Insulin For Glycaemic Control in Acute Ischaemic Stroke (Review)

Cochrane Database of Systematic Reviews
Insulin for glycaemic control in acute ischaemic stroke

(Review)
Bellolio MF, Gilmore RM, Ganti L
Bellolio MF, Gilmore RM, Ganti L.

Insulin for glycaemic control in acute ischaemic stroke.
Cochrane Database of Systematic Reviews 2014, Issue 1. Art. No.: CD005346.
DOI: 10.1002/14651858.CD005346.pub4.
www.cochranelibrary.com
Insulin for glycaemic control in acute ischaemic stroke (Review)

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS
HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Analysis 1.1. Comparison 1 Dependency or death, Outcome 1 Dependency or death at the end of the follow-up. . 29
Analysis 1.2. Comparison 1 Dependency or death, Outcome 2 Death. . . . . . . . . . . . . . . . . 30
Analysis 1.3. Comparison 1 Dependency or death, Outcome 3 Diabetes mellitus versus no diabetes mellitus. . . . 31
Analysis 1.4. Comparison 1 Dependency or death, Outcome 4 Less than 30 days versus 90 days of follow-up. . . . 32
Analysis 2.1. Comparison 2 Functional neurological outcome, Outcome 1 NIHSS or ESS at the end of the follow-up. 33
Analysis 2.2. Comparison 2 Functional neurological outcome, Outcome 2 Independent in daily activities. . . . . 34
Analysis 2.3. Comparison 2 Functional neurological outcome, Outcome 3 Diabetes mellitus versus no diabetes mellitus. 35
Analysis 2.4. Comparison 2 Functional neurological outcome, Outcome 4 Less than 30 days versus 90 days of follow-up. 36
Analysis 3.1. Comparison 3 Hypoglycaemia, Outcome 1 Symptomatic hypoglycaemia. . . . . . . . . . . . 37
Analysis 3.2. Comparison 3 Hypoglycaemia, Outcome 2 Hypoglycaemia (with or without symptoms). . . . . . 38
Analysis 4.1. Comparison 4 Mean glucose level, Outcome 1 Mean glucose level. . . . . . . . . . . . . . 39
ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Insulin for glycaemic control in acute ischaemic stroke (Review) i

[Intervention Review]
Insulin for glycaemic control in acute ischaemic stroke
M Fernanda Bellolio1 , Rachel M Gilmore1 , Latha Ganti2

1 Department of Emergency Medicine, Mayo Clinic, Rochester, Minnesota, USA. 2 NFSG Veterans Affairs Medical Center, Gainesville,
Florida, USA
Contact address: M Fernanda Bellolio, Department of Emergency Medicine, Mayo Clinic, Generose Building-G410, 200 First Street
SW, Rochester, Minnesota, 55905, USA. [email protected].
Editorial group: Cochrane Stroke Group.

Publication status and date: New search for studies and content updated (no change to conclusions), published in Issue 1, 2014.
Citation: Bellolio MF, Gilmore RM, Ganti L. Insulin for glycaemic control in acute ischaemic stroke. Cochrane Database of Systematic
Reviews 2014, Issue 1. Art. No.: CD005346. DOI: 10.1002/14651858.CD005346.pub4.
ABSTRACT
Background
People with hyperglycaemia concomitant with an acute stroke have greater mortality, stroke severity, and functional impairment when
compared with those with normoglycaemia at stroke presentation. This is an update of a Cochrane Review first published in 2011.
Objectives
To determine whether intensively monitoring insulin therapy aimed at maintaining serum glucose within a specific normal range (4 to
7.5 mmol/L) in the first 24 hours of acute ischaemic stroke influences outcome.
Search methods
We searched the Cochrane Stroke Group Trials Register (September 2013), CENTRAL (The Cochrane Library 2013, Issue 8), MED-
LINE (1950 to September 2013), EMBASE (1980 to September 2013), CINAHL (1982 to September 2013), Science Citation Index
(1900 to September 2013), and Web of Science (ISI Web of Knowledge) (1993 to September 2013). We also searched ongoing trials
registers and SCOPUS.
Selection criteria
Randomised controlled trials (RCTs) comparing intensively monitored insulin therapy versus usual care in adults with acute ischaemic
stroke.
Data collection and analysis
We obtained a total of 1565 titles through the literature search. Two review authors independently selected the included articles and
extracted the study characteristics, study quality, and data to estimate the odds ratio (OR) and 95% confidence interval (CI), mean
difference (MD) and standardised mean difference (SMD) of outcome measures. We resolved disagreements by discussion.
Main results
We included 11 RCTs involving 1583 participants (791 participants in the intervention group and 792 in the control group). We
found that there was no difference between the treatment and control groups in the outcomes of death or dependency (OR 0.99, 95%
CI 0.79 to 1.23) or final neurological deficit (SMD -0.09, 95% CI -0.19 to 0.01). The rate of symptomatic hypoglycaemia was higher
in the intervention group (OR 14.6, 95% CI 6.6 to 32.2). In the subgroup analyses of diabetes mellitus (DM) versus non-DM, we
found no difference for the outcomes of death and disability or neurological deficit. The number needed to treat was not significant
for the outcomes of death and final neurological deficit. The number needed to harm was nine for symptomatic hypoglycaemia.
Insulin for glycaemic control in acute ischaemic stroke (Review) 1
Authors’ conclusions
After updating the results of our previous review, we found that the administration of intravenous insulin with the objective of
maintaining serum glucose within a specific range in the first hours of acute ischaemic stroke does not provide benefit in terms of
functional outcome, death, or improvement in final neurological deficit and significantly increased the number of hypoglycaemic
episodes. Specifically, those people whose glucose levels were maintained within a tighter range with intravenous insulin experienced a
greater risk of symptomatic and asymptomatic hypoglycaemia than those people in the control group.
PLAIN LANGUAGE SUMMARY
Controlling high blood sugar levels with insulin in people who have had an acute ischaemic stroke
After a stroke, people with high levels of sugar in their blood have increased mortality regardless of their age, how severe the stroke was,
or what type of stroke they had. Insulin can reduce blood sugar levels. We do not know what the optimal level of blood sugar should
be after a stroke. We searched for trials that compared usual care with intensive insulin treatment (trying to keep blood sugar levels
within the normal range of 4 to 7.5 mmol/L) after stroke. We found 11 trials involving 1583 participants. Trying to keep the blood
sugar level within a tight range immediately after a stroke did not improve the outcomes of neurological deficit and dependency. It did,
however, significantly increase the chance of experiencing very low blood sugar levels (hypoglycaemia), which can be harmful and can
cause brain damage and death. On balance, the trials did not show any benefit from intensive control of blood sugar levels after stroke.
BACKGROUND 1982; Rehncrona 1981). Enhanced acidosis may exaggerate is-

chaemic neuronal injury by facilitating free radical formation, ac-
tivating pH-dependent endonucleases, or altering intracellular cal-
Description of the condition cium regulation (Barber 1967; Barry 1993; Combs 1992; Regli
1996; Rehncrona 1981; Siesjo 1985; Siesjo 1996). Moderately and
Hyperglycaemia (where the blood glucose concentration is greater
severely elevated blood glucose levels have been shown to enhance
than 6.1 mmol/L) on admission to hospital is common after acute
cortical intracellular acidosis and significantly worsen mitochon-
ischaemic stroke and occurs in up to two-thirds of all patients
drial function (Anderson 1999; Hoxworth 1999).
(Scott 1999). There has been much debate over the years as to
It has also been postulated that intracellular acidosis enhances
whether hyperglycaemia in people with acute stroke is the re-
glutamate release. Glutamate plays a central role in neuronal death
sult of a stress response (Candelise 1985; Jorgensen 1994), poor
because it activates post-synaptic glutamate receptors and leads to
glycaemic control in people with diabetes, or unrecognised dia-
an excessive influx of calcium, subsequent mitochondrial injury,
betes. Regardless, the majority of clinical trials have concluded
and cell death (Li 2000; Nedergaard 1996). In addition, the pre-
that hyperglycaemia predicts increased stroke mortality indepen-
synaptic release of glutamate mediates repeated waves of spreading
dently of age, stroke severity, or stroke type (Capes 2001; Gray
depression (a decrease of activity in the cortex), which is another
1987; Melamed 1976; Weir 1997). However, the uncertainty over
mechanism believed to propagate the necrosis of penumbral tissue
whether hyperglycaemia worsens prognosis through the augmen-
(Koistinaho 1999; Lo 2003).
tation of acute brain injury (Jorgensen 1994; Kiers 1992; Van
Previous studies in humans showed that people with acute is-
Kooten 1993; Weir 1997) or is merely a physiological response
chaemic stroke and hyperglycaemia had worse outcomes when
to acute stroke (Murros 1992; Murros 1993; O’Neill 1991; Woo
compared with those with normal glycaemic levels. Multiple ob-
1988) still remains.
servational studies have showed that people with hyperglycaemia
Hyperglycaemia has been shown to have a deleterious effect on
exhibited greater stroke severity and greater functional impair-
ischaemic brain tissue in numerous animal studies, and many
ment than those with normoglycaemia; people with hypergly-
mechanisms have been proposed to explain this phenomenon (de
caemia were 2.3 times more likely to be dead at 90 days compared
Courten-Myers 1989; Gisselsson 1999; Lin 1998; Myers 1977;
with those with normal glucose levels in one study (Stead 2009).
Pulsinelli 1982). One of the most consistent findings is the asso-
To date, most of the work done in humans in the field of hyper-
ciation between hyperglycaemia and cerebral acidosis (Pulsinelli
glycaemia and stroke has been limited to correlating the presence OBJECTIVES
of hyperglycaemia to ischaemic stroke outcome. The next logical
To determine whether intensively monitoring insulin therapy
step in this debate is to ascertain whether treating hyperglycaemia
aimed at maintaining serum glucose within a specific normal range
following stroke does, in fact, reduce mortality and improve func-
(4 to 7.5 mmol/L) in the first 24 hours of acute ischaemic stroke
tional outcome.
influences outcome.
Description of the intervention METHODS

Insulin has a number of effects on glucose metabolism including
inhibition of glycogenolysis and gluconeogenesis, increased glu-
cose transport into fat and muscle, increased glycolysis, and stim- Criteria for considering studies for this review
ulation of glycogen synthesis (Ramnanan 2010). All these mech-
anisms decrease the levels of circulating glucose. Insulin serves to
co-ordinate the use of alternative fuels (glucose and free fatty acids)
Types of studies
to meet the energy demands of the organism during fasting, exer-
cise, and stress (Farese 1991; Fielding 1998). All randomised controlled trials (RCTs) comparing insulin with
placebo, low dose insulin versus high dose insulin, or close mon-
itoring of glucose versus loose monitoring in people with acute
ischaemic stroke with blood glucose levels greater than 6.1 mmol/
How the intervention might work L were eligible for inclusion.
In acute stroke the initial ischaemia is rarely complete due to col-
lateral blood supply. The ischaemic penumbra is the area of hypop-
Types of participants
erfused, but still viable, tissue that surrounds a densely ischaemic,
hypoxic core. This is the area that is vulnerable to the metabolic Trials that included adults, older than 18 years of age, presenting
imbalances mentioned above. With hyperglycaemia, there is a re- within 24 hours of acute ischaemic stroke with serum glucose levels
duction in penumbral salvage and an increase in brain lactate lev- greater than 6.1 mmol/L were eligible for inclusion. We included
els (Parsons 2002). Therefore, it is believed that if the detrimen- people of either gender both with and without diabetes.
tal effects associated with hyperglycaemia are mediated primar- Definite ischaemic stroke implies people in whom computed to-
ily through metabolic mechanisms, tight glycaemic control dur- mography (CT) or magnetic resonance imaging (MRI) was per-
ing acute stroke may reduce the extent of ultimate brain injury formed before randomisation. Presumed ischaemic stroke refers
(Bruno 2004). Acute hyperglycaemia is associated by neuroimag- to people who did not have a CT or MRI so haemorrhage could
ing with reduced salvage of penumbral tissue and greater final in- not be excluded.
farct size (Parsons 2002), and hyperglycaemia is associated with
reduced benefit from recanalization with thrombolytic therapy
Types of interventions
(Bruno 2002). Also, experimental studies have showed that ad-
ministration of insulin during focal and global ischaemia may re- We included trials that evaluated intervention with insulin, or
duce subsequent brain damage (Auer 1998; Hamilton 1995; Voll tight glycaemic control, to maintain a glycaemic level between 4
1989; Voll 1991; Wass 1996; Zhu 1994). and 7.5 mmol/L (72 to 135 mg/dL). The intervention had to start
within 24 hours of symptom onset. The control interventions were
placebo, no treatment, or loose control with insulin.
Why it is important to do this review

Types of outcome measures
Previous versions of our review (Bellolio 2008; Bellolio 2011)
showed that the use of insulin to maintain glycaemia in a tight
range had no improvement in functional outcome, death, or final
neurological deficit and significantly increased the number of hy- Primary outcomes
poglycaemic episodes. Death or dependency at the end of the scheduled follow-up. De-
The purpose of this review is to update the previous review with pendency is defined as being severely dependent on others in ac-
new information to ascertain whether the use of insulin to control tivities of daily living, or being significantly disabled; this corre-
hyperglycaemia within a tight range in the acute phase of ischaemic sponds to a Barthel Index score 60 or less, or a modified Rankin
stroke influences outcome. Scale grade 3 to 6 (Uyttenboogaart 2007).

