1. Premorbid mRS >1.
2. Artificial ventilation for more than 4 days prior to
TT.
3. Any emergency situation that compromises the
patient’s safety, such as:
a. ICP persistently >25 mmHg,
b. difficult airway management, or anticipated tech-
nical problems with extubation/re-intubation,
c. contraindications for a PDT (Supplementary
Table 3, which can be found online with this
article, http://wso.sagepub.com),

International Journal of Stroke, 11(3)

372 International Journal of Stroke 11(3)

Table 2. In-house score for estimation of two-week ventila-

tion need
Screening and randomization
After admission, clinical and radiological examination,
Area of assessment Situation Points
verification of inclusion/exclusion criteria, and informed
Neurological Dysphagia 4 consent from the patients or (expectedly the rule) their
function legal representative, patients will be randomly assigned
to one of the two treatment groups: (1). PDT as soon as
Observed aspiration 3
possible, but not later than five days from intubation.
GCS on admission <10 3 (2). Ongoing orotracheal intubation (with attempts
to wean and extubate) and PDT from day 10 after
Neurological Brainstem 4 intubation if extubation was not accomplished.
lesion Randomization will be performed using a central web-
based randomization tool to achieve comparable treat-
Space-occupying 3
ment groups (www.randomizer.at). Block randomiza-
cerebellar
tion will be applied and stratified for the participating
Ischemic infarct >2/3 4 centers to achieve equal group sizes per centre.
MCA territory

ICH volume >25 ml 4 Intervention and treatment

a
Diffuse lesion 3 Intervention
Hydrocephalus 4 The intervention (PDT) is performed in the ‘‘early TT’’
group within five days from intubation, and in the
General organ (Neuro) surgical 2 ‘‘standard’’ group from day 10 after intubation if they
function/procedure Intervention
have not already been successfully extubated. For prag-
Additional respiratory 3 matic reasons, and still consistent with many trials on
disease early TT, the cut-off for early TT was put on the fifth day
from intubation. Since we still assume that most sites
PaO2/FiO2 < 150 2 will be able to achieve enrollment and – in the respective
patients – TT within three days, we do not consider a
APS (of APACHEII) > 20 4
relevant deviation from the pilot trial that would affect
LIS > 1 2 sample size calculations for SETPOINT2. The cut-off at
day 10 was determined to be a common standard of
Sepsis 3 care, based on several trials including the results of an
Note: Non-validated scoring tool based on TT predictors from the lit- analysis of TT practices in the US National Inpatient
erature, for details see reference 31. Estimation of at least 2 weeks of Sample,1,32,34,38 and surveys of >250 US and German
ventilatory support, if score sums up to > 10. Confirmation or refusal by neurointensivists (DS, JB, unpublished data). Each of
an experienced neurointensivist, who has the last judgment. these entities identified day 10 as a very common and
APS: acute physiology score; LIS: lung injury score; GCS: Glasgow
coma scale, MCA: middle cerebral artery; ICH: intracerebral
consistent point in time of the ICU course to proceed to
hemorrhage; PaO2: partial pressure of oxygen; FiO2: fraction of inspired TT if extubation failed or was not feasible.
oxygen. In PDT, the trachea is punctured through an anter-
a
Diffuse categorizes brain pathologies that do not have a clear one-sided ior neck incision, optionally under direct broncho-
hemispherical affection, such as in multilocular ICH, AIS or SAH. scopic visualization, and a blunt-tipped guidewire
inserted. Dilation is performed over the guidewire,
d. oxygenation impairment: positive end-expiratory and a TT tube is inserted on the final pass.42,43 In
pressure (PEEP) >12, or FiO2 >0.6. order to minimize variability in the process, this trial
will be limited to centers employing PDT techniques.
4. Expected need for a permanent surgical TT. The PDT procedure has been routinely employed at the
5. Pregnancy. leading trial centers for more than 15 years, and is con-
6. Participation in any other interventional trial. sidered to be standard of care.12 All study participants are
7. Life expectancy < 3 weeks. planned for PDT: contraindications for PDT that are
8. Patient/family unwilling or unlikely to opt for at apparent at screening lead to exclusion from the trial
least three weeks of aggressive therapy prior to con- (Supplementary Table 3). If conversion from PDT to sur-
sideration of transition to comfort measures/dis- gical TT becomes warranted just before, during or after
continuation of life support measures. the procedure, it is performed surgically at the bedside by

International Journal of Stroke, 11(3)

