Acta Neurol Scand 2010: 122: 404–408 DOI: 10.1111/j.1600-0404.2010.01326.
x Copyright  2010 The Authors
Fever is associated with doubling of odds of
short-term mortality in ischemic stroke: an
updated meta-analysis
Prasad K, Krishnan PR. Fever is associated with doubling of odds of
short-term mortality in ischemic stroke: an updated meta-analysis.
Acta Neurol Scand: 2010: 122: 404–408.
 2010 The Authors Journal compilation  2010 Blackwell Munksgaard.
Objective – Association between fever and ischemic stroke mortality is
known, but the magnitude and independence of the association is
controversial. This paper aims to determine the size of independent effect
of fever on short term mortality in acute ischemic stroke. Methods – We
searched the Medline and Cochrane library databases for papers
studying the relationship between fever in acute ischemic stroke and
short term mortality from January, 1990 to November, 2008. Two
authors independently selected the studies for inclusion in the review
using explicit criteria. Data was entered into software Revman 4.2.8.
Heterogeneity was assessed using I2 and chi-square statistics. Odds
ratios (OR) from logistic regression were combined. Magnitude of
association was determined using meta-analysis of the adjusted odds
ratio using fixed effects model. Results – Six cohort studies involving
2986 patients were included. There was no significant heterogeneity
among studies reporting short-term mortality (I2 = 21.2%, P = 0.28).
Meta-analysis yielded a combined OR of 2.20 (95% CI 1.59–3.03,
P < 0.00001). Conclusions – This meta-analysis suggests that fever
within first 24 h of hospitalization in patients with ischemic stroke is
associated with doubling of odds of mortality within one month of the
onset of stroke.
Introduction
Stroke is the third leading cause of morbidity and
mortality globally. Several factors influence mor-
tality after acute ischemic stroke (1–3). Fever is one
of them. Experiments on rat middle cerebral artery
occlusion model have shown that elevated body
temperature increases the volume of cerebral infarct
(4, 5). Several human studies have been reported but
with inconsistent results. Kammersgaard et al. (6)
and others (7–15) found significantly higher
mortality rates among stroke patients with fever as
compared with patients without fever. Almost an
equal number of studies (16–22) did not find fever to
be associated with higher mortality in stroke. Hajat
et al. (23) in their meta-analysis stated that pyrexia
after acute stroke is associated with marked increase
in morbidity and mortality, but the confidence
interval was wide (OR 1.19; 95% confidence interval
404
Journal compilation  2010 Blackwell Munksgaard
ACTA NEUROLOGICA
SCANDINAVICA
K. Prasad1, P. R. Krishnan2
1
Department of Neurology, All India Institute of Medical
Sciences, New Delhi, India; 2Department of Neurology,
Fortis Hospitals, Bangalore, India.
Key words: ischemic; stroke; short-term; mortality;
fever; meta-analysis
Kameshwar Prasad, Department of Neurology, Room
704, Neurosciences Centre, All India Institute of
Medical Sciences, New Delhi, India – 110029
Tel ⁄ fax: +91 11 26588979
e-mail: drkameshwarprasad@yahoo.co.in
Accepted for publication December 18, 2009
(CI) 0.99–1.43) indicating need for more data to
improve precision. We added more data to the meta-
analysis by including the studies reported after its
publication. We also considered that any indepen-
dent effect of fever will be seen mainly in the short
term (in hospital or within 30 days), and mortality
after this period will be influenced by many other
factors, like level of control of risk factors, occur-
rence of new vascular events, myocardial infarction
and others. Therefore, the objective of this study
was to determine the size of independent effect of
fever on the outcome of mortality within one month
after acute ischemic stroke.
Methods
We searched Medline and Cochrane library from
January 1990 to November 2008 with the search
terms: stroke, pyrexia, fever and combined them