Secondary outcomes Two review authors (MFB and RMG) independently screened the
1. Measures of neurological deficit after intervention for titles and abstracts of the records identified from the electronic
hyperglycaemia, including the National Institutes of Health searches and excluded obviously irrelevant studies. We obtained
Stroke Scale (NIHSS) and the European Stroke Scale (ESS). the full texts of the remaining papers and the same review authors
2. Number of deaths in each group at the end of the scheduled selected studies for inclusion based on the predefined criteria. The
follow-up. review authors were not blinded to the journal, institution, or
3. Hypoglycaemia, defined as glucose concentration less than study authors.
3 mmol/L. Symptomatic hypoglycaemia was defined as
confusion, visual disturbances, seizures, sweating, or hunger in a
person with a glucose level lower than 3 mmol/L. The numbers
Data extraction and management
of hypoglycaemic events (symptomatic and asymptomatic) in
each group were compared. Using a standardised data extraction form and working in dupli-
cate, two review authors (MFB and RMG) abstracted the following
descriptive data from every study: description of randomisation,
concealment of participants, blinding, intention to treat and lost
Search methods for identification of studies
to follow-up as part of the quality assessment. The same two re-
See the ’Specialized register’ section in the Cochrane Stroke Group view authors also collected information on: definitive or presumed
module. We did not apply any language restrictions and arranged acute ischaemic stroke (based on the use of CT or MRI before or
translation of relevant articles published in languages other than after randomisation), number of participants, gender, age, propor-
English. tion of people with diabetes, setting (stroke unit, critical care unit,
or non-critical care unit), characteristics of treatment and control
interventions (insulin dose, frequency, route, and duration), co-
Electronic searches interventions, and frequency of the glucose level checks.
We searched the Cochrane Stroke Group Trials Register (last We abstracted the specific stroke outcomes from every study. We
searched September 2013). We also searched MEDLINE (1950 to collected end-of-study deaths, Barthel score, Rankin Scale, Na-
September 2013) (Appendix 1), the Cochrane Central Register of tional Institutes of Health Stroke Scale (NIHSS), European Stroke
Controlled Trials (CENTRAL) (The Cochrane Library 2013, Issue Scale (ESS), and the number of hypoglycaemic events (symp-
8) (Appendix 2), EMBASE (1980 to September 2013) (Appendix tomatic and asymptomatic). Outcomes were collected at the end
3), Science Citation Index (1900 to September 2013) (Appendix of follow-up. Follow-up was defined for each study as ≤ 30 or ≤
4), Web of Science (1993 to September 2013) (Appendix 5), and 90 days. We contacted all the authors of the included studies, if
CINAHL (1982 to September 2013) (Appendix 6). The Cochrane data were missing or unclear, to obtain full details.
Stroke Group Trials Search Co-ordinator developed the MED- The same two review authors resolved any discrepancies between
LINE search strategy, which was adapted for the other databases the two sets of data through discussion. One author (MFB) entered
by a reference librarian who also ran the searches. the data into the Review Manager software, RevMan 5.2 (RevMan
In an effort to identify further published, unpublished, and ongo- 2012) and performed the analyses.
ing trials we searched:
1. Stroke Trials Directory (www.strokecenter.org/trials/),
2. ClinicalTrials.gov ( http://clinicaltrials.gov/),
3. Current Controlled Trials ( www.controlled-trials.com/), Assessment of risk of bias in included studies
4. SCOPUS ( www.scopus.com/). Two review authors (MFB and RMG) independently assessed risk
of bias for each study using the criteria outlined in the Cochrane
Handbook for Systematic Reviews of Interventions (Higgins 2011).
Searching other resources We resolved any disagreements by discussion. We assessed the risk
We checked the reference lists of the selected articles to identify of bias according to the following domains.
additional trials. 1. Random sequence generation.
2. Allocation concealment.
3. Blinding of participants and personnel.
Data collection and analysis 4. Blinding of outcome assessment.
5. Incomplete outcome data.
We graded the risk of bias for each domain as high, low, or unclear
and provided information from the study report together with a
Selection of studies justification for our judgment in the ’Risk of bias’ tables.

Measures of treatment effect off value of less than 3 mmol/L and defined symptomatic hypo-
The treatment was insulin and the measurements of treatment glycaemia as confusion, visual disturbances, seizures, sweating, or
effect were glucose levels and functional scales including death, hunger in a person with a glucose level lower than 3 mmol/L.
Barthel score, Rankin Scale, NIHSS, ESS, and the number of For continuous outcomes we determined the pooled mean dif-
hypoglycaemic events (symptomatic and asymptomatic). ference (MD) between the treatment and control interventions
and the associated 95% confidence interval (CI) using a fixed-
effect model. For continuous outcomes using different scales (for
Unit of analysis issues example for assessment of the final deficit with the NIHSS and
The glucose level is a continuous variable and was provided in mg/ ESS) we determined the pooled SMD between the treatment and
dL or mml/L; all the values were converted to the same unit (mml/ control interventions and the associated 95% CI using a fixed-
L) prior to any analysis. effect model. For binary outcomes we determined the odds ratio
When the scales used different units of measure, that is Barthel and (OR) between the treatment and control interventions and the
ESS to evaluate functional outcome, we used standardised mean associated 95% CI using a fixed-effect model.
differences (SMDs) as recommended by the Cochrane Handbook When any of the values in the outcome was zero, a 0.5 continuity
for Systematic Reviews of Interventions (Higgins 2011). correction factor was applied automatically by the RevMan soft-
ware (0.5 added to all cells).
Dealing with missing data

We contacted all the authors of the included studies, if data were Subgroup analysis and investigation of heterogeneity
missing or unclear, to obtain full details regarding study design, We determined subgroup analyses a priori. We compared:
outcomes, and attrition rates. When the data were not available, we 1. diabetic versus non-diabetic cohorts for the outcomes of
included the study in the qualitative analysis. For missing measures death and dependency, and final neurological deficit. Most of the
of precision, that is missing standard deviations, we estimated these studies did not report results separated by diabetic status so we
with the available information, like interquartile range or standard did the analysis comparing the cohorts with more than 50%
error, according to the Cochrane Handbook for Systematic Reviews diabetic versus less than 50% diabetic;
of Interventions (Higgins 2011). 2. outcomes measured at less than 30 days versus 90 days.
To explore these hypotheses, we estimated the difference in treat-
Assessment of heterogeneity ment effects between subgroups or treatment-subgroup interac-
tions (Altman 2003).
We assessed clinical heterogeneity by determining whether the
We quantified heterogeneity and inconsistency in the studies using
characteristics of the participants, interventions, outcome mea-
the I2 statistic, which describes the proportion of variance across
sures, and timing of outcome measurement were similar across
studies not due to chance.
studies. We assessed statistical heterogeneity using the I2 statistic.
Assessment of reporting biases Sensitivity analysis

When the data were not available we contacted the study authors We determined sensitivity analyses a priori. We excluded:
for more information. If data were still missing we classified the 1. studies with presumed diagnosis of ischaemic stroke (studies
study as unclear. For the assessment of publication bias, we evalu- that did not perform CT or MRI before the randomisation,
ated funnel plots if at least 10 studies examined the same outcome. therefore haemorrhagic strokes were not excluded);
2. studies where the controls received insulin (loose control
studies);
Data synthesis 3. studies with inadequate allocation concealment;
We used RevMan 5.2 (RevMan 2012) for all the meta-analyses. 4. the largest study.
We analysed the primary outcome of death or dependency at the
end of the scheduled follow-up as a binary outcome.
For the secondary outcome of neurological deficit after interven-
tion, the NIHSS and the ESS are ordinal scales with a minimum
increment of one point and we used these as continuous variables RESULTS
for the purposes of the analyses. The NIHSS ranges from 0 to
42 points, with 0 points being no measurable neurological deficit.
The ESS ranges from 0 to 100 points, with 0 points being bedrid- Description of studies
den. We analysed hypoglycaemia as a binary outcome with a cut-

Results of the search
We obtained a total of 1565 articles and, after removing duplicates,
we manually screened 1380 records. We retrieved a total of 16
trials for more detailed evaluation. We included four new trials
and one study was the published version of a previously included
study (Staszewski 2011). See Figure 1 for details.
Figure 1. Study flow diagram.
proportion of people with diabetes mellitus.

Included studies
In GIST-UK 2007,16% of the cohort had a disease different from
We included a total of 11 trials involving 1583 participants (791 ischaemic stroke but with a measurable neurological impairment
participants in the intervention group and 792 in the control (for example intracranial haemorrhage).
group). Three ongoing studies at the time of our previous review In nine studies participants had a definitive diagnosis of acute
were completed and have been included in this review (Azevedo ischaemic stroke prior to randomisation. All 11 studies reported
2009; INSULINFARCT 2012; McCormick 2010). intervention with an intravenous insulin infusion. Vinychuk 2005
Enrolled participants were typically older people (mean age 74 reported the results separately for participants with and without
years) with equal gender distribution. Baseline characteristics were diabetes mellitus.
similar in the intervention and control groups, with higher baseline For the functional outcome, four studies used the Barthel score
glycaemic levels measured at admission in the cohorts with a higher
and six studies used the modified Rankin Scale. Walters 2006 did 2010), or (3) stroke was an outcome of the study but the study
not report functional outcome scales. was not done on stroke participants (CIMT Trial; Miyashita 2008;
One study in the previous version of this review (Staszewski 2007) VADT).
is now included as a published study (Staszewski 2011).
For the neurological deficit assessment, six studies reported the Risk of bias in included studies
NIHSS and one reported the ESS at the end of the follow-up.
The NIHSS is a 42-point score; a higher score indicates a worse All studies were randomised with a parallel design. GRASP 2009
functional outcome and higher level of disability. The ESS score was stratified by glucose concentration, and Vriesendorp 2009 was
ranges from 0 to 100 points, 0 being bedridden and a higher stratified by dysphagia and diabetes mellitus (DM). All studies
score showing a better functional outcome. Ten studies reported except Vriesendorp 2009 had adequate generation of randomisa-
hypoglycaemia and eight reported deaths; Vinychuk 2005 did not tion. There was a high risk of bias in allocation concealment in
report these outcomes. five of the 11 studies, and most of the assessors of outcomes were
Azevedo 2009 was available in abstract format only and we con- not blinded.
tacted the author to obtain more information. We also contacted Four studies had a high risk of bias secondary to inadequate al-
the author of an ongoing study (NCT00373269) and the study location: in Staszewski 2011 the list was read by the investigator
is currently inactive. We contacted other authors for the previous entering the participant into the trial; Vriesendorp 2009 used con-
version of this review. secutive envelopes; and Azevedo 2009 and Kreisel 2009 did not
report this information. In none of the studies were physicians
blinded to the intervention. Participants were blinded to the in-
tervention in Azevedo 2009 and THIS 2008, and in four studies
Excluded studies the outcome assessment was blinded (GIST-UK 2007; GRASP
Excluded studies were those that did not meet the inclusion crite- 2009; Staszewski 2011; THIS 2008).
ria. Some studies were not included because: (1) they were not ran- GRASP 2009 and Walters 2006 were pilot studies, and GIST-UK
domised trials, (2) they had a mixed critical care population with 2007 was stopped early due to the slow enrolment rate. See the
less than 5% of the participants having ischaemic stroke (Green summary risk of bias in Table 1 and Figure 2.

Figure 2. Risk of bias summary: review authors’ judgements about each risk of bias item for each included
study.

Allocation
level in the intervention group was 6.7 mmol/L, and 7.3 mmol/L
Six studies had adequate allocation, two studies had a high risk of in the loose control group.
bias because either the list was read by the investigator (Staszewski
2011) or consecutive envelopes were used (Vriesendorp 2009),
and two studies did not provide this information (Kreisel 2009; Dependency or death
McCormick 2010). Modified Rankin grades and Barthel scores were used. Meta-anal-
ysis of the nine comparisons (1516 participants) found no signif-
icant difference between the treatment and control groups, with
Blinding an OR of 0.99 (95% CI 0.79 to 1.23, Analysis 1.1).
All the studies had a high risk of bias with regard to blinding.
Most of the studies were not blinded to the treatment arm and
had different treatments and co-interventions for the intervention Death
and control groups. In no studies were the physicians blinded to Meta-analysis of the nine comparisons (1422 participants) found
the intervention. Two studies (Azevedo 2009; THIS 2008) had no significant difference between the treatment and control
participants blinded to the intervention. Four studies had outcome groups, with an OR of 1.09 (95% CI 0.85 to 1.41, Analysis 1.2).
assessors blinded to the allocation group (GIST-UK 2007; GRASP
2009; Staszewski 2011; THIS 2008).
Final neurological deficit
NIHSS and ESS were used. Meta-analysis of the eight comparisons
Incomplete outcome data (1432 participants) found no significant difference between the
Most of the studies had an adequate attrition rate and the loss treatment and control groups, with a pooled SMD of -0.09 (95%
to follow-up was between 0% (Staszewski 2011; THIS 2008; CI -0.19 to 0.01, Analysis 2.1).
Vinychuk 2005; Walters 2006) to 15.2% (Vriesendorp 2009). The When using the dichotomised outcome of dependency on others
loss to follow-up was 7.4% in GIST-UK 2007, 7.5% in Kreisel for functional activities, the comparison of the nine studies had
2009, 1.4% in GRASP 2009, unknown in Vinychuk 2005 but an OR of 1.03 (95% CI 0.81 to 1.32, Analysis 2.2).
estimated to be 2.3%, estimated to be 15.2% in Vriesendorp 2009,
and 2.2% in INSULINFARCT 2012. Loss to follow-up was not
Hypoglycaemia
reported in Azevedo 2009 or McCormick 2010. All the studies
used intention-to-treat (ITT) analyses. One study reported results Meta-analysis of the 10 comparisons (1455 participants) found a
per protocol and used ITT (INSULINFARCT 2012). GIST-UK significant difference in the incidence of hypoglycaemia between
2007 was terminated early because of the low enrolment rate. the treatment and control groups, with an OR of 14.6 (95% CI
6.6 to 32.2) for symptomatic hypoglycaemia and an OR of 18.4
(95% CI 9.1 to 37.3) in participants with or without symptoms
Selective reporting of hypoglycaemia (Analysis 3.1 and Analysis 3.2).
It is possible that there was reporting bias. However, we did not
detect significant selective reporting among the included studies.
Subgroup analyses
Most of the studies had a functional outcome reported as well as
death and hypoglycaemia as outcomes.
Diabetes mellitus versus no diabetes mellitus (DM)
Other potential sources of bias For the outcome of dependence or death we compared those stud-
Azevedo 2009 was available in abstract format only thereby in- ies with 50% or more of the cohort having DM versus the studies
creasing the risk of bias and risk to methodological quality. Full with less than 50% of the cohort with DM. Vinychuk 2005 re-
text articles were available for the other 10 studies. ported these cohorts separately. The studies included in the DM
group were THIS 2008 (100% DM in the intervention group
and 73.3% in the control group), Walters 2006 (53.9% in the
Effects of interventions intervention and 50% in the control group), and GRASP 2009
The mean glycaemic level during treatment was significantly lower (50% in the intervention and 64% in the control group). The
in the intervention group than in the control group, with an MD of studies included in the no DM cohort were GIST-UK 2007 (17%
-0.63 (95% CI -0.80 to -0.46, Analysis 4.1). The average glucose in the intervention and 16% in the control), Staszewski 2011 (0%