Schönenberger et al.
open operation including the attachment of the tracheal
mucosa to the skin thus creating an epithelialized tra-
cheostoma. This conversion does not lead to exclusion
from the trial but is documented as an event of interest.
Treatment
Conservative intensive care treatment is based on local
standards and in accordance with established American
Heart Association/American Stroke Association (AHA/
ASA) Guidelines,44–47 applied to both study groups in
identical fashion. The principle aim for both groups is to
realize an ICU-stay as short as possible and a transfer to
rehabilitation as fast as the individual state of disease
allows. In this medically complex patient population,
however, discussions and considerations among medical
team, caregivers and family on adequacy or withdrawal
of life support therapies (WLST) on a regular basis are
also part of the standard treatment, and are also applied
to both study groups without difference.
1. Intubation and mechanical ventilation: Patients
are intubated at a GCS score < 8, when there
are any signs of respiratory insufficiency (arterial
pO2 < 60 mmHg and/or pCO2 > 48 mmHg), re-
duced swallowing or coughing reflexes, or when
the airway is compromised. Earlier intubation (i.e.
for diagnostic/therapeutic procedures) is left at the
discretion of the physician in charge. Ventilation is
based on each institution’s standard operating pro-
cedure (SOP), and is weaned as soon as possible.
2. Sedation and analgesia: Analgesia and sedation are
routinely applied for pain, agitation and anxiety, and
to improve the comfort of intubation, mechanical
ventilation, and invasive procedures. The mode of
analgesia and sedation depends on the
estimated time for need of sedation, and is performed
by using either short-lasting or longer-lasting agents
according to standard local practice. Locally custom-
ary agents are routinely utilized to assure patient
comfort, and titrated to sedation scales such as a
sedation and agitation scale (SAS) of 3–448,49 or a
Richmond agitation-sedation scale at 2 to 4.50,51
3. Monitoring consists of standard ICU monitoring
comprising continuous blood pressure, heart rate,
arterial O2 saturation, respiratory rate, and tempera-
ture measurement. Disease-dependent extended
multi-modal neurological monitoring may include
varying patterns of ICP, cerebral perfusion, cerebral
oxygenation, cerebral temperature, and neurochem-
ical measurements. This bedside monitoring is com-
plemented by intermittent neuroradiological
imaging. Monitoring is disease-dependent and insti-
tution-specific and will not be controlled between
study groups.
373
4. Feeding and gastrointestinal management: This is
applied according to institutional SOPs to all
patients. Early enteral feeding and augmentation
of gastrointestinal motility are advised.
5. Blood pressure control: Blood pressure is managed
according to current AHA/ASA guidelines, and the
latest literature on the treatment of AIS, ICH, and
SAH, following institutional SOPs.
6. Body core temperature: Normothermia is recom-
mended. Elevated body temperature is treated as
soon as it exceeds 37.5 C. The maintenance of nor-
mothermia or the application of hypothermia is
based on institutional SOPs.
7. Blood glucose level: The recommended target
blood glucose level is 80–144 mg/dl (up to
8 mmol/l), using insulin if necessary.
Hypoglycemia is treated with infusions of 10% or
20% glucose solution.
8. Hemoglobin concentration: This is based on insti-
tutional SOPs.
9. Infection control: Standardized hygienic measures are
in place to avoid infections. Infections are screened
for daily by clinical examination, continuous tem-
perature measurements and laboratory assessments.
Antibiotic treatment is based on institutional SOPs.
10. ICP management: ICP is measured and managed
according to institutional SOPs; in summary, it is rec-
ommended at a sustained ICP over 20 mmHg, to
apply osmotic agents such as mannitol and hyper-
tonic saline that are applied as boluses or infusions,
before more definitive measures (e.g. surgical decom-
pression) are undertaken. Hy-perventilation is only
applied short-term as a ‘‘bridge’’ before surgery.
Escalation of non-surgical measures may include
the use of barbiturates or hypothermia.
11. Vasospasm management: Vasospasm, a common
complication of SAH, is diagnosed and managed
by institutional SOPs. Patients with aneurysmal
SAH will be managed according to guidelines,44
with identical strategies in both study groups.
12. Neurosurgery: Neurosurgical interventions, such
as decompressive hemicraniectomy, occipital trep-
anation, hematoma evacuation, aneurysm clipping,
and placement of ventricular drains or shunts, are
applied according to clinical necessity after consult-
ation of local neurosurgeons without principal dif-
ferences in approach between the two study groups.
13. Neuroendovascular interventions: In case of acute
large brain vessel occlusions, cerebral or precereb-
ral arterial stenoses, cerebral venous thrombosis,
cerebral aneurysms, or vasospasm, neuroendovas-
cular catheter interventions such as balloon dilata-
tion, stent placement, thrombectomy, coiling,
embolization, or flow-diversion may be warranted.
These measures are decided upon individually after
International Journal of Stroke, 11(3)
374 International Journal of Stroke 11(3)

Figure 1. Study flow chart of SETPOINT2.