with mortality. All resulting titles were scanned
and relevant articles were analyzed in detail. We
also sought for potential studies for inclusion by
searching bibliographies of identified studies. We
attempted to contact authors for clarification of
unclear or insufficient information. Two authors
independently assessed the studies for inclusion in
the review and any disagreement was resolved by
discussion. Studies were eligible in case they met
the following inclusion criteria: (i) they had
patients with acute ischemic stroke with mortality
within 30 days or in-hospital mortality as outcome,
(ii) had performed multivariable analysis (iii) had
adjusted for age and stroke severity (as our aim
was to determine the association between fever and
outcome, independent of age and stroke severity).
We separately extracted adjusted odds ratios and
standard errors from logistic regression analysis of
published studies and used generic inverse variance
method to combine the results. Chi-square and I2
were used to examine heterogeneity. Statistical
analysis was done using Cochrane Collaboration
software, Revman 4.2.8 (24).
Results
We found six cohort studies (7, 8, 12, 15, 18, 22)
eligible for inclusion in the meta-analysis. The
remaining nine studies (6, 9, 11, 13, 14, 16, 19–21)
were excluded as they did not provide odds ratios
from logistic regression or did not provide mortal-
ity data.
Six cohort studies selected as per inclusion
criteria had reported odds ratios and standard
errors for mortality. These studies included a total
of 2986 patients (Table 1). Table 2 details excluded
studies. Table 3 details the variables included in
multivariate analysis in various studies.
To estimate the size of effect of fever on short-
term mortality, we combined five studies and
excluded the study (18) reporting one year mortal-
Table 1 Included studies
Total No of
number of patients
Study patients with fever
Azzimondi 183 78 Fever within 7 days Temp >37.9C
Hanchaiphiboolkul 332 88
Jorgensen 166 NA From 6–12 h of stroke Temp >37.4C
McWalter 1628 73
Prasad 287 12
Reith 390 97
NA, data not available.
Meta-analysis of fever and stroke mortality
ity. We used fixed effects model as there was no
significant heterogeneity across the studies
(P = 0.28, I2 = 21.2%). The combined estimate
of the five studies yielded an odds ratio (OR) of
2.20 (95% CI 1.59–3.03, P < 0.00001) (Fig. 1).
We also conducted meta-analysis of all six
studies using a fixed effect model, to assess the
impact of including the sixth study. As expected,
the heterogeneity across the studies became statis-
tically significant (P = 0.0002, I2 = 79.7%). Fur-
ther, meta-analysis with random effects model
yielded an OR of 1.86 (95% CI 1.10–3.15,
P = 0.02) (Fig. 2).
In both cases, a statistically significant increase
(by about two times) in odds of death in patients
with fever was observed.
Discussion
Our meta-analysis shows that fever occurring
within the first 24 h of onset of ischemic stroke
is associated with almost twice the risk of short
term mortality. This association seems to be
independent of age, stroke severity or co-morbid-
ity (hypertension, diabetes, hyperlipidemia)
because the odds ratios were obtained from the
results of logistic regression analysis of the five
studies that were adjusted for confounding fac-
tors. The results are consistent across the included
studies as the test of heterogeneity has a statisti-
cally non-significant chi square test as well as the
I2 value of 21.2% (Fig. 1). This meta-analysis
builds on previous one by adding new studies and
combining the adjusted odds ratios. Greer et al. in
a recent meta-analysis of 14,431 patients with
stroke and other brain injuries found that fever is
consistently associated with worse outcomes
across multiple outcome measures (25). They
found that patients with hemorrhagic or ischemic
stroke or traumatic brain injury who had fever or
raised body temperature had a prolongation of
Inclusion criteria Mortality Cause of fever Comments
53 NA 30-day mortality
Less than 72 h Temp >37.5C 19 Infectious cause In hospital mortality
in 28 patients
NA NA In hospital mortality
Less than 24 h Temp >37.5C NA NA 12-month mortality
Less than 24 h Temp >37.5C 3 NA In hospital mortality
Less than 6 h Temp >37.4C NA Infectious in In hospital mortality
61 patients
405
Prasad & Krishnan

Table 2 Excluded Studies A randomized trial of acetaminophen in acute

stroke by Dippel et al. (26) reported a reduction of
Excluded study Reason for exclusion
body temperature up to 0.4C with daily doses of
Kamersgaard (6) Cox proportional hazards analysis 6000 mg; however, there was no statistically sig-
for 3 months survival nificant difference in functional outcome even at
Castillo 1994 (9) No logistic regression data 1 month. Hence therapeutic measures to help
Huo (11) No logistic regression data (Chinese)
Roy (13) No logistic regression data
overcome the harmful effect of fever on ischemic
Georgilis (14) Logistic regression not done stroke need further study.
Castillo 1998 (16) Multiple logistic regression done, but not den Hertog et al. published their study on
for mortality paracetamol in stroke (PAIS trial) along with an
Boysen (19) Multinomial logistic regression for mRS
updated meta-analysis based on data of six phar-
at 3 months, no mortality data
Sharma (20) No logistic regression data macological temperature-lowering trials. Their trial
Wong (21) No multivariate analysis as well as the updated meta-analysis failed to show
a statistically significant change in outcome with
the use of paracetamol. However, post-hoc analysis
Table 3 Variables included in multivariate analysis in included studies of patients with baseline body temperature
37–39C treated with paracetamol was associated
Year of Variables included in with improved outcome (27, 28). We agree with
Study publication multivariate analysis
their conclusion that in a subgroup of patients with
Azzimondi (8) 1995 Level of conciousness, a baseline temperature of 37–39C, there is a
fever, age, glycemia potentially important clinical benefit. For every
Hanchaiphiboolkul (12) 2005 Highest temp recorded
1000 patients treated, 90 more show improvement
during 72 hours of
admission, conciousness, beyond expectation, which is clinically important.
ischemic heart disease However, this observation comes from post-hoc
Jorgensen (15) 1996 Initial stroke severity, analysis, and hence requires confirmation in an
infarct size, mortality, independent large clinical trial. Our results are
outcome in survivors
McWalter (18) 1998 Body temp, type of stroke,
consistent with the findings of the meta-analysis by
pre and post stroke ADL, Greer et al. who observed a relative risk of 1.5 for
admission neurological score, mortality in patients with raised body temperature
mentality status, smoking in ischemic stroke (25).
Prasad (22) 2008 GCS score, fever, mortality
Reith (7) 1996 Infarct size, SSS score on
Two questions arise: (1) Why are the reports of
admission, mortality, SSS the association of stroke mortality and fever
score at discharge conflicting? (2) What is the pathophysiologic
basis for this association?