DM), Kreisel 2009 (30% DM in the intervention and 35% in DISCUSSION
the control group), McCormick 2010 (28% in the intervention
and 40% in the control group), INSULINFARCT 2012 (8% in Our updated systematic review and meta-analyses found similar
the intervention and 17% in the control group), and Vriesendorp results to those published two years ago and suggest that the admin-
2009 (35% in the intervention and 40% in the control group). istration of intravenous insulin in an attempt to maintain serum
The meta-analysis found no significant difference between the glucose within a specific normal range in the first hours of acute
treatment and control groups for the outcome of dependency or ischaemic stroke does not provide benefit in terms of functional
death when stratified by DM status, with an OR of 0.66 (95% CI outcome or deaths at 30 days or 90 days, or improvement in final
0.35 to 1.24) for DM and an OR of 1.02 (95% CI 0.81 to 1.30) neurological deficit.
for participants without DM (Analysis 1.3).
While there was no difference in net benefit noted between the
For the outcome of final neurological deficit, the SMD for partic-
treatment and control groups, there was a significant difference
ipants with DM was -0.06 (95% CI -0.43 to 0.31) and for partic-
between the groups with regard to the adverse event of hypo-
ipants without DM it was -0.08 (95% CI -0.19 to 0.03, Analysis
glycaemia. Specifically, those participants who were maintained
2.3).
within a more tight range of glycaemia with intravenous insulin
experienced a greater risk of symptomatic and asymptomatic hy-
Follow-up poglycaemia than those in the control group.
For the outcome of dependency or death, we compared the stud-
When examining the endpoint of dependency or death, subgroup
ies with the outcome measured earlier than 30 days versus 90
analyses showed that there was no difference between participants
days. The studies with 30 days of follow-up were: Azevedo 2009,
with DM and those without, and there was no difference when
Kreisel 2009, McCormick 2010, Staszewski 2011, Vinychuk
looking at the outcome at 30 days or 90 days.
2005, Vriesendorp 2009, and Walters 2006. However, Staszewski
2011 and Walters 2006 did not report this outcome. The studies For the endpoint of final neurological deficit, we found no dif-
with 90 days of follow-up were GIST-UK 2007, GRASP 2009, ference in the mean NIHSS or ESS scores between the interven-
INSULINFARCT 2012, and THIS 2008. Meta-analysis of the tion and control groups. In the subgroup analyses there was no
comparisons found no significant difference between the treat- statistically significant difference between the DM and non-DM
ment and control groups when stratified by follow-up, with an cohorts. This means that the treatment did not appear to be more
OR of 0.74 (95% CI 0.43 to 1.25) for 30 days and an OR of 1.05 effective in one or the other group of participants.
(95% CI 0.82 to 1.34) for 90 days (Analysis 1.4).
For the outcome of final neurological deficit, we found a statisti- For the subgroup analyses of neurological deficit at 30 days ver-
cally significant difference between 30 days and 90 days of follow- sus 90 days, the effect of the intervention was favourable in those
up, in favour of the intensive glucose control group, with a SMD studies that reported the outcome at 30 days but not at 90 days;
of -0.47 (95% CI -0.72 to -0.23) for 30 days and 0.00 (95% CI this result is difficult to interpret because the natural history of
-0.12 to 0.11) for 90 days of follow-up (Analysis 2.4). ischaemic stroke is that the outcomes improve and stabilize by
three months, which is likely to be why this time point is more
common and reliable in studies of ischaemic stroke. Another pos-
Sensitivity analyses
sible explanation for this result is better quality of the studies that
1. Excluding the studies with a presumed diagnosis of reported the outcomes at 90 days; as these studies were bigger and
ischaemic stroke (GIST-UK 2007; Vriesendorp 2009): the OR accounted for 81% of the total number of participants included in
for death or dependency (primary outcome) was 0.82 (95% CI the review; two of the three studies were blinded to the allocation
0.58 to 1.17). group when evaluating the neurological and functional outcomes.
2. Excluding studies where the controls received insulin: in all In comparison, for the 30-day outcome only one of the five studies
the studies except Walters 2006 the control group received was blinded for the neurological and functional outcome assess-
insulin. The range of glycaemia was different in the different ments.
studies. The study by Vinychuk 2005 did not report the
intervention in the control group. When plotting the results in a two by two table we found that
3. Excluding studies with inadequate allocation concealment the number needed to treat (NNT) or to harm (NNH) for the
(Azevedo 2009; Kreisel 2009; McCormick 2010; Staszewski outcome of death was 91, meaning that one in every 91 patients
2011; Vriesendorp 2009 ). The outcome of dependency or death could be harmed by the treatment, with an absolute risk increase
had an OR of 1.0 (95% CI 0.79 to 1.26). of 1.10% in the treatment group (95% CI -3.3 to 5.5%). Because
4. Excluding the largest study: this was the GIST-UK 2007 the confidence interval includes positive and negative numbers,
study because the results were the same as above with an OR of the NNT or NNH was not significant between the treatment and
0.82 (95% CI 0.58 to 1.17). control groups for this outcome.

The NNT for the outcome of dependency or death was 419, We included 11 trials involving 1583 participants (791 interven-
meaning that one in every 419 patients could benefit from the tion and 792 controls). There was an intermediate risk of bias
treatment, with an absolute risk reduction of 0.24% % in the overall. The allocation concealment was adequate in most of the
treatment group (95% CI -4.5% to 5.0%). Because the confidence studies, but most of the studies were not blinded. The heterogene-
interval includes positive and negative numbers, the NNT was ity across the studies was adequate.
not significant between the treatment and control groups for this
outcome.
The NNH for the outcome of symptomatic hypoglycaemia was Potential biases in the review process
nine, meaning that one in every nine patients will have a symp- The results of this systematic review and meta-analysis should be
tomatic hypoglycaemic event with an absolute risk increase of interpreted with caution for several reasons. First, only 11 studies
11.5%% in the treatment group (95% CI 9.1 to 13.9%). This is were available for inclusion in the analyses. Second, the studies
statistically significant. The 95% CI for the NNT ranges from 7.2 were performed in a stroke unit and not in the general ward. Spe-
to 11.0. cialised units for the management of people with stroke have been
shown to improve the outcomes of those people presenting with
Remaining questions include how tight the control must be to ob-
haemorrhagic and ischaemic strokes, and should be the preferred
tain the most benefit without increasing the risk for severe hypo-
setting (Langhorne 1993).
glycaemia, and whether an acuity level exists in which this benefit
For the meta-analyses we decided to use the fixed-effect method
is not clearly visualized (Bochicchio 2008). The extent to which
instead of the random-effects method because of a suspected ac-
intensive insulin therapy and tight control of blood glucose im-
ceptable level of heterogeneity across the studies (Cook 1995).
prove the outcome after ischaemic neurological insults remains
The random-effects method assumes that the different studies are
unclear. The benefit of such treatment regimes may be negated
estimating different, yet related, treatment effects (DerSimonian
by the hypoglycaemic episodes, which may aggravate neurologi-
1986). The random and inverse variance methods will give similar
cal injury (Prakash 2008). Although it seems sensible to control
results when there is no heterogeneity among the studies. Where
hyperglycaemia in people with neurological injury the treatment
there is heterogeneity, confidence intervals for the average treat-
must account for potential hypoglycaemic episodes, which can be
ment effect will be wider if the random-effects model rather than a
detrimental. Therefore, it seems prudent that we accept slightly
fixed-effect model is used, and corresponding claims of statistical
less tight blood glucose control than in critically ill people without
significance of the treatment effect will be more conservative. It is
neurological injury.
also possible that the central estimate will change if there are rela-
tionships between the observed treatment effects and sample sizes.
The fixed-effect model is based on the mathematical assumption
Summary of main results that a single common effect underlies every study in the meta-
Our review suggests that the administration of intravenous insulin analysis; this assumes that there is low heterogeneity among the
in an attempt to maintain serum glucose within a specific normal studies.
range in the first hours of acute ischaemic stroke does not provide When evaluating heterogeneity the readers must be cautious in in-
benefit in terms of functional outcome, disability, improvement terpreting the I2 statistic because of the moderate number of stud-
in final neurological deficit, or death at 30 or 90 days, and confers ies included and the small number of participants in each study.
a greater risk of symptomatic and asymptomatic hypoglycaemia. Therefore, tests that do not reject the null hypothesis do not nec-
essarily indicate absence of heterogeneity (Montori 2003). How-
ever, we decided that it was sensible to pool the results. Hetero-
geneity comes from clinical differences. For instance, the biggest
Overall completeness and applicability of study (GIST-UK 2007) included 16% of participants without an
evidence ischaemic stroke, contributing to some degree to heterogeneity.
The studies identified in this review are sufficient to address the Another source of heterogeneity is differences in the treatment ef-
objectives of the review, and the population that is included al- fects secondary to a different methodological approach (Montori
lows for good generalization of the results and external validity. 2003); for instance, few studies were blinded to the allocation
The studies include more than 1000 participants from different group when assessing the outcome. To explore heterogeneity we
countries overall. These results discourage the maintenance of a performed subgroup analyses. One particular problem with the
tight glycaemic control. GIST-UK 2007, the biggest included study, is that there might
be confounding that is secondary to lower blood pressure levels
in the intervention group. This study found that the overall mean
plasma glucose level and mean systolic blood pressure were sig-
Quality of the evidence
nificantly lower in the glucose-insulin-potassium group (MD in

glucose 0.57 mmol/L, P < 0.001; MD in blood pressure 9.0 mm Studies in favour of tight glucose control include Van den Berghe
Hg, P < 0.0001). 2001, which showed that insulin therapy decreases mortality in
For the secondary outcome of final neurological deficit, we used people with DM and those undergoing surgery. Another study re-
ordinal scales (NIHSS and ESS) as continuous outcomes, and we ported beneficial outcomes with strict blood glucose control after
made the assumption that they followed a normal distribution for cardiac surgery (Ingels 2006). A study by Lazar et al found that
the purposes of the meta-analyses, but we acknowledge that in tight glycaemic control in people with DM undergoing coronary
clinical practice most of the functional scales have a skewed distri- artery bypass graft (CABG) improved perioperative outcomes, en-
bution in favour of better scores. We are aware that the methods hanced survival, and decreased the incidence of ischaemic events
for meta-analysis of continuous data assume that the outcomes and wound complications (Lazar 2004). A study by Vriesendorp
have a normal distribution in each treatment arm in each study. et al, on people undergoing surgery for oesophageal cancer, re-
This assumption may not always be met, although this is less im- ported an association between postoperative glucose levels and an
portant in large studies. increased length of hospital stay, but not an increase in infectious
For the final neurological deficit, we used the SMD approach complications (Vriesendorp 2004).
instead of the MD because the scales were different (NIHSS and Our results are in keeping with other more recent trials of tight
ESS). In this case, the standard deviation was used to standardise glycaemic control. In a study of 400 people after cardiac surgery,
the MDs to a single scale as well as in the computation of study there was an increased incidence of death and stroke in the in-
weights. It is assumed that a variation between standard deviations tensive insulin treatment group (Gandhi 2007). In a later study,
reflects only differences in measurement scales and not differences Van den Berghe et al studied 1200 people and showed that there
in the reliability of the outcome measures or variability among trial was no change in mortality among people in medical intensive
populations. These limitations of the methods should be borne in care units (ICU) (Van den Berghe 2006). Brunkhorst et al stud-
mind where an unexpected variation of standard deviations across ied 537 people with sepsis and found no significant difference be-
studies is observed. tween the intensive and conventional therapy groups in the rate
Sensitivity analyses test the robustness of the results relative to fea- of death or mean score for organ failure (Brunkhorst 2008). The
tures of the primary studies and to key assumptions and decisions CREATE-ECLA 2005 study compared glucose-insulin-potassium
(Cook 1995), and they test for bias due to the retrospective nature infusion with usual care in 20,000 people presenting with ST-seg-
of systematic reviews. Sensitivity analyses in meta-analysis allow ment elevation myocardial infarction and found a lack of benefit
more appropriate and reliable conclusions when problems such on mortality, cardiac arrest, and cardiogenic shock. The multi-
as unavailable estimates are present (Riley 2004). We were able centre Normoglycaemia in Intensive Care Evaluation and Survival
to perform only a minor percentage of the previously proposed Using Glucose Algorithm Regulation (NICE-SUGAR) study had
sensitivity analyses due to the fact that, for the primary outcome, a mixed ICU population with 6104 people. This study provided
some studies did not report enough data. evidence that intensive insulin treatment was associated with in-
Reviewing research for systematic reviews implies a retrospective creased mortality and increased hypoglycaemia without significant
nature and therefore is subject to random and systematic error differences in critical care unit or hospital length of stay, number
(Montori 2003). In order to decrease this error, we wrote a proto- of days of mechanical ventilation, or need for renal replacement
col before the search was made and two review authors worked in- therapy. In a meta-analysis of 29 RCTs involving 8432 partici-
dependently in the extraction of the information, with substantial pants (Wiener 2008) there was no difference in hospital mortality
agreement in the decisions on which studies should be included. between tight and usual glucose control (21.6% versus 23.3%).
Publication bias could affect our review despite our extensive
search procedure. More than 10 studies (ideally more than 30
studies) should be included in order to provide a funnel plot (Lau
2006). We did not contact pharmaceutical companies manufac- AUTHORS’ CONCLUSIONS
turing insulin. Our meta-analysis included only RCTs and we
found only one unpublished study in the previous version of the Implications for practice
review that was subsequently published in 2011. We contacted
Evidence from this systematic review indicates that, compared
the authors for clarification in regards to study information when
with control intervention, the administration of insulin with the
appropriate.
aim of maintaining glucose within a specific range immediately af-
ter acute ischaemic stroke does not reduce dependency, neurolog-
ical deficit, or mortality at 30 days or 90 days. These findings did
Agreements and disagreements with other not change in the subgroup analysis of those with diabetes mel-
studies or reviews litus compared with those without diabetes mellitus. There was,
Clinical trials that investigated the role of intensive glycaemic con- however, a significant increase in episodes of hypoglycaemia in
trol on critically ill people have produced controversial results. the treatment group. This review provides no evidence to support