Visit 1: Screening

of ICU-patients with AIS or ICH or SAH ventilated for less than 4 days

Check of inclusion / exclusion criteria, Informed consent

Randomization

Visit 2: Prolonged intubation

Visit 2: Early tracheostomy
Weaning / extubation or
Within 5 days after intubation tracheostomy
+ intensive care treatment ≥ 10 days after intubation
(n=190) + intensive care treatment
(n=190)

Visit 3: Endpoint documentation during hospital stay

Visit 4: Follow-up telephone interview 6 months after admission

AIS: acute ischemic stroke; ICH: intracerebral hemorrhage; SAH: subarachnoid hemorrhage; ICU: intensive care unit.

consultation with neuroendovascular specialists Secondary outcomes

without differences between the study groups, and
independent of the study protocol.
By this guideline-based and systematically standar-
dized management, we hope to ensure a uniform treat-
ment approach in both study groups.
The duration of acute inpatient therapy is highly
variable, and may range from 2 to 8 weeks or longer.
According to the trial protocol, screening, inclusion,
randomization, and documentation of the ICU-stay
are followed by a telephone interview with the patient,
caregiver, and/or treating physician at six months after
the stroke. Documentation of the hospital stay and tele-
phone interview is performed by investigators not
involved in the patient’s treatment, and blinded to the
timing of PDT. The six-month outcome interview is the
last follow-up in both arms (Study flow chart Figure 1,
Study protocol Supplementary Table 4).
Primary outcome
. Dichotomized functional outcome mRS score of 0–4
(favorable outcome) vs. 5,6 (unfavorable outcome)
at six months after admission to ICU.
. Mortality (death of any cause during the ICU-stay
or within six months after admission).
. ICU-LOS (days from admission until discharge from
the ICU).
. Hospital LOS (days spent at centre hospital from
admission to discharge).
. Accumulated duration of ventilation (sum of days
on the ventilator until the patient is ventilator-inde-
pendent for 24 h).
. Quality of life assessment (EuroQol will be used to
assess quality of life at six-months).
. Accumulated duration of weaning (sum of days
spent under the possible application of a weaning
protocol and spent within specific step-wise phases
of such a protocol).
. Accumulated duration until decannulation (sum of
days until decannulation).
. Accumulated duration of analgosedation depend-
ence (sum of days requiring the application of seda-
tives and analgesics which are also specified).
. Accumulated duration of vasopressor dependence
(sum of days spent under vasopressors).
. Accumulated rounds of antibiotic treatment (sum of
rounds of antibiotic treatment).

International Journal of Stroke, 11(3)