the use of intensive insulin therapy for tight glucose control after that tight glycaemic control improves functional outcome or re-
acute stroke. The number needed to treat is not significant for the duces mortality in people with acute ischaemic stroke. At this time,
outcome of death or the composite outcome of dependency and the benefits of maintaining tight glucose control do not outweigh
death, and the number needed to harm is nine for symptomatic the potential risks of hypoglycaemia to the already vulnerable is-
hypoglycaemia. chaemic penumbra in ischaemic stroke patients.
Implications for research

ACKNOWLEDGEMENTS
In animal studies and human studies of acute stroke, hypergly-
caemia has been shown to have a deleterious effect on the critical To Hazel Fraser and the Cochrane Stroke Group for their contin-
area of ischaemic penumbra. This review and other similar studies uous support, and to Daniel Cabrera for his support in selecting
of glucose control have been unable to provide reliable evidence articles and preparing the manuscript.
REFERENCES
References to studies included in this review glycemic control in nondiabetic acute stroke patients with
mild hyperglycemia. Journal of Stroke and Cerebrovascular
Azevedo 2009 {published data only (unpublished sought but not used)} Diseases 2011;20(2):150–4. [PUBMED: PMID:
Azevedo JRA, Azevedo RP, Miranda MA, Costa NNR, 20621520]
Araujo LO. Management of hyperglycemia in patients with
THIS 2008 {published and unpublished data}
acute ischemic stroke: comparison of two strategies. Critical
Bruno A, Kent TA, Coull BM, Shankar RR, Saha C, Becker
Care 2009;13 Suppl 3:48.
KJ, et al. Treatment of hyperglycemia in ischemic stroke
GIST-UK 2007 {published and unpublished data} (THIS): a randomized pilot trial. Stroke 2008;39:384–9.
Gray CS, Hildreth AJ, Sandercock PA, O’Connell JE,
Vinychuk 2005 {published data only}
Johnston DE, Cartlidge NE, et al. Glucose-potassium-
Vinychuk S, Melnyk V, Margitich V. Hyperglycemia
insulin infusions in the management of post-stroke
after acute ischemic stroke: prediction, significance
hyperglycaemia: the UK Glucose Insulin in Stroke Trial
and immediate control with insulin-potassium-saline-
(GIST-UK). Lancet Neurology 2007;6:397–406.
magnesium infusions. Heart Drug 2005; 5:197–204.
GRASP 2009 {published data only}
Vriesendorp 2009 {published data only}
Johnston KC, Hall CE, Kissela BM, Bleck TP, Conaway
Vriesendorp TM, Roos YB, Kruyt ND, Biessels GJ,
MR, GRASP Investigators. Glucose Regulation in Acute
Kappelle LJ, Vermeulen M, et al. Efficacy and safety of two
Stroke Patients (GRASP) trial: a randomized pilot trial.
5 day insulin dosing regimens to achieve strict glycaemic
Stroke 2009;40:3804–9.
control in patients with acute ischaemic stroke. Journal of
INSULINFARCT 2012 {published data only} Neurology, Neurosurgery, and Psychiatry 2009;80(9):1040–3.
Rosso C, Corvol JC, Pires C, Crozier S, Attal Y, [PUBMED: PMID: 19684236]
Jacqueminet S, et al. Intensive versus subcutaneous Walters 2006 {published data only}
insulin in patients with hyperacute stroke: results from Walters MR, Weir CJ, Lees KR. A randomised, controlled
the randomized INSULINFARCT trial. Stroke 2012;43: pilot study to investigate the potential benefit of intervention
2343–9. [NCT00472381; PUBMED: PMID: 22700528] with insulin in hyperglycaemic acute ischaemic stroke
Kreisel 2009 {published data only} patients. Cerebrovascular Diseases 2006;22:116–22.
Kreisel SH, Berschin UM, Hammes HP, Leweling H,
Bertsch T, Hennerici MG, et al. Pragmatic management References to studies excluded from this review
of hyperglycaemia in acute ischaemic stroke: safety and CIMT Trial {published data only}
feasibility of intensive intravenous insulin treatment. Lundby Christensen L, Almdal T, Boesgaard T, Breum L,
Cerebrovascular Diseases 2009;27(2):167–75. Dunn E, Gade-Rasmussen B, et al. Study rationale and
McCormick 2010 {published data only} design of the CIMT trial: the Copenhagen Insulin and
McCormick M, Hadley D, McLean JR, Macfarlane Metformin Therapy trial. Diabetes, Obesity & Metabolism
JA, Condon B, Muir KW. Randomized controlled trial 2009;11:315–22.
of insulin for acute poststroke hyperglycemia. Annals of De Azevedo 1997 {published and unpublished data}
Neurology 2010;67:570–8. [PUBMED: PMID: 20437554] De Azevedo JR, Silva LGS. Prognosis of neurological
Staszewski 2011 {published and unpublished data} and neurosurgical patients submitted to prevention of
Staszewski J, Brodacki B, Kotowicz J, Stepien A. hyperglycemia. Revista Brasileira de Terapia Intensiva 1997;
Intravenous insulin therapy in the maintenance of strict 9(2):77–81.
GIST 1999 {published data only} Barry 1993
Scott JF, Robinson GM, French JM, O’Connell JE, Alberti Barry MA, Reynolds JE, Eastman A. Etoposide-induced
KG, Gray CS. Glucose potassium insulin infusions in the apoptosis in human HL-60 cells is associated with
treatment of acute stroke patients with mild to moderate intracellular acidification. Cancer Research 1993;53:
hyperglycemia: the Glucose Insulin in Stroke Trial (GIST). 2349–57.
Stroke 1999;30(4):793–9. Bochicchio 2008
GLUCOVAS {published data only} Bochicchio GV, Scalea TM. Glycemic control in the ICU.
Kruyt N. Glucose regulation by continuous tube feeding Advances in Surgery 2008;42:261–75.
and Vildagliptin in addition to insulin in hyperglycemic Brunkhorst 2008
acute stroke patients. http://www.trialregister.nl/trialreg/ Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A,
admin/rctview.asp?TC=1874. [Nederlands Trial Register Ragaller M, Weiler N, et al. Intensive insulin therapy and
(http://www.trialregister.nl). 2009.] pentastarch resuscitation in severe sepsis. New England
Green 2010 {published data only} Journal of Medicine 2008;358(2):125–39.
Green DM, O’Phelan KH, Bassin SL, Change CWJ, Stern Bruno 2002
TS, Asai SM. Intensive versus conventional insulin therapy Bruno A, Levine SR, Frankel MR, Brott TG, Lin Y, Tilley
in critically ill neurologic patients. Neurocritical Care 2010; BC, et al. Admission glucose level and clinical outcomes in
13:299–306. the NINDS rt-PA Stroke Trial. Neurology 2002;59(5):669.
Miyashita 2008 {published data only} [PUBMED: 12221155]
Miyashita Y, Nishimura R, Nemoto M, Matsudaira T, Bruno 2004
Kurata H, Yokota T, et al. Prospective randomized study Bruno A, Williams LS, Kent TA. How important is
for optimal insulin therapy in type 2 diabetic patients with hyperglycemia during acute brain infarction?. Neurologist
secondary failure. Cardiovascular Diabetology 2008;7:16. 2004;10(4):195–200.
VADT {published data only} Candelise 1985
Duckworth W, Abraira C, Moritz T, Reda D, Emanuele Candelise L, Landi G, Orazio EN, Boccardi E. Prognostic
N, Reaven PD, et al. Glucose control and vascular significance of hyperglycemia in acute stroke. Archives of
complications in veterans with type 2 diabetes. New Neurology 1985;42:661–3.
England Journal of Medicine 2009;360:129–39. Capes 2001
Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC.
References to ongoing studies Stress hyperglycemia and prognosis of stroke in nondiabetic
and diabetic patients: a systematic overview. Stroke 2001;
NCT00373269 {published data only} 32:2426–32.
Gentile NT. Effect of insulin on infarct size and neurologic
Combs 1992
outcome after acute stroke. http://clinicaltrials.gov/show/
Combs DJ, Dempsey RJ, Donaldson D, Kindy MS.
NCT00373269.
Hyperglycemia suppresses c-fos mRNA expression following
SHINE {published data only} transient cerebral ischemia in gerbils. Journal of Cerebral
Johnston KC. Stroke Hyperglycemia Insulin Network Effort Blood Flow and Metabolism 1992;12:169–72.
(SHINE). ClinicalTrials.gov. [electronic database] 2011. Cook 1995
Cook DJ, Sackett DL, Spitzer WO. Methodologic
Additional references
guidelines for systematic reviews of randomized control
trials in health care from the Potsdam Consultation on
Altman 2003
Meta-Analysis. Journal of Clinical Epidemiology 1995;48:
Altman DG, Bland JM. Interaction revisited: the difference
167–71.
between two estimates. BMJ 2003;326:219.
CREATE-ECLA 2005
Anderson 1999 Mehta SR, Yusuf S, Díaz R, Zhu J, Pais P, Xavier D, et al.
Anderson RE, Tan WK, Martin HS, Meyer FB. Effects Effect of glucose-insulin-potassium infusion on mortality
of glucose and PaO2 modulation on cortical intracellular in patients with acute ST-segment elevation myocardial
acidosis, NADH redox state, and infarction in the ischemic infarction: the CREATE-ECLA randomized controlled
penumbra. Stroke 1999;30(1):160–70. trial. JAMA 2005;293(4):437–46.
Auer 1998 de Courten-Myers 1989
Auer RN. Insulin, blood glucose levels, and ischemic brain de Courten-Myers GM, Kleinholz M, Wagner KR, Myers
damage. Neurology 1998;51(3 Suppl 3):39–43. RE. Fatal strokes in hyperglycemic cats. Stroke 1989;20:
Barber 1967 1707–15.
Barber AA, Bernheim F. Lipid peroxidation: its DerSimonian 1986
measurement, occurrence, and significance in animal tissues. DerSimonian R, Laird N. Meta-analysis in clinical trials.
Advances in Gerontological Research 1967;2:355–403. Controlled Clinical Trials 1986;7:177–88.
Farese 1991 Koistinaho 1999
Farese RV Jr, Yost TJ, Eckel RH. Tissue-specific regulation Koistinaho J, Pasonen S, Yrjanheikki J, Chan PH. Spreading
of lipoprotein lipase activity by insulin/glucose in normal- depression-induced gene expression is regulated by plasma
weight humans. Metabolism 1991;40(2):214. [PUBMED: glucose. Stroke 1999;30(1):114–9.
1988780] Langhorne 1993
Fielding 1998 Langhorne P, Williams BO, Gilchrist W, Howie K. Do
Fielding BA, Frayn KN. Lipoprotein lipase and the stroke units save lives?. Lancet 1993;342:395–8.
disposition of dietary fatty acids. British Journal of Nutrition Lau 2006
1998;80(6):495. [PUBMED: 10211047] Lau J, Ioannidis JP, Terrin N, Schmid CH, Olkin I. The case
Gandhi 2007 of the misleading funnel plot. BMJ 2006;333:597–600.
Gandhi GY, Nuttall GA, Abel MD, Mullany CJ, Schaff HV, Lazar 2004
O’Brien PC, et al. Intensive intraoperative insulin therapy Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H,
versus conventional glucose management during cardiac Apstein CS. Tight glycemic control in diabetic coronary
surgery: a randomized trial. Annals of Internal Medicine artery bypass graft patients improves perioperative outcomes
2007;146(4):233–43. and decreases recurrent ischemic events. Circulation 2004;
Gisselsson 1999 109(12):1497–502.
Gisselsson L, Smith ML, Siesjo BK. Hyperglycemia and Li 2000
focal brain ischemia. Journal of Cerebral Blood Flow and Li PA, Shuaib A, Miyashita H, He QP, Siesjo BK, Warner
Metabolism 1999;19:288–97. DS. Hyperglycemia enhances extracellular glutamate
Gray 1987 accumulation in rats subjected to forebrain ischemia. Stroke
Gray CS, Taylor R, French JM, Alberti KG, Venables 2000;31(1):183–92.
GS, James OF, et al. The prognostic value of stress Lin 1998
hyperglycaemia and previously unrecognized diabetes in Lin B, Ginsberg MD, Busto R. Hyperglycemic exacerbation
acute stroke. Diabetic Medicine 1987;4:237–40. of neuronal damage following forebrain ischemia:
Hamilton 1995 microglial, astrocytic and endothelial alterations. Acta
Hamilton MG, Tranmer BI, Auer RN. Insulin reduction of Neuropathologica 1998;96:610–20.
cerebral infarction due to transient focal ischemia. Journal Lo 2003
of Neurosurgery 1995;82(2):262–8. Lo EH, Dalkara T, Moskowitz MA. Mechanisms, challenges
Higgins 2011 and opportunities in stroke. Nature Reviews Neuroscience
Higgins JPT, Green S (editors). Cochrane Handbook 2003;4(5):399–415.
for Systematic Reviews of Interventions Version 5.1.0 Melamed 1976
[updated March 2011]. The Cochrane Collaboration, Melamed E. Reactive hyperglycaemia in patients with acute
2011. Available from www.cochrane-handbook.org. stroke. Journal of the Neurological Sciences 1976;29:267–75.
Hoxworth 1999 Montori 2003
Hoxworth JM, Xu K, Zhou Y, Lust WD, LaManna JC. Montori VM, Swiontkowski MF, Cook DJ. Methodologic
Cerebral metabolic profile, selective neuron loss, and issues in systematic reviews and meta-analyses. Clinical
survival of acute and chronic hyperglycemic rats following Orthopaedics and Related Research 2003;413:43–54.
cardiac arrest and resuscitation. Brain Research 1999;821 Murros 1992
(2):467–79. Murros K, Fogelholm R, Kettunen S, Vuorela AL, Valve J.
Ingels 2006 Blood glucose, glycosylated haemoglobin, and outcome
Ingels C, Debaveye Y, Milants I, Buelens E, Peeraer A, of ischemic brain infarction. Journal of the Neurological
Devriendt Y, et al. Strict blood glucose control with insulin Sciences 1992;111:59–64.
during intensive care after cardiac surgery: impact on 4- Murros 1993
years survival, dependency on medical care, and quality-of- Murros K, Fogelholmd R. Diabetes and stress
life. European Heart Journal 2006;27(22):2716–24. hyperglycaemia in stroke. Journal of Neurology, Neurosurgery,
Jorgensen 1994 and Psychiatry 1993;56:428.
Jorgensen H, Nakayama H, Raaschou HO, Olsen TS. Myers 1977
Stroke in patients with diabetes. The Copenhagen Stroke Myers RE, Yamaguchi S. Nervous system effects of cardiac
Study. Stroke 1994;25:1977–84. arrest in monkeys. Preservation of vision. Archives of
Kiers 1992 Neurology 1977;34:65–74.
Kiers L, Davis SM, Larkins R, Hopper J, Tress B, Rossiter Nedergaard 1996
SC, et al. Stroke topography and outcome in relation Nedergaard M. Spreading depression as a contributor to
to hyperglycaemia and diabetes. Journal of Neurology, ischemic brain damage. Advances in Neurology 1996;71:
Neurosurgery, and Psychiatry 1992;55:263–70. 75–83.

NICE-SUGAR vitro. Journal of Cerebral Blood Flow and Metabolism 1985;
The NICE-SUGAR Study Investigators. Intensive versus 5:253–8.
conventional glucose control in critically ill patients. New Siesjo 1996
England Journal of Medicine 2009;360(13):1283–97. Siesjo BK, Katsura K, Kristian T. Acidosis-related damage.
[ClinicalTrials identifier NCT00220987] Advances in Neurology 1996;71:209–33.
O’Neill 1991 Stead 2009
O’Neill PA, Davies I, Fullerton KJ, Bennett D. Stress Stead LG, Gilmore RM, Bellolio MF, Mishra S, Bhagra A,
hormone and blood glucose response following acute stroke Vaidyanathan L, et al. Hyperglycemia as an independent
in the elderly. Stroke 1991;22:842–7. predictor of worse outcome in non-diabetic patients
Parsons 2002 presenting with acute ischemic stroke. Neurocritical Care
Parsons MW, Barber PA, Desmond PM, Baird TA, Darby 2009;10(2):181–6.
DG, Byrnes G, et al. Acute hyperglycemia adversely Uyttenboogaart 2007
affects stroke outcome: a magnetic resonance imaging and Uyttenboogaart M, Luijckx GJ, Vroomen PC, Stewart RE,
spectroscopy study. Annals of Neurology 2002;52(1):20–8. De Keyser J. Measuring disability in stroke: relationship
Prakash 2008 between the modified Rankin scale and the Barthel index.
Prakash A, Matta BF. Hyperglycaemia and neurological Journal of Neurology 2007;254(8):1113–7.
injury. Current Opinion in Anaesthesiology 2008;21:565–9. Van den Berghe 2001
Pulsinelli 1982 Van den Berghe G, Wouters P, Weekers F, Verwaest C,
Pulsinelli WA, Waldman S, Rawlinson D, Plum F. Bruyninckx F, Schetz M, et al. Intensive insulin therapy in
Moderate hyperglycemia augments ischemic brain damage: the critically ill patients. New England Journal of Medicine
a neuropathologic study in the rat. Neurology 1982;32: 2001;345(19):1359–67.
1239–46. Van den Berghe 2006
Ramnanan 2010 Van den Berghe G, Wilmer A, Hermans G, Meersseman W,
Ramnanan CJ, Edgerton DS, Rivera N, Irimia-Dominguez Wouters PJ, Milants I, et al. Intensive insulin therapy in the
J, Farmer B, Neal DW, et al. Molecular characterization of medical ICU. New England Journal of Medicine 2006;354
insulin-mediated suppression of hepatic glucose production (5):449–61.
in vivo. Diabetes 2010;59(6):1302. [PUBMED: 20185816] Van Kooten 1993
Regli 1996 Van Kooten F, Hoogerbrugge N, Naarding P, Koudstaal PJ.
Regli L, Held MC, Anderson RE, Meyer FB. Nitric oxide Hyperglycemia in the acute phase of stroke is not caused by
synthase inhibition by L-NAME prevents brain acidosis stress. Stroke 1993;24:1129–32.
during focal cerebral ischemia in rabbits. Journal of Cerebral Voll 1989
Blood Flow and Metabolism 1996;16:988–95. Voll CL, Whishaw IQ, Auer RN. Postischemic insulin
Rehncrona 1981 reduces spatial learning deficit following transient forebrain
Rehncrona S, Rosen I, Siesjo BK. Brain lactic acidosis and ischemia in rats. Stroke 1989;20(5):646–51.
ischemic cell damage: 1. Biochemistry and neurophysiology. Voll 1991
Journal of Cerebral Blood Flow and Metabolism 1981;1: Voll CL, Auer RN. Insulin attenuates ischemic brain
297–311. damage independent of its hypoglycemic effect. Journal of
RevMan 2012 [Computer program] Cerebral Blood Flow and Metabolism 1991;11(6):1006–14.
The Nordic Cochrane Centre, The Cochrane Collaboration. Vriesendorp 2004
Review Manager (RevMan). Version 5.2. Copenhagen: Vriesendorp TM, DeVries JH, Hulscher JB, Holleman
The Nordic Cochrane Centre, The Cochrane Collaboration, F, Van Lanschot JJ, Hoekstra JB. Early postoperative
2012. hyperglycaemia is not a risk factor for infectious
Riley 2004 complications and prolonged in-hospital stay in patients
Riley RD, Sutton AJ, Abrams KR, Lambert PC. Sensitivity undergoing oesophagectomy: a retrospective analysis of a
analyses allowed more appropriate and reliable meta-analysis prospective trial. Critical Care 2004;8(6):R437–42.
conclusions for multiple outcomes when missing data was Wass 1996
present. Journal of Clinical Epidemiology 2004;57:911–24. Wass CT, Scheithauer BW, Bronk JT, Wilson RM,
Scott 1999 Lanier WL. Insulin treatment of corticosteroid-associated
Scott JF, Robinson GM, French JM, O’Connell JE, Alberti hyperglycemia and its effect on outcome after forebrain
KG, Gray CS. Prevalence of admission hyperglycaemia ischemia in rats. Anesthesiology 1996;84(3):644–51.
across clinical subtypes of acute stroke. Lancet 1999;353: Weir 1997
376–7. Weir CJ, Murray GD, Dyker AG, Lees KR. Is
Siesjo 1985 hyperglycaemia an independent predictor of poor outcome
Siesjo BK, Bendek G, Koide T, Westerberg E, Wieloch T. after acute stroke? Results of a long-term follow up study.
Influence of acidosis on lipid peroxidation in brain tissues in BMJ 1997;314:1303–6.

Wiener 2008 Bellolio 2008
Wiener RS, Wiener DC, Larson RJ. Benefits and risks of Bellolio MF, Gilmore RM, Stead LG. Interventions for
tight glucose control in critically ill adults: a meta-analysis. controlling hyperglycaemia in acute ischaemic stroke.
JAMA 2008;300(8):933–44. Cochrane Database of Systematic Reviews 2008, Issue 2.
Woo 1988 DOI: 10.1002/14651858.CD005346.pub2
Woo E, Ma JT, Robinson JD, Yu YL. Hyperglycemia is a
stress response in acute stroke. Stroke 1988;19:1359–64.
Bellolio 2011
Zhu 1994
Bellolio MF, Gilmore RM, Stead LG. Insulin for
Zhu CZ, Auer RN. Intraventricular administration of
glycaemic control in acute ischaemic stroke. Cochrane
insulin and IGF-1 in transient forebrain ischemia. Journal
Database of Systematic Reviews 2011, Issue 9. DOI:
of Cerebral Blood Flow and Metabolism 1994;14(2):237–42.
10.1002/14651858.CD005346.pub3; PUBMED: PMID:
References to other published versions of this review 21901697
∗
Indicates the major publication for the study

CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Azevedo 2009
Methods Randomised, parallel design
Participants 34 consecutive acute ischaemic strokes; 14 randomised to intensive insulin therapy and
20 control carbohydrate
Interventions Continous intravenous insulin infusion

Control: carbohydrate restrictive strategy, intravenous infusion glucose-free and enteral
nutrition 33.3% carbohydrates, regular insulin SQ Goal < 150 mg/dL
Outcomes Follow-up until discharge

Outcomes: glucose level, NIHSS, death, hypoglycaemia
Notes Abstract format only

Author contacted and provided extra information
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection Low risk Computer-generated random-number table and sealed
bias) envelopes
Allocation concealment (selection bias) Unclear risk Not reported
Blinding of participants and personnel High risk Not blinded

(performance bias)
All outcomes
Blinding of outcome assessment (detection High risk Not blinded

bias)
All outcomes
Incomplete outcome data (attrition bias) Unclear risk Not reported

All outcomes
GIST-UK 2007
Participants 933 participants: 464 in the intervention group and 469 in the control group

GIST-UK 2007 (Continued)
Interventions Glucose-insulin-potassium (10% dextrose, 20 mmol KCl, 16 units insulin); continuous

intravenous infusion for 24 hours to maintain capillary glycaemia 4 to 7 mmol/L, plasma
glucose 4.6 to 8 mmol/L, measured every 8 hours
Outcomes At 90 days
ESS and Rankin Scale
Notes 71/464 (15.3%) non-ischaemic strokes in the intervention group, 78/469 (16.6%) non-
ischaemic strokes in the control group; stopped early
Risk of bias
Random sequence generation (selection Low risk Adequate

bias)
Allocation concealment (selection bias) Low risk Adequate

(performance bias)
All outcomes
Blinding of outcome assessment (detection Low risk Blinded

bias)
All outcomes
Incomplete outcome data (attrition bias) High risk 7.4% lost to follow-up and early termination
All outcomes
GRASP 2009
Methods Randomised, parallel design, stratified by glucose concentration
Participants 74 participants (3 arms): 24 participants in the intervention group (tight control), 50

participants in the control group
Interventions Intravenous insulin infusion in normal saline (plus glucose 5% and 20 mEq/L potassium
infusion) and subcutaneous insulin with each meal; continuous intravenous insulin in-
fusion individually adjusted; target 70 to 110 mg/dL, or 3.9 to 6.1 mmol/L in the tight
control group, with capillary glucose check every 1 to 4 hours for 5 days or discharge
Outcomes At 90 days
NIHSS and Rankin
Notes
Risk of bias

GRASP 2009 (Continued)

bias)

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) Low risk 1.4% lost to follow-up
All outcomes
INSULINFARCT 2012
Methods Randomised, parallel design, not blinded
Participants 180 enrolled: 90 each arm, ITT and 85 each arm per protocol analysis
Interventions Intravneous continuous infusion of Actrapid insulin with 1 hourly glucose check and
dose adjustment; aim to get glucose < 7; continued for 24 hours
Control subcutaneous insulin, glucose check every 4 hours x 24 hours; stop point of 8
mmol/L at which no insulin given
Outcomes At 90 days
Glucose level, NIHSS, Rankin, death , hypoglycaemia, infarct volume by MRI
Notes
Risk of bias

bias)