Schönenberger et al.
. Number and type of early adverse events (AEs) asso-
ciated with the procedure (from 2 h after TT to dis-
charge, number, and types of complications related
to TT (i.e. bleeding, mispositioning, malfunction,
replacement demand, etc.)).
. Number and type of late AEs associated with the
procedure (from discharge to follow-up, number,
and types of complications related to TT (i.e. recur-
rent or chronic infection, tracheal stenosis, surgical
revision of stoma, etc.)).
. Number and type of severe adverse events (SAE)
associated with the procedure (any AEs, related to
the procedure or not, that occur after enrollment
into the study with one of the following conse-
quences: death, life-threatening situation, prolonged
hospital stay or re-admission to hospital, and related
prolonged deterioration of health).
. Number of ICP increases > 25 mmHg (number of
episodes requiring treatment before and after TT).
. Costs of treatment (sum of actual hospital
charges).
Patient safety and data monitoring
AE monitoring will be managed through the primary
coordinating centers in Freiburg and Heidelberg in
Germany, and Maine in the United States, with an
actualized report of events issued to the medical moni-
tors every three months. The monitors will recommend
continuation or discontinuation of the trial to the
Steering Committee. All data specified in the trial proto-
col will be documented in standardized electronic case
report forms (eCRF) in the Research Electronic Data
Capture (RedCap) database by the co-investigators of
the centers and collected by the coordinating centre in
Heidelberg, Germany. Data validation, including con-
trol of completeness, consistency, and plausibility of
the data documented in the CRF is at the responsibility
of the local principal investigator (PI) and the document-
ing sub-investigators. Planning and implementation of
the trial procedures as well as the analysis are done in
consultation with the Institute of Medical Biometry and
Informatics (IMBI), Heidelberg.
The PI (JB) and the co-coordinators (WDN for
Germany, DS for USA) are trained by certification in
the practice and conduct of clinical trials according to
Good Clinical Practice (GCP) and the declaration of
Helsinki. They supervise all trial inclusions and adher-
ence to the trial protocol. All protocol-required infor-
mation for the clinical data collected during the trial
must be entered by the investigator, or by a designated
representative, in the eCRF. The investigator is respon-
sible for ensuring that all sections of the eCRF are com-
pleted correctly and that entries can be verified against
375
source data. Any entry and correction in the Remote
Data Entry System will be recorded automatically in
an audit file. Once the documentation of a patient is
completed and checked for plausibility, the investigator
is asked to date and sign it via electronic identification. In
order to guarantee a high quality of the data the com-
pleteness, validity and plausibility of data are examined
by validating programs, which thereby generate queries.
The sites are obliged to clarify or explain the queries. The
data will be managed and analyzed according to the
appropriate SOPs valid in the IMBI. All data manage-
ment and supervising procedures are aimed to be per-
formed according to SOPs and in accordance to
ICH-GCP Guidelines (E6) and the declaration of
Helsinki.
Statistical analysis and sample size
Sample size
Sample size considerations are based on the primary
endpoint ‘‘treatment success (mRS 0–4) after 6
months’’. Analysis of the pilot trial SETPOINT
showed a treatment success rate after six months of
30% in the control group (prolonged orotracheal intub-
ation) and a clinically relevant higher success rate of
48% in the intervention group (TT before day 3).
However, since several considerations of the authors
of the publication of the SETPOINT study40 led to
the conclusion that the true treatment effect might
indeed be slightly smaller than observed in this pilot
trial and since the authors considered an effect size of
15% still highly clinically relevant, sample size calcula-
tion for SETPOINT2 was performed assuming a treat-
ment success rate after six months of 30% in the control
group and of 45% in the experimental group.
The sample size was determined for the implemented
two-stage group sequential design according to O’Brien
and Fleming52 at overall two-side level a ¼ 0.05 and
with power 1  b ¼ 0.80 at the above specified alterna-
tive when applying a chi-square test. The calculations
were carried out using ADDPLAN, version 6.1.1.
A total of 326 patients (163 patients per group) are
required under these assumptions. For the primary
analysis, a binary logistic regression model will be
applied including the factor treatment group and the
covariates age, GCS at admission, and center. Applying
this adjusted analysis, it can be expected that the actual
power will be increased as compared to the power pro-
vided by the chi-square test. In analogy to the pilot trial
SETPOINT, we assume a rate of about 15% for drop-
outs and patients lost to follow-up. Although missing
values for outcome will be imputed for the primary ana-
lysis, there will be some loss of information due to incom-
plete data. For this reason, the total number of patients to
International Journal of Stroke, 11(3)
376