(performance bias)
All outcomes

INSULINFARCT 2012 (Continued)

bias)
All outcomes
Incomplete outcome data (attrition bias) Low risk Adequate

All outcomes
Kreisel 2009
Participants 40 participants: 20 in each arm
Interventions Continuous intravenous insulin infusion individually adjusted; target 80 to 110 mg/
dL, or 4.44 to 6.1 mmol/L with capillary glucose check every 1 to 4 hours for 5 days
Outcomes 120 days

Rankin Scale
Notes Main objective of the study was feasibility of insulin for control of glycaemia and hypo-
glycaemic events
Risk of bias

bias)
Allocation concealment (selection bias) Unclear risk Not enough information

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) Low risk 10% lost to follow-up (4 of 40 lost to clinical follow-up,
All outcomes 1 of 40 (2.5%) lost to survival follow-up)

McCormick 2010
Methods Randomised, parallel design, placebo controlled
Participants 40 participants: 25 intervention, 15 control
Interventions Glucose-insulin-potassium
Continuous intravenous infusion 100ml/hour x 24 hours (10 participants), 48 hours
(5 participants) or 72 hours (10 participants); dose adjusted to keep capillary blood
glucose between 4 to 7 mmol/L; glucose checked 1 hourly until euglycaemia achieved
and subsequently 2 hourly
Outcomes Hypoglycaemia, Rankin Scale, death
Notes 3 different durations of treatment
Risk of bias

bias)
Allocation concealment (selection bias) Unclear risk Not enough information

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) Unclear risk Not reported

All outcomes
Staszewski 2011
Participants 50 participants: 26 intervention and 24 controls
Interventions Continuous intravenous insulin infusion adjusted to maintain euglycaemia (4.5 to 7

mmol/L); capillary glucose every 1 hour initially (every 4 hours once the participant was
stable) for 24 hours
Outcomes At 30 days
Glucose level, NIHSS, Rankin, death, hypoglycaemia
Notes Same study used in 2007 with unpublished results. Now has been published

Staszewski 2011 (Continued)
Risk of bias

bias)
Allocation concealment (selection bias) High risk Inadequate (random list was read by investigator)

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) Low risk None lost to follow-up
All outcomes
THIS 2008
Participants 46 participants: 31 in the intervention group, 15 in the control group
Interventions Continuous intravenous insulin infusion individually adjusted; target 5 to 7.2 mmol/L;
capillary glucose every 1 hour for 72 hours
Outcomes At 90 days
NIHSS and Barthel
Notes Almost all had diabetes mellitus (100% of intervention and 73% of control participants)
Risk of bias

bias)
Blinding of participants and personnel High risk Participants were blinded, clinicians and data collectors
(performance bias) were not blinded
All outcomes

THIS 2008 (Continued)

bias)
All outcomes
All outcomes
Vinychuk 2005
Participants 128 participants: 61 intervention, 67 control
Interventions Continuous intravenous infusion 100 ml/hour per 4 hours with insulin doses adjusted
by glucose level, until desired glucose level reached (< 7 mmol/L); measured every 4
hours
Outcomes At 30 days
NIHSS and Barthel
Notes Diabetes mellitus and non-diabetes mellitus cohorts reported separately
Risk of bias

bias)

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) High risk Reported none lost to follow-up; however, estimated to
All outcomes be at least 2.3% based on the percentages of the figures
at follow-up

Vriesendorp 2009
Participants 33 participants: 13 basal insulin, 10 meal-related insulin, 10 controls
Interventions Stopped oral glucose lowering drugs

Basal group: intravenous infusion insulin adjusted every hour until < 6.1 mmol/l and
bolus after meals
Meal group: SQ long acting as basal and SQ rapid acting as meal related insulin, measured
before and 2 hours after meals, and twice at night to detect hypoglycaemia Target: 4.4
to 6.1 mmol/l
Outcomes At 5 days
NIHSS and hypoglycaemia
Notes
Risk of bias
Random sequence generation (selection High risk Consecutive envelopes

bias)
Allocation concealment (selection bias) High risk Consecutive envelopes stratified for dysphagia and dia-
betes mellitus

(performance bias)
All outcomes

bias)
All outcomes
Incomplete outcome data (attrition bias) High risk 15% without outcome documented at 5 days
All outcomes
Walters 2006
Participants 25 participants: 13 in the intervention group, 12 in the control group
Interventions Continuous intravenous insulin infusion for 48 hours; target 5 to 7.9 mmol/L; monitored
every 2 hours
Outcomes 30 days

Walters 2006 (Continued)
Notes Pilot study looking at feasibility
Risk of bias

bias)

(performance bias)
All outcomes

bias)
All outcomes
All outcomes
ESS: European Stroke Scale

ITT: intention-to-treat
MRI: magnetic resonance imaging
NIHSS: National Institutes of Health Stroke Scale
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
CIMT Trial Not stroke participants; stroke is an outcome
De Azevedo 1997 Intervention is diet; both groups received the same regimen of insulin
GIST 1999 Overlapping cohort with GIST-UK 2007
GLUCOVAS Not randomised clinical trial design, these are cohorts with intervention
Green 2010 Only 3.7% of the cohort is ischaemic stroke
Miyashita 2008 Not stroke participants; stroke is an outcome

(Continued)
VADT Not stroke participants; stroke is an outcome
Characteristics of ongoing studies [ordered by study ID]
NCT00373269
Trial name or title Effect of insulin on infarct size and neurologic outcome after acute stroke
Methods RCT
Participants Adults with acute stroke
Interventions Insulin
Outcomes Primary outcome is change in infarct volume measured on diffusion-perfusion MRI
Starting date January 2004 to January 2010
Contact information Nina T Gentile, MD
Notes clinicaltrials.gov NCT00373269
SHINE
Trial name or title Stroke Hyperglycemia Insulin Network Effort (SHINE)
Methods RCT
Participants Acute ischaemic stroke < 12 hours from onset
Interventions Intravenous insulin to maintain target glucose concentration of 80 to 130 mg/dL
Outcomes Modified Rankin Scale score at 3 months

Hypoglycaemia
Starting date April 2012
Contact information Karen C Johnston, MD
Notes ClinicalTrials.gov NCT01369069
MRI: magnetic resonance imaging

RCT: randomised controlled trial

DATA AND ANALYSES
Comparison 1. Dependency or death
No. of No. of
Outcome or subgroup title studies participants Statistical method Effect size
1 Dependency or death at the end 9 1516 Odds Ratio (M-H, Fixed, 95% CI) 0.99 [0.79, 1.23]
of the follow-up
2 Death 9 1422 Odds Ratio (M-H, Fixed, 95% CI) 1.09 [0.85, 1.41]
3 Diabetes mellitus versus no 8 1482 Odds Ratio (M-H, Fixed, 95% CI) 0.97 [0.77, 1.21]
diabetes mellitus
3.1 Diabetes mellitus 3 194 Odds Ratio (M-H, Fixed, 95% CI) 0.66 [0.35, 1.24]
3.2 No diabetes mellitus 6 1288 Odds Ratio (M-H, Fixed, 95% CI) 1.02 [0.81, 1.30]
4 Less than 30 days versus 90 days 9 1516 Odds Ratio (M-H, Fixed, 95% CI) 0.99 [0.79, 1.23]
of follow-up
4.1 30 days 5 289 Odds Ratio (M-H, Fixed, 95% CI) 0.74 [0.43, 1.25]
4.2 90 days 4 1227 Odds Ratio (M-H, Fixed, 95% CI) 1.05 [0.82, 1.34]
Comparison 2. Functional neurological outcome
No. of No. of
1 NIHSS or ESS at the end of the 8 1432 Std. Mean Difference (IV, Fixed, 95% CI) -0.09 [-0.19, 0.01]
follow-up
2 Independent in daily activities 9 1224 Odds Ratio (M-H, Fixed, 95% CI) 1.03 [0.81, 1.32]
3 Diabetes mellitus versus no 8 1432 Std. Mean Difference (IV, Fixed, 95% CI) -0.07 [-0.18, 0.03]
diabetes mellitus
3.1 Diabetes mellitus 3 146 Std. Mean Difference (IV, Fixed, 95% CI) -0.06 [-0.43, 0.31]
3.2 No diabetes mellitus 6 1286 Std. Mean Difference (IV, Fixed, 95% CI) -0.08 [-0.19, 0.03]
4 Less than 30 days versus 90 days 8 1432 Std. Mean Difference (IV, Fixed, 95% CI) -0.09 [-0.19, 0.01]
of follow-up
4.1 30 days 5 273 Std. Mean Difference (IV, Fixed, 95% CI) -0.47 [-0.72, -0.23]
4.2 90 days 3 1159 Std. Mean Difference (IV, Fixed, 95% CI) -0.00 [-0.12, 0.11]

Comparison 3. Hypoglycaemia
No. of No. of
1 Symptomatic hypoglycaemia 10 1455 Odds Ratio (M-H, Fixed, 95% CI) 14.60 [6.62, 32.21]
2 Hypoglycaemia (with or without 10 1455 Odds Ratio (M-H, Fixed, 95% CI) 18.41 [9.09, 37.27]
symptoms)
Comparison 4. Mean glucose level
No. of No. of
1 Mean glucose level 8 1398 Mean Difference (IV, Fixed, 95% CI) -0.63 [-0.80, -0.46]
Analysis 1.1. Comparison 1 Dependency or death, Outcome 1 Dependency or death at the end of the
follow-up.
Review: Insulin for glycaemic control in acute ischaemic stroke
Comparison: 1 Dependency or death
Outcome: 1 Dependency or death at the end of the follow-up
Study or subgroup Experimental Control Odds Ratio Weight Odds Ratio

n/N n/N M-H,Fixed,95% CI M-H,Fixed,95% CI
Azevedo 2009 13/14 15/20 0.6 % 4.33 [ 0.45, 42.02 ]
GIST-UK 2007 336/461 330/467 56.5 % 1.12 [ 0.84, 1.49 ]
GRASP 2009 14/24 34/49 5.9 % 0.62 [ 0.22, 1.70 ]
INSULINFARCT 2012 49/90 49/90 14.2 % 1.00 [ 0.56, 1.80 ]
Kreisel 2009 15/20 15/20 2.4 % 1.00 [ 0.24, 4.18 ]
McCormick 2010 21/25 13/15 1.7 % 0.81 [ 0.13, 5.05 ]
Staszewski 2011 14/26 17/24 5.2 % 0.48 [ 0.15, 1.55 ]
THIS 2008 16/31 8/15 3.3 % 0.93 [ 0.27, 3.21 ]
Vinychuk 2005 14/60 22/65 10.3 % 0.59 [ 0.27, 1.31 ]
Total (95% CI) 751 765 100.0 % 0.99 [ 0.79, 1.23 ]
0.05 0.2 1 5 20
Favours experimental Favours control
(Continued . . . )

(. . . Continued)
Total events: 492 (Experimental), 503 (Control)
Heterogeneity: Chi2 = 6.25, df = 8 (P = 0.62); I2 =0.0%
Test for overall effect: Z = 0.11 (P = 0.91)
Test for subgroup differences: Not applicable
0.05 0.2 1 5 20
Analysis 1.2. Comparison 1 Dependency or death, Outcome 2 Death.
Outcome: 2 Death

Azevedo 2009 4/14 6/20 3.1 % 0.93 [ 0.21, 4.20 ]
GIST-UK 2007 139/464 128/469 77.4 % 1.14 [ 0.86, 1.51 ]
GRASP 2009 3/24 7/50 3.4 % 0.88 [ 0.21, 3.74 ]
INSULINFARCT 2012 9/90 14/90 10.9 % 0.60 [ 0.25, 1.47 ]
Kreisel 2009 5/20 3/20 2.0 % 1.89 [ 0.38, 9.27 ]
McCormick 2010 2/25 0/15 0.5 % 3.30 [ 0.15, 73.45 ]
Staszewski 2011 1/26 2/24 1.7 % 0.44 [ 0.04, 5.19 ]
THIS 2008 2/31 0/15 0.5 % 2.63 [ 0.12, 58.20 ]
Walters 2006 1/13 0/12 0.4 % 3.00 [ 0.11, 80.95 ]
Total (95% CI) 707 715 100.0 % 1.09 [ 0.85, 1.41 ]

0.02 0.1 1 10 50

Analysis 1.3. Comparison 1 Dependency or death, Outcome 3 Diabetes mellitus versus no diabetes mellitus.
Outcome: 3 Diabetes mellitus versus no diabetes mellitus

1 Diabetes mellitus
GRASP 2009 14/24 34/49 6.0 % 0.62 [ 0.22, 1.70 ]
THIS 2008 16/31 8/15 3.3 % 0.93 [ 0.27, 3.21 ]
Vinychuk 2005 7/36 12/39 5.9 % 0.54 [ 0.19, 1.58 ]
Subtotal (95% CI) 91 103 15.2 % 0.66 [ 0.35, 1.24 ]

2 No diabetes mellitus
GIST-UK 2007 336/461 330/467 56.9 % 1.12 [ 0.84, 1.49 ]
INSULINFARCT 2012 49/90 49/90 14.3 % 1.00 [ 0.56, 1.80 ]
Kreisel 2009 15/20 15/20 2.4 % 1.00 [ 0.24, 4.18 ]
McCormick 2010 21/25 13/15 1.7 % 0.81 [ 0.13, 5.05 ]
Staszewski 2011 14/26 17/24 5.2 % 0.48 [ 0.15, 1.55 ]
Vinychuk 2005 7/24 10/26 4.3 % 0.66 [ 0.20, 2.15 ]
Subtotal (95% CI) 646 642 84.8 % 1.02 [ 0.81, 1.30 ]

Total (95% CI) 737 745 100.0 % 0.97 [ 0.77, 1.21 ]
Test for subgroup differences: Chi2 = 1.66, df = 1 (P = 0.20), I2 =40%
0.1 0.2 0.5 1 2 5 10


Analysis 1.4. Comparison 1 Dependency or death, Outcome 4 Less than 30 days versus 90 days of follow-up.
Outcome: 4 Less than 30 days versus 90 days of follow-up

1 30 days
Azevedo 2009 13/14 15/20 0.6 % 4.33 [ 0.45, 42.02 ]
Kreisel 2009 15/20 15/20 2.4 % 1.00 [ 0.24, 4.18 ]
McCormick 2010 21/25 13/15 1.7 % 0.81 [ 0.13, 5.05 ]
Staszewski 2011 14/26 17/24 5.2 % 0.48 [ 0.15, 1.55 ]
Vinychuk 2005 14/60 22/65 10.3 % 0.59 [ 0.27, 1.31 ]
Subtotal (95% CI) 145 144 20.1 % 0.74 [ 0.43, 1.25 ]

2 90 days
GIST-UK 2007 336/461 330/467 56.5 % 1.12 [ 0.84, 1.49 ]
GRASP 2009 14/24 34/49 5.9 % 0.62 [ 0.22, 1.70 ]
INSULINFARCT 2012 49/90 49/90 14.2 % 1.00 [ 0.56, 1.80 ]
THIS 2008 16/31 8/15 3.3 % 0.93 [ 0.27, 3.21 ]
Subtotal (95% CI) 606 621 79.9 % 1.05 [ 0.82, 1.34 ]

Total (95% CI) 751 765 100.0 % 0.99 [ 0.79, 1.23 ]
0.1 0.2 0.5 1 2 5 10


Analysis 2.1. Comparison 2 Functional neurological outcome, Outcome 1 NIHSS or ESS at the end of the
follow-up.
Comparison: 2 Functional neurological outcome
Outcome: 1 NIHSS or ESS at the end of the follow-up
Std. Std.
Mean Mean
Study or subgroup Experimental Control Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI
GIST-UK 2007 464 73.4 (24.6) 469 74.5 (23.8) 66.1 % -0.05 [ -0.17, 0.08 ]
INSULINFARCT 2012 90 11 (3) 90 10.5 (3.8) 12.7 % 0.15 [ -0.15, 0.44 ]
Kreisel 2009 20 8.79 (6.8) 20 9.79 (6.55) 2.8 % -0.15 [ -0.77, 0.47 ]
Staszewski 2011 26 4.4 (3.45) 24 6.54 (3.52) 3.4 % -0.60 [ -1.17, -0.04 ]
THIS 2008 31 8.45 (11.36) 15 5.62 (4.68) 2.8 % 0.29 [ -0.33, 0.90 ]
Vinychuk 2005 60 8.7 (1.2) 65 9.4 (1.1) 8.5 % -0.61 [ -0.96, -0.25 ]
Vriesendorp 2009 23 2 (5.3) 10 2 (4.8) 2.0 % 0.0 [ -0.74, 0.74 ]
Walters 2006 13 4 (4.3) 12 6.5 (2.8) 1.7 % -0.66 [ -1.47, 0.15 ]
Total (95% CI) 727 705 100.0 % -0.09 [ -0.19, 0.01 ]

Heterogeneity: Chi2 = 17.42, df = 7 (P = 0.01); I2 =60%
-1 -0.5 0 0.5 1

Analysis 2.2. Comparison 2 Functional neurological outcome, Outcome 2 Independent in daily activities.
Outcome: 2 Independent in daily activities

Azevedo 2009 1/14 5/20 3.0 % 0.23 [ 0.02, 2.24 ]
GIST-UK 2007 94/309 105/327 56.6 % 0.92 [ 0.66, 1.29 ]
GRASP 2009 10/24 15/49 4.6 % 1.62 [ 0.59, 4.46 ]
INSULINFARCT 2012 41/90 41/90 17.8 % 1.00 [ 0.56, 1.80 ]
Kreisel 2009 5/20 5/20 3.0 % 1.00 [ 0.24, 4.18 ]
McCormick 2010 4/25 2/15 1.7 % 1.24 [ 0.20, 7.74 ]
Staszewski 2011 12/26 7/24 3.1 % 2.08 [ 0.65, 6.71 ]
THIS 2008 16/31 7/15 3.6 % 1.22 [ 0.35, 4.19 ]
Vinychuk 2005 13/60 11/65 6.6 % 1.36 [ 0.56, 3.32 ]
Total (95% CI) 599 625 100.0 % 1.03 [ 0.81, 1.32 ]

0.2 0.5 1 2 5

Analysis 2.3. Comparison 2 Functional neurological outcome, Outcome 3 Diabetes mellitus versus no
diabetes mellitus.
Outcome: 3 Diabetes mellitus versus no diabetes mellitus
Std. Std.
Mean Mean
1 Diabetes mellitus
THIS 2008 31 8.45 (11.36) 15 5.62 (4.68) 2.9 % 0.29 [ -0.33, 0.90 ]
Vinychuk 2005 36 8.6 (1.1) 39 8.9 (1.3) 5.4 % -0.25 [ -0.70, 0.21 ]
Walters 2006 13 4 (0) 12 6.5 (0) Not estimable
Subtotal (95% CI) 80 66 8.3 % -0.06 [ -0.43, 0.31 ]

2 No diabetes mellitus
GIST-UK 2007 464 73.4 (24.6) 469 74.5 (23.8) 67.3 % -0.05 [ -0.17, 0.08 ]
INSULINFARCT 2012 90 11 (3) 90 10.5 (3.8) 13.0 % 0.15 [ -0.15, 0.44 ]
Kreisel 2009 20 8.79 (6.8) 20 9.79 (6.55) 2.9 % -0.15 [ -0.77, 0.47 ]
Staszewski 2011 26 4.4 (3.45) 24 6.54 (3.52) 3.4 % -0.60 [ -1.17, -0.04 ]
Vinychuk 2005 24 8.9 (1.3) 26 10.1 (1) 3.2 % -1.02 [ -1.62, -0.43 ]
Vriesendorp 2009 23 2 (5.3) 10 2 (4.8) 2.0 % 0.0 [ -0.74, 0.74 ]
Subtotal (95% CI) 647 639 91.7 % -0.08 [ -0.19, 0.03 ]

Total (95% CI) 727 705 100.0 % -0.07 [ -0.18, 0.03 ]
Test for subgroup differences: Chi2 = 0.01, df = 1 (P = 0.94), I2 =0.0%
-2 -1 0 1 2

Analysis 2.4. Comparison 2 Functional neurological outcome, Outcome 4 Less than 30 days versus 90 days
of follow-up.
Outcome: 4 Less than 30 days versus 90 days of follow-up
Std. Std.
Mean Mean
1 30 days
Kreisel 2009 20 8.79 (6.8) 20 9.79 (6.55) 2.8 % -0.15 [ -0.77, 0.47 ]
Staszewski 2011 26 4.4 (3.45) 24 6.54 (3.52) 3.4 % -0.60 [ -1.17, -0.04 ]
Vinychuk 2005 60 8.7 (1.2) 65 9.4 (1.1) 8.5 % -0.61 [ -0.96, -0.25 ]
Vriesendorp 2009 23 2 (5.3) 10 2 (4.8) 2.0 % 0.0 [ -0.74, 0.74 ]
Walters 2006 13 4 (4.3) 12 6.5 (2.8) 1.7 % -0.66 [ -1.47, 0.15 ]
Subtotal (95% CI) 142 131 18.3 % -0.47 [ -0.72, -0.23 ]

2 90 days
GIST-UK 2007 464 73.4 (24.6) 469 74.5 (23.8) 66.1 % -0.05 [ -0.17, 0.08 ]
INSULINFARCT 2012 90 11 (3) 90 10.5 (3.8) 12.7 % 0.15 [ -0.15, 0.44 ]
THIS 2008 31 8.45 (11.36) 15 5.62 (4.68) 2.8 % 0.29 [ -0.33, 0.90 ]
Subtotal (95% CI) 585 574 81.7 % 0.00 [ -0.12, 0.11 ]

Total (95% CI) 727 705 100.0 % -0.09 [ -0.19, 0.01 ]
-4 -2 0 2 4

Analysis 3.1. Comparison 3 Hypoglycaemia, Outcome 1 Symptomatic hypoglycaemia.
Comparison: 3 Hypoglycaemia
Outcome: 1 Symptomatic hypoglycaemia

Azevedo 2009 1/14 2/20 25.3 % 0.69 [ 0.06, 8.47 ]
GIST-UK 2007 73/464 0/469 6.9 % 176.29 [ 10.89, 2854.26 ]
GRASP 2009 0/24 1/50 16.0 % 0.67 [ 0.03, 17.14 ]
INSULINFARCT 2012 0/90 0/90 Not estimable
Kreisel 2009 3/20 0/20 6.9 % 8.20 [ 0.40, 169.90 ]
McCormick 2010 1/25 0/15 9.7 % 1.90 [ 0.07, 49.60 ]
Staszewski 2011 2/26 0/24 7.8 % 5.00 [ 0.23, 109.62 ]
THIS 2008 5/31 0/15 9.1 % 6.43 [ 0.33, 124.43 ]
Vriesendorp 2009 1/23 0/10 10.6 % 1.40 [ 0.05, 37.33 ]
Walters 2006 1/13 0/12 7.7 % 3.00 [ 0.11, 80.95 ]
Total (95% CI) 730 725 100.0 % 14.60 [ 6.62, 32.21 ]

Test for overall effect: Z = 6.64 (P < 0.00001)
0.05 0.2 1 5 20

Analysis 3.2. Comparison 3 Hypoglycaemia, Outcome 2 Hypoglycaemia (with or without symptoms).
Comparison: 3 Hypoglycaemia
Outcome: 2 Hypoglycaemia (with or without symptoms)

Azevedo 2009 1/14 2/20 22.8 % 0.69 [ 0.06, 8.47 ]
GIST-UK 2007 73/464 0/469 6.2 % 176.29 [ 10.89, 2854.26 ]
GRASP 2009 7/24 2/50 13.7 % 9.88 [ 1.87, 52.29 ]
INSULINFARCT 2012 5/90 1/90 14.1 % 5.24 [ 0.60, 45.74 ]
Kreisel 2009 7/20 1/20 9.7 % 10.23 [ 1.12, 93.34 ]
McCormick 2010 19/25 1/15 4.5 % 44.33 [ 4.78, 410.94 ]
Staszewski 2011 4/26 0/24 6.5 % 9.80 [ 0.50, 192.41 ]
THIS 2008 12/31 0/15 6.1 % 19.87 [ 1.09, 362.72 ]
Vriesendorp 2009 1/23 0/10 9.6 % 1.40 [ 0.05, 37.33 ]
Walters 2006 1/13 0/12 6.9 % 3.00 [ 0.11, 80.95 ]
Total (95% CI) 730 725 100.0 % 18.41 [ 9.09, 37.27 ]

0.01 0.1 1 10 100


Analysis 4.1. Comparison 4 Mean glucose level, Outcome 1 Mean glucose level.
Comparison: 4 Mean glucose level
Outcome: 1 Mean glucose level
Mean Mean
Azevedo 2009 14 7.73 (1.44) 20 8.16 (2.18) 1.9 % -0.43 [ -1.65, 0.79 ]
GIST-UK 2007 464 6.87 (1.68) 469 7.23 (1.79) 55.4 % -0.36 [ -0.58, -0.14 ]
GRASP 2009 24 6.17 (4) 50 8.39 (2) 1.0 % -2.22 [ -3.91, -0.53 ]
INSULINFARCT 2012 90 5.7 (0.54) 90 6.6 (1.26) 34.3 % -0.90 [ -1.18, -0.62 ]
Kreisel 2009 20 6.49 (2.19) 20 8.01 (3.06) 1.0 % -1.52 [ -3.17, 0.13 ]
McCormick 2010 26 5.8 (4.74) 15 7 (4.74) 0.3 % -1.20 [ -4.21, 1.81 ]
Staszewski 2011 26 6 (1.55) 24 6.89 (1) 5.3 % -0.89 [ -1.61, -0.17 ]
THIS 2008 31 7.39 (0.89) 15 10.56 (3.56) 0.8 % -3.17 [ -5.00, -1.34 ]
Total (95% CI) 695 703 100.0 % -0.63 [ -0.80, -0.46 ]

-4 -2 0 2 4
ADDITIONAL TABLES
Table 1. Risk of bias summary
Study Generation of ran- Allocation Blinding: Blinding: Lost to follow-up

domisation concealment participants and outcome to alloca- (%)
physicians tion group
Vinychuk 2005 Low risk Low risk High risk High risk 0
GIST-UK 2007 Low risk Low risk High risk Low risk 7.4
Staszewski 2011 Low risk High risk High risk Low risk 0
THIS 2008 Low risk Low risk High risk Low risk 0

Table 1. Risk of bias summary (Continued)
Walters 2006 Low risk Low risk High risk High risk 0
GRASP 2009 Low risk Low risk High risk Low risk 1.4
Kreisel 2009 Low risk Unclear risk High risk High risk 10
McCormick 2010 Low risk Unclear risk Unclear risk High risk Not reported
INSULINFARCT Low risk Low risk High risk High risk 2.2
2012
Vriesendorp 2009 High risk High risk High risk High risk 15.2
Azevedo 2009 Low risk Unclear risk High risk High risk Not reported
APPENDICES
Appendix 1. MEDLINE search strategy