be randomized will be increased by about 15% to com-
pensate the potential dilution of the treatment effect
caused by information loss. The total number of patients
to be randomized in the SETPOINT2 trial is therefore
chosen as 380 (190 per group).
Statistical design and analysis
Group sequential design. To allow for early stopping in
case of an overwhelming large treatment effect, a two-
stage group sequential design with one interim analysis
is employed. The critical levels for the two-sided P-
values for rejecting the null hypotheses assessed in con-
firmatory analysis are calculated according to O’Brien
and Fleming52 to assure an overall two-sided level of
a ¼ 0.05.For this design, early stopping occurs only in
case of a large treatment effect (i.e. a small P-value
observed in the interim analysis), and (as a conse-
quence) the significance level to be applied in the final
analysis has to be adjusted only slightly. The interim
analysis is performed when the result for the primary
endpoint is available for one-third of the maximum
total sample size, i.e. for 127 patients). The related
two-sided local type I error rates for the interim and
final analysis, respectively, are a1 ¼ 0.0006 and
a2 ¼ 0.498 (calculations performed using ADDPLAN,
version 6.1.1.).
Statistical hypotheses
The primary aim of the trial is a comparison between
early TT (experimental treatment) and ongoing orotra-
cheal intubation (control) with respect to the propor-
tion of patients with mRS 0–4 (treatment success) six
months after admission to ICU. Secondarily, the over-
all death rates six months after admission to ICU
will be compared. The related null hypotheses state
that that there is no difference in the success rate or
overall death rate, respectively, between the treatment
groups.
Analysis sets
Each patient’s allocation to the different analysis popu-
lations (full analysis set (FAS), per-protocol (PP) ana-
lysis set, safety analysis set) will be defined prior to the
analysis. The allocation will be documented in the stat-
istical analysis plan. During the data review, deviations
from the protocol will be assessed as ‘‘minor’’ or
‘‘major’’. FAS is defined according to the ITT principle
and will thus include all randomized patients. Major
deviations from the protocol will lead to the exclusion
of FAS patients from the PP analysis set. The safety
analysis set includes all patients treated with one of the
interventions studied in this trial.
International Journal of Stroke, 11(3)
International Journal of Stroke 11(3)
Confirmatory analysis of primary endpoints
The multiplicity of the test problems to be assessed and
the hierarchy with respect to the importance of the
research questions are taken into account by applying
the multiple test procedure for a priori hierarchically
ordered hypotheses: The null hypothesis concerning
treatment success is tested first at local level ai, i ¼ 1,2
denoting the interim of final analysis, respectively. The
test procedure stops with acceptance of both null
hypotheses if these null hypotheses cannot be rejected;
the null hypothesis concerning overall death is tested at
local level ai. The procedure controls the experiment
wise type I error rate at a ¼ 0.05 if the local levels of
the two-stage O’Brien and Fleming group sequential
design52 given in the preceding subsection are applied
as described previously.53 The confirmatory analysis of
the two null hypotheses to be assessed will be done
using a binary logistic regression model including the
factor treatment group and adjusting for the covariates
age, GCS at admission, and centre (where US-
American and European centers are combined) at
two-sided local type I error rates a1 ¼ 0.0006 and
a2 ¼ 0.498 at the interim or final analysis, respectively
(see description of the multiple testing procedure
above). The crude and adjusted rate difference together
with the corresponding two-sided 95% as well as
repeated confidence intervals will be calculated. The
primary analysis will be conducted based on the full
analysis set which includes all randomized patients. In
case of patients lost to follow-up before evaluation of
the primary endpoint, missing data for the outcome will
be imputed using multiple imputation as described
elsewhere.54,55
Analysis of secondary endpoints
Analyses of secondary endpoints will be descriptive and
will include the calculation of appropriate summary
measures of the empirical distributions (continuous
variables: mean, standard deviation, median, interquar-
tile range, minimum, maximum; categorical variables:
frequencies, and percentages) as well as of descriptive
two-sided P-values. Graphical methods (e.g. boxplots,
Kaplan-Meier estimator curves for overall survival, and
length of ICU-stay) will be used to visualize the find-
ings. Sensitivity analyses will be conducted for the PP
set (patients without major protocol violations) and for
predefined subgroups. The safety analysis includes a
comparison of frequencies of AE and frequencies stra-
tified by intensity and causality. Furthermore, statis-
tical methods are used to assess the quality of data
and the homogeneity of intervention groups with
respect to baseline data. All analyses will be done
using SAS version 9.1 or higher or SPSS. A detailed
Schönenberger et al.
description of the analysis will be given in a statistical
analysis plan that will be finalized before start of the
evaluation.