We used the following search strategy for MEDLINE (Ovid).
1. cerebrovascular disorders/ or basal ganglia cerebrovascular disease/ or exp brain ischemia/ or carotid artery diseases/ or carotid artery
thrombosis/ or intracranial arterial diseases/ or cerebral arterial diseases/ or exp “intracranial embolism and thrombosis”/ or exp stroke/
2. (isch?emi$ adj6 (stroke$ or apoplex$ or cerebral vasc$ or cerebrovasc$ or cva or attack$)).tw.
3. ((brain or cerebr$ or cerebell$ or vertebrobasil$ or hemispher$ or intracran$ or intracerebral or infratentorial or supratentorial or
middle cerebr$ or mca$ or anterior circulation) adj5 (isch?emi$ or infarct$ or thrombo$ or emboli$ or occlus$ or hypoxi$)).tw.
4. 1 or 2 or 3
5. exp insulins/ or insulin infusion systems/
6. (insulin$ or novolin or humulin or iletin or velosulin).tw.
7. insulin$.nm.
8. 5 or 6 or 7
9. 4 and 8
10. exp animals/ not humans.sh
11. 9 not 10

Appendix 2. Cochrane CENTRAL search strategy
#1 [mh ˆ“cerebrovascular disorders”] or [mh ˆ“basal ganglia cerebrovascular disease”] or [mh ˆ“brain ischemia”] or [mh “brain infarction”]
or [mh ˆ“hypoxia-ischemia, brain”] or [mh ˆ“carotid artery diseases”] or [mh ˆ“carotid artery thrombosis”] or [mh ˆ“carotid artery,
internal, dissection”] or [mh ˆ“intracranial arterial diseases”] or [mh ˆ“cerebral arterial diseases”] or [mh ˆ“infarction, anterior cerebral
artery”] or [mh ˆ“infarction, middle cerebral artery”] or [mh ˆ“infarction, posterior cerebral artery”] or [mh “intracranial embolism
and thrombosis”] or [mh stroke] or [mh ˆ“vertebral artery dissection”]
#2 isch*mi* near/6 (stroke* or apoplex* or cerebral next vasc* or cerebrovasc* or cva or attack*):ti,ab,kw (Word variations have been
searched)
#3 (brain or cerebr* or cerebell* or vertebrobasil* or hemispher* or intracran* or intracerebral or infratentorial or supratentorial or
middle next cerebr* or mca* or “anterior circulation”) near/5 (isch*mi* or infarct* or thrombo* or emboli* or occlus* or hypoxi*):
ti,ab,kw (Word variations have been searched)
#4 #1 or #2 or #3
#5 [mh insulins] or [mh ˆ“insulin infusion systems”]
#6 insulin* or novolin or humulin or iletin or velosulin:ti,ab,kw (Word variations have been searched)
#7 #5 or #6
#8 #4 and #7
Appendix 3. EMBASE search strategy

EMBASE (Ovid)
1. cerebrovascular disease/ or brain infarction/ or brain stem infarction/ or cerebellum infarction/ or exp brain ischemia/ or carotid artery
disease/ or exp carotid artery obstruction/ or cerebral artery disease/ or exp cerebrovascular accident/ or exp occlusive cerebrovascular
disease/ or stroke/ or stroke patient/
2. (isch?emi$ adj6 (stroke$ or apoplex$ or cerebral vasc$ or cerebrovasc$ or cva or attack$)).tw.
3. ((brain or cerebr$ or cerebell$ or vertebrobasil$ or hemispher$ or intracran$ or intracerebral or infratentorial or supratentorial or
middle cerebr$ or mca$ or anterior circulation) adj5 (isch?emi$ or infarct$ or thrombo$ or emboli$ or occlus$ or hypoxi$)).tw.
4. 1 or 2 or 3
5. insulin/ or biphasic insulin/ or bovine insulin/ or globin zinc insulin/ or human insulin/ or insulin aspart/ or insulin aspart plus insulin
degludec/ or insulin degludec/ or insulin detemir/ or insulin glargine/ or insulin glulisine/ or insulin lispro/ or insulin peglispro/ or
insulin tregopil/ or insulin zinc suspension/ or isophane insulin/ or long acting insulin/ or neutral insulin/ or pig insulin/ or recombinant
human insulin/ or short acting insulin/ or synthetic insulin/
6. exp insulin treatment/
7. (insulin$ or novolin or humulin or iletin or velosulin).tw.
8. 5 or 6 or 7
9. Randomized Controlled Trial/
10. Randomization/
11. Controlled Study/
12. control group/
13. clinical trial/ or phase 1 clinical trial/ or phase 2 clinical trial/ or phase 3 clinical trial/ or phase 4 clinical trial/ or controlled clinical
trial/
14. Double Blind Procedure/
15. Single Blind Procedure/
16. triple blind procedure/
17. placebo/
18. “types of study”/
19. random$.tw.
20. (controlled adj5 (trial$ or stud$)).tw.
21. (clinical$ adj5 trial$).tw.
22. ((control or treatment or experiment$ or intervention) adj5 (group$ or subject$ or patient$)).tw.
23. (quasi-random$ or quasi random$ or pseudo-random$ or pseudo random$).tw.
24. ((control or experiment$ or conservative) adj5 (treatment or therapy or procedure or manage$)).tw.
25. ((singl$ or doubl$ or tripl$ or trebl$) adj5 (blind$ or mask$)).tw.
26. (placebo$ or sham).tw.
27. trial.ti.
28. (assign$ or allocat$).tw.
29. (RCT or RCTs).tw.
30. or/9-29
31. 4 and 8 and 30
32. (exp animals/ or exp invertebrate/ or animal experiment/ or animal model/ or animal tissue/ or animal cell/ or nonhuman/) not
(human/ or normal human/ or human cell/)
33. 31 not 32
Appendix 4. SCOPUS search strategy

Query: (TITLE-ABS-KEY((stroke* OR cerebrovascular OR brain OR cerebral) AND (infarct* OR accident* OR ischem* OR is-
chaem*)) AND TITLE-ABS-KEY(insulin AND acute AND hyperglyc*))
Appendix 5. Web of Science search strategy

TS=(hyperglyc*)
TS=((brain or cerebrovascular) same (ischem* or ischaem* or infarct* or accident))
TS=(trial* or intervention* or random* or placebo* or hyperglyc*)
TS=(ischemi* OR ischaem*)
TS=(insulin)
TS=((acute or emergen*) SAME stroke*)
Appendix 6. CINAHL (EBSCO)

S10 .S5 AND S9
S9 .S6 OR S7 OR S8
S8 .TI ( insulin* or novolin or humulin or iletin or velosulin ) OR AB ( insulin* or novolin or humulin or iletin or velosulin )
S7 .(MH “Insulin Infusion Systems”) OR (MH “Insulin Injection (Saba CCC)”)
S6 .(MH “Insulins+”)
S5 .S1 OR S2 OR S3 OR S4
S4 .TI ( (brain or cerebr* or cerebell* or vertebrobasil* or hemispher* or intracran* or intracerebral or infratentorial or supratentorial
or middle cerebr* or mca* or anterior circulation) N5 (isch?emi* or infarct* or thrombo* or emboli* or occlus* or hypoxi*) ) OR AB (
(brain or cerebr* or cerebell* or vertebrobasil* or hemispher* or intracran* or intracerebral or infratentorial or supratentorial or middle
cerebr* or mca* or anterior circulation) N5 (isch?emi* or infarct* or thrombo* or emboli* or occlus* or hypoxi*) )
S3 .TI ( isch#emi* N6 (stroke* or apoplex* or cerebral vasc* or cerebrovasc* or cva or attack*) ) OR AB ( isch#emi* N6 (stroke* or
apoplex* or cerebral vasc* or cerebrovasc* or cva or attack*) )
S2 .(MH “Stroke Patients”)
S1 .(MH “Cerebrovascular Disorders”) OR (MH “Basal Ganglia Cerebrovascular Disease”) OR (MH “Carotid Artery Diseases”) OR
(MH “Carotid Artery Thrombosis”) OR (MH “Cerebral Ischemia+”) OR (MH “Intracranial Arterial Diseases”) OR (MH “Cerebral
Arterial Diseases”) OR (MH “Intracranial Embolism and Thrombosis+”) OR (MH “Stroke”)

WHAT’S NEW
Date Event Description
30 October 2013 New search has been performed We updated the literature searches to September 2013.
We included four new studies. There are now 11 studies
involving 1583 participants included in the review. The
conclusions have not changed
30 October 2013 New citation required but conclusions have not Updated review. Conclusions not changed.
changed
HISTORY
Protocol first published: Issue 3, 2005
Review first published: Issue 9, 2011
Date Event Description
17 March 2008 Amended Converted to new review format.
5 February 2008 New citation required and major changes The protocol has been extensively revised and updated and replaces
the previously published version. There has been a change of authors
CONTRIBUTIONS OF AUTHORS
MF Bellolio wrote the background section, collected the data, did the analyses and wrote the results and conclusion of the full review
and the updated review.
RM Gilmore wrote the background section, collected the data, wrote the results and conclusion of the full review, and abstracted data
and quality measurements for the updated review.
LG Stead devised the initial review and wrote the original protocol.
MFB and RMG updated this version of the review.
The authors approve and take full responsibility for the contents of this systematic review.

DECLARATIONS OF INTEREST
None of the authors has any conflicts of interest to disclose.
SOURCES OF SUPPORT
Internal sources
• Mayo Clinic, USA.
Protected time for research to the principal investigator
External sources
• No sources of support supplied
INDEX TERMS
Medical Subject Headings (MeSH)

Blood Glucose [∗ metabolism]; Diabetes Mellitus [mortality]; Hyperglycemia [blood; complications; ∗ drug therapy; mortality]; Hypo-
glycemia [blood; complications; mortality]; Hypoglycemic Agents [∗ administration & dosage]; Insulin [∗ administration & dosage];
Prognosis; Randomized Controlled Trials as Topic; Reference Values; Stroke [∗ blood; complications]
MeSH check words

Aged; Female; Humans; Male


Insulin For Glycaemic Control in Acute Ischaemic Stroke (Review)

Uploaded by

Copyright:

Available Formats

Insulin For Glycaemic Control in Acute Ischaemic Stroke (Review)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Insulin For Glycaemic Control in Acute Ischaemic Stroke (Review)

Uploaded by

Copyright:

Available Formats

Cochrane Database of Systematic Reviews

Insulin for glycaemic control in acute ischaemic stroke

Bellolio MF, Gilmore RM, Ganti L

Bellolio MF, Gilmore RM, Ganti L.

Insulin for glycaemic control in acute ischaemic stroke (Review)

Insulin for glycaemic control in acute ischaemic stroke (Review) i

Insulin for glycaemic control in acute ischaemic stroke

M Fernanda Bellolio1 , Rachel M Gilmore1 , Latha Ganti2

Editorial group: Cochrane Stroke Group.

PLAIN LANGUAGE SUMMARY

BACKGROUND 1982; Rehncrona 1981). Enhanced acidosis may exaggerate is-

Description of the intervention METHODS

Why it is important to do this review

Insulin for glycaemic control in acute ischaemic stroke (Review) 3

Insulin for glycaemic control in acute ischaemic stroke (Review) 4

Dealing with missing data

Assessment of reporting biases Sensitivity analysis

Insulin for glycaemic control in acute ischaemic stroke (Review) 5

Figure 1. Study flow diagram.

proportion of people with diabetes mellitus.

Insulin for glycaemic control in acute ischaemic stroke (Review) 7

Insulin for glycaemic control in acute ischaemic stroke (Review) 8

Insulin for glycaemic control in acute ischaemic stroke (Review) 9

Insulin for glycaemic control in acute ischaemic stroke (Review) 10

Insulin for glycaemic control in acute ischaemic stroke (Review) 11

Insulin for glycaemic control in acute ischaemic stroke (Review) 12

Implications for research

Insulin for glycaemic control in acute ischaemic stroke (Review) 15

Insulin for glycaemic control in acute ischaemic stroke (Review) 16

Insulin for glycaemic control in acute ischaemic stroke (Review) 17

Characteristics of included studies [ordered by study ID]

Methods Randomised, parallel design

Interventions Continous intravenous insulin infusion

Outcomes Follow-up until discharge

Notes Abstract format only

Bias Authors’ judgement Support for judgement

Allocation concealment (selection bias) Unclear risk Not reported

Blinding of participants and personnel High risk Not blinded

Blinding of outcome assessment (detection High risk Not blinded

Incomplete outcome data (attrition bias) Unclear risk Not reported

Methods Randomised, parallel design

Insulin for glycaemic control in acute ischaemic stroke (Review) 18

Interventions Glucose-insulin-potassium (10% dextrose, 20 mmol KCl, 16 units insulin); continuous

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Low risk Adequate

Allocation concealment (selection bias) Low risk Adequate

Blinding of participants and personnel High risk Not blinded

Blinding of outcome assessment (detection Low risk Blinded

Methods Randomised, parallel design, stratified by glucose concentration

Participants 74 participants (3 arms): 24 participants in the intervention group (tight control), 50

Insulin for glycaemic control in acute ischaemic stroke (Review) 19

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Low risk Adequate

Allocation concealment (selection bias) Low risk Adequate

Blinding of participants and personnel High risk Not blinded

Blinding of outcome assessment (detection Low risk Blinded

Methods Randomised, parallel design, not blinded

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Low risk Adequate