Duration
Based on the number of patients treated in the partici-
pating centers within the past five years, the estimated
number of recruited patients averages 16 per month.
Thus, the duration of recruitment is estimated 2
years, which results in an estimated total trial duration
of 2.5 years after addition of 0.5 year for documenta-
tion of the last six-months-outcome endpoint.
Summary
It is presently unclear whether and when ventilated
patients with ischemic or hemorrhagic stroke should
receive a TT. Retrospective subgroup analyses suggest
that these patients could be particularly fast to wean off
the ventilator after early TT because their main impair-
ment is control of the upper airway and not breathing.
Thus, sedation could be stopped and rehabilitation
activities initiated sooner with earlier TT. The only ran-
domized pilot trial on this question (SETPOINT) sug-
gested benefits of an early TT including reduced need
for sedatives, reduced necessity of fully controlled ven-
tilation, and reduced mortality. However, the latter was
not the primary endpoint, and the pilot trial was not
sufficiently powered to demonstrate differences in func-
tional outcome. SETPOINT2 is the first randomized
multicentre trial on early TT vs. prolonged intubation
in ventilated patients with severe ischemic and hemor-
rhagic stroke. It aims to investigate potential benefits in
long-term functional outcome in the first place, and
additional parameters of the hospital and ICU course
of these patients in the second. The design is multicen-
tre, prospective, randomized, open-blinded endpoint
(PROBE). It has the potential to fundamentally
change airway management in cerebrovascular neuro-
critical care and to improve the functional outcomes of
patients with severe stroke.
Study organization
Steering committee
The Steering Committee consists of the PI (JB), the
co-coordinator for Germany (WDN) and the co-
coordinator for the USA (DS). They conduct the trial
and are responsible for development of the trial proto-
col, design of patient’s and legal representative’s infor-
mation and informed consent documents, approval of
the trial protocol, later amendments by legal authorities
and ethics committees, design of the CRF, organization
377
of a system for patient randomization, any decisions on
changes, amendments, communication with the local
ethics committee, or interruption of the trial. This is
an investigator-driven trial.
Data Safety and Monitoring Board (DSMB)
Data safety and monitoring will be in place before
enrolment begins, and monitoring will be performed
throughout subject enrolment and treatment. The
Data Safety Monitoring Board (DSMB) consists of
Dr. José I Suarez (lead, United States), Dr. Eric
Jüttler (Germany), and Dr. Niklas Nielsen (Sweden),
all neuro intensivists and very experienced clinical trial-
ists. The DSMB will be responsible for ongoing moni-
toring of reports on SAEs and AEs submitted by the
clinical centers to ensure good clinical care and to
quickly identify safety concerns. If necessary, it will
suggest measures to be taken to prevent the occurrence
of particular AE, e.g., modifying the protocol to require
frequent measurement of laboratory values predictive
of the event. In the event of unexpected SAEs or an
unduly high rate of SAEs, the DSMB will be respon-
sible for notifying the Steering Committee, which may
even result in stopping the trial according to stopping
rules created independently by the DSMB.
Ethical approval
The trial was approved by the Heidelberg local ethics
committee (Ethikkommission Medizinische Fakultät
Heidelberg), code S-489/2014. It will have to be
approved by equivalent institutions at each centre.
Trial registration
SETPOINT2 is a registered trial at
https://clinicaltrials.gov/ct2/show/NCT02377167
Initiated, actively enrolling centers (17 more
planned for participation)
Department of Neurology, University of Heidelberg,
Im Neuenheimer Feld 400, D-69120 Heidelberg,
Germany. Department of Neurology, University of
Freiburg, Breisacher Strasse 64, D-79106 Freiburg
i.Br., Germany. Department of Critical Care Services,
Maine Medical Center, 22 Bramhall St, Portland, Maine
04102, USA.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest
with respect to the research, authorship, and/or publi-
cation of this article.
International Journal of Stroke, 11(3)
378
Funding
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this
article: There is limited funding by local resources of the PI
and from a research grant by the foundation of the German
Neurocritical Care Society (DGNI-Stiftung) that will be used
for data management/analysis and administration costs. At
present, neither the patients, nor the centers will be compen-
sated financially – time of the local investigators and nurse-
coordinators will be donated by the local institutions. Should
funding become available in the course of the trial this may
change and/or, the trial may be extended with regard to sub-
investigations of additional questions or parameters. The
industry has no role in this trial.
References
1. Walcott BP, Kamel H, Castro B, Kimberly WT and
Sheth KN. Tracheostomy after severe ischemic stroke: a
population-based study. J Stroke Cerebrovasc Dis 2014;
23: 1024–1029.
2. Grotta J, Pasteur W, Khwaja G, Hamel T, Fisher M
and Ramirez A. Elective intubation for
neurologic deterioration after stroke. Neurology 1995;
45: 640–644.
3. Bushnell CD, Phillips-Bute BG, Laskowitz DT, Lynch
JR, Chilukuri V and Borel CO. Survival and outcome
after endotracheal intubation for acute stroke.
Neurology 1999; 52: 1374–1381.
4. Steiner T, Mendoza G, De Georgia M, Schellinger P,
Holle R and Hacke W. Prognosis of stroke patients
requiring mechanical ventilation in a neurological critical
care unit. Stroke 1997; 28: 711–715.
5. Gujjar AR, Deibert E, Manno EM, Duff S and Diringer
MN. Mechanical ventilation for ischemic stroke and
intracerebral hemorrhage: indications, timing, and out-
come. Neurology 1998; 51: 447–451.
6. Skolarus LE, Morgenstern LB, Zahuranec DB, Burke JF,
Langa KM and Iwashyna TJ. Acute care and long-term
mortality among elderly patients with intracerebral hem-
orrhage who undergo chronic life-sustaining procedures.
J Stroke Cerebrovasc Dis 2013; 22: 15–21.
7. Harvey RL, Roth EJ, Heinemann AW, Lovell LL,
McGuire JR and Diaz S. Stroke rehabilitation: clinical
predictors of resource utilization. Arch Phys Med
Rehabil 1998; 79: 1349–1355.
8. Seder DB, Riker RR, Jagoda A, et al. Emergency neuro-
logical life support: airway, ventilation, and sedation.
Neurocritical Care 2012; 17(Suppl 1): S4–S20.
9. McPhee LSD. The Neurocritical Care Airway. In: LK
(ed.) Neuro-ICU Book. New York: McGraw Hill, 2012,
pp. 655–671.
10. Nieszkowska A, Combes A, Luyt CE, et al. Impact of
tracheotomy on sedative administration, sedation level,
and comfort of mechanically ventilated intensive care
unit patients. Crit Care Med 2005; 33: 2527–2533.
11. Antonelli M, Michetti V, Di Palma A, et al. Percutaneous
translaryngeal versus surgical tracheostomy: a rando-
mized trial with 1-yr double-blind follow-up. Crit Care
Med 2005; 33: 1015–1020.
International Journal of Stroke, 11(3)
International Journal of Stroke 11(3)
12. Cobean R, Beals M, Moss C and Bredenberg CE.
Percutaneous dilatational tracheostomy. A safe, cost-
effective bedside procedure. Arch Surg 1996; 131: 265–271.
13. Delaney A, Bagshaw SM and Nalos M. Percutaneous
dilatational tracheostomy versus surgical tracheostomy
in critically ill patients: a systematic review and meta-
analysis. Critical Care 2006; 10: R55.
14. Freeman BD, Isabella K, Cobb JP, et al. A prospective,
randomized study comparing percutaneous with surgical
tracheostomy in critically ill patients. Critical Care Med
2001; 29: 926–930.
15. Freeman BD, Isabella K, Lin N and Buchman TG. A
meta-analysis of prospective trials comparing percutan-
eous and surgical tracheostomy in critically ill patients.
Chest 2000; 118: 1412–1418.
16. Seder DB, Lee K, Rahman C, et al. Safety and feasibility
of percutaneous tracheostomy performed by neurointen-
sivists. Neurocritical Care 2009; 10: 264–268.
17. Silvester W, Goldsmith D, Uchino S, et al. Percutaneous
versus surgical tracheostomy: a randomized controlled
study with long-term follow-up. Critical Care Med
2006; 34: 2145–2152.
18. Chevron V, Menard JF, Richard JC, Girault C, Leroy J
and Bonmarchand G. Unplanned extubation: risk factors
of development and predictive criteria for reintubation.
Critical Care Med 1998; 26: 1049–1053.
19. Coplin WM, Pierson DJ, Cooley KD, Newell DW and
Rubenfeld GD. Implications of extubation delay in
brain-injured patients meeting standard weaning criteria.
Am J Respir Crit Care Med 2000; 161: 1530–1536.
20. Nava S, Ambrosino N, Clini E, et al. Noninvasive mech-
anical ventilation in the weaning of patients with respira-
tory failure due to chronic obstructive pulmonary disease.
A randomized, controlled trial. Ann Intern Med 1998;
128: 721–728.
21. Freeman BD and Morris PE. Tracheostomy practice in
adults with acute respiratory failure. Crit Care Med 2012;
40: 2890–2896.
22. Rumbak MJ, Newton M, Truncale T, Schwartz SW,
Adams JW and Hazard PB. A prospective, randomized,
study comparing early percutaneous dilational tracheot-
omy to prolonged translaryngeal intubation (delayed
tracheotomy) in critically ill medical patients. Crit Care
Med 2004; 32: 1689–1694.
23. Griffiths J, Barber VS, Morgan L and Young JD.
Systematic review and meta-analysis of studies of the
timing of tracheostomy in adult patients undergoing arti-
ficial ventilation. BMJ 2005; 330: 1243.
24. Young D, Harrison DA, Cuthbertson BH, Rowan K and
TracMan C. Effect of early vs. late tracheostomy place-
ment on survival in patients receiving mechanical venti-
lation: the TracMan randomized trial. JAMA 2013; 309:
2121–2129.
25. Terragni PP, Antonelli M, Fumagalli R, et al. Early vs.
late tracheotomy for prevention of pneumonia in mech-
anically ventilated adult ICU patients: a randomized con-
trolled trial. JAMA 2010; 303: 1483–1489.
26. Scales DC, Thiruchelvam D, Kiss A and Redelmeier DA.
The effect of tracheostomy timing during critical illness on
long-term survival. Crit Care Med 2008; 36: 2547–2557.
Schönenberger et al.
27. Siempos II, Ntaidou TK, Filippidis FT and Choi AM.
Effect of early versus late or no tracheostomy on mortal-
ity and pneumonia of critically ill patients receiving
mechanical ventilation: a systematic review and meta-
analysis. Lancet Respir Med 2015; 3: 150–158.
28. Rodriguez JL, Steinberg SM, Luchetti FA, Gibbons KJ,
Taheri PA and Flint LM. Early tracheostomy for pri-
mary airway management in the surgical critical care
setting. Surgery 1990; 108: 655–659.
29. Bouderka MA, Fakhir B, Bouaggad A, Hmamouchi B,
Hamoudi D and Harti A. Early tracheostomy versus pro-
longed endotracheal intubation in severe head injury.
J Trauma 2004; 57: 251–254.
30. Dunham CM and LaMonica C. Prolonged tracheal
intubation in the trauma patient. J Trauma 1984; 24:
120–124.
31. Bösel J, Schiller P, Hacke W and Steiner T. Benefits of
early tracheostomy in ventilated stroke patients? Current
evidence and study protocol of the randomized pilot trial
SETPOINT (Stroke-related Early Tracheostomy vs.
Prolonged Orotracheal Intubation in Neurocritical care
Trial). Int J Stroke 2012; 7: 173–182.
32. Villwock JA, Villwock MR and Deshaies EM.
Tracheostomy timing affects stroke recovery. J Stroke
Cerebrovasc Dis 2014; 23: 1069–1072.
33. Kocaeli H, Korfali E, Taskapilioglu O and Ozcan T.
Analysis of intracranial pressure changes during early
versus late percutaneous tracheostomy in a neuro-inten-
sive care unit. Acta Neurochir 2008; 150: 1263–1267;
Discussion 7.
34. Qureshi AI, Suarez JI, Parekh PD and Bhardwaj A.
Prediction and timing of tracheostomy in patients with
infratentorial lesions requiring mechanical ventilatory
support. Crit Care Med 2000; 28: 1383–1387.
35. Rabinstein AA and Wijdicks EF. Outcome of survivors
of acute stroke who require prolonged ventilatory assist-
ance and tracheostomy. Cerebrovasc Dis 2004; 18:
325–331.
36. van der Lely AJ, Veelo DP, Dongelmans DA,
Korevaar JC, Vroom MB and Schultz MJ. Time to
wean after tracheotomy differs among subgroups of
critically ill patients: retrospective analysis in a mixed
medical/surgical intensive care unit. Respir Care 2006;
51: 1408–1415.
37. Huttner HB, Kohrmann M, Berger C, Georgiadis D and
Schwab S. Predictive factors for tracheostomy in neuro-
critical care patients with spontaneous supratentorial
hemorrhage. Cerebrovasc Dis 2006; 21: 159–165.
38. Pinheiro Bdo V, Tostes Rde O, Brum CI, Carvalho EV,
Pinto SP and Oliveira JC. Early versus late tracheostomy
in patients with acute severe brain injury. J Bras Pneumol
2010; 36: 84–91.
39. Szeder V, Ortega-Gutierrez S, Ziai W and Torbey MT.
The TRACH Score: clinical and radiological predictors
of tracheostomy in supratentorial spontaneous intracer-
ebral hemorrhage. Neurocrit Care 2010; 13: 40–46.
40. Bösel J, Schiller P, Hook Y, et al. Stroke-related Early
Tracheostomy versus Prolonged Orotracheal Intubation
in Neurocritical Care Trial (SETPOINT): a randomized
pilot trial. Stroke 2013; 44: 21–28.
379
41. Lee YC, Kim TH, Lee JW, Oh IH and Eun YG.
Comparison of complications in stroke subjects undergo-
ing early versus standard tracheostomy. Respir Care
2015; 60: 651–657.
42. Mallick A and Bodenham AR. Tracheostomy in critically
ill patients. Eur J Anaesthesiol 2010; 27: 676–682.
43. Durbin CG Jr. Tracheostomy: why, when, and how?
Respir Care 2010; 55: 1056–1068.
44. Connolly ES Jr, Rabinstein AA, Carhuapoma JR, et al.
Guidelines for the management of aneurysmal subarach-
noid hemorrhage: a guideline for healthcare professionals
from the American Heart Association/American Stroke
Association. Stroke 2012; 43: 1711–1737.
45. Morgenstern LB, Hemphill JC 3rd, Anderson C, et al.
Guidelines for the management of spontaneous intracer-
ebral hemorrhage: a guideline for healthcare profes-
sionals from the American Heart Association/American
Stroke Association. Stroke 2010; 41: 2108–2129.
46. Wijdicks EF, Sheth KN, Carter BS, et al.
Recommendations for the management of cerebral and
cerebellar infarction with swelling: a statement for health-
care professionals from the American Heart Association/
American Stroke Association. Stroke 2014; 45:
1222–1238.
47. Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines
for the early management of adults with ischemic stroke:
a guideline from the American Heart Association/
American Stroke Association Stroke Council, Clinical
Cardiology Council, Cardiovascular Radiology and
Intervention Council, and the Atherosclerotic Peripheral
Vascular Disease and Quality of Care Outcomes in
Research Interdisciplinary Working Groups: the
American Academy of Neurology affirms the value of
this guideline as an educational tool for neurologists.
Circulation 2007; 115: e478–e534.
48. Riker RR, Picard JT and Fraser GL. Prospective evalu-
ation of the sedation-agitation scale for adult critically ill
patients. Crit Care Med 1999; 27: 1325–1329.
49. Simmons LE, Riker RR, Prato BS and Fraser GL.
Assessing sedation during intensive care unit mechanical
ventilation with the Bispectral Index and the sedation-
agitation scale. Crit Care Med 1999; 27: 1499–1504.
50. Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond
agitation-sedation scale: validity and reliability in adult
intensive care unit patients. Am J Respir Crit Care Med
2002; 166: 1338–1344.
51. Ely EW, Truman B, Shintani A, et al. Monitoring sed-
ation status over time in ICU patients: reliability and
validity of the Richmond Agitation-Sedation Scale
(RASS). JAMA 2003; 289: 2983–2991.
52. O’Brien PC and Fleming TR. A multiple testing proced-
ure for clinical trials. Biometrics 1979; 35: 549–556.
53. Kieser M BP and Lehmacher W. Inference on multiple
endpoints in clinical trials with adaptive interim analyses.
Biom J 1999; 41: 261–277.
54. PD A. Missing Data. Sage University Papers Series on
Quantitative Applications in the Social Sciences
2001:Series 07-136.
55. van Buuren S. Flexible imputation of missing data.
Chapman & Hall/CRC Press, 2012.
International Journal of Stroke, 11(3)