hospital stay by a factor of 3.2 and had 1.3 times

Reasons for conflicting reports
increase in probability of poorer outcome com-
pared to patients without fever or raised body Data on the association of stroke mortality and
temperature. fever is not homogeneous and is exemplified by the
Review: Fever is associated with doubling of odds of short term mortality in ischemic stroke: an updated meta-analysis
Comparison: 01 With fever versus without fever
Outcome: 01 mortality
Study Odds ratio (fixed) Weight Odds ratio (fixed)
or sub-category log [Odds Ratio] (SE) 95% Cl % 95% Cl

Azzimondi 1995 1.2240 (0.5280) 9.71 3.40 [1.21, 9.57]

Hanchaiphiboolkul 1.3740 (0.3900) 17.80 3.95 [1.84, 8.49]
Jorgensen 1996 0.7420 (0.3680) 19.99 2.10 [1.02, 4.32]
Prasad 2008 –0.0800 (0.7280) 5.11 0.92 [0.22, 3.85]
Reith 1996 0.5880 (0.2390) 47.39 1.80 [1.13, 2.88]

Total (95% Cl) 100.00 2.20 [1.59, 3.03]

Test for heterogeneity: χ2 = 5.08, df = 4 (P = 0.28), I 2 = 21.2%
Test for overall effect Z = 4.78 (P < 0.00001)

0.001 0.01 0.1 1 10 100 1000

Fever decreases risk Fever increases risk

Figure 1. Forest plot of adjusted odds ratios of all studies except study by McWalter et al. (fixed effect model).

406

Review: Fever is associated with doubling of odds of short term mortality in ischemic stroke: an updated meta-analysis
Comparison: 01 With fever versus without fever
Outcome: 01 mortality
Study
or sub-category log [Odds Ratio] (SE)
Azzimondi 1995 1.2240 (0.5280)
Hanchaiphiboolkul 1.3740 (0.3900)
Jorgensen 1996 0.7420 (0.3680)
McWalter 1998 0.0120 (0.0600)
Prasad 2008 –0.0800 (0.7280)
Reith 1996 0.5880 (0.2390)
Total (95% Cl)
Test for heterogeneity: χ2 = 24.63, df = 5 (P = 0.0002), I 2 = 79.7%
Test for overall effect Z = 2.33 (P = 0.02)
0.001 0.01 0.1
Fever decreases risk
Figure 2. Forest plot of adjusted odds ratio of all studies (random effects model).
fact that ten studies (6–15) found significantly
higher mortality rates among patients with fever as
compared to patients without fever, whereas seven
studies (16–22) did not find fever to be associated
with higher mortality in stroke. This may be due to
one or more of the following reasons:
1. Selection bias: If patient population among the
studies varies in time since onset, it introduces
different degrees of survivor bias in different
studies. For example, if a study has a time
window of 1 week since onset, those presenting
towards the end of a week represent patients
ÔsurvivingÕ the risk of mortality in the first few
days.
2. The definition of fever varies from study to
study. Fever was defined as temperature >37.4
degrees centigrade in two studies (7, 15), 37.5C
in three and (12, 18, 22), 37.9C in one (8).
3. The most common reason is the inadequate
sample size to detect the association adjusted
for other prognostic factors like age, GCS score,
etc. This is reflected in statistically non-signifi-
cant but wide CI in some of the studies.
4. Some studies may not have conducted an
adjusted (regression) analysis. A crude analysis
may overestimate or underestimate the associ-
ation that can be found in a regression analysis.
Pathophysiologic rationale
What is the pathophysiologic basis for this asso-
ciation? Neuronal damage produced by ischemia
generates neurotoxic substances, such as superox-
ide, nitric oxide, or cytokines, which could induce a
rise in temperature after the ischemic episode.
Interleukin-1b induces fever and is expressed in the
Meta-analysis of fever and stroke mortality
Odds ratio (random) Weight Odds ratio (random)
95% Cl % 95% Cl
12.65 3.40 [1.21, 9.57]
16.29 3.95 [1.84, 8.49]
16.93 2.10 [1.02, 4.32]
24.59 1.01 [0.90, 1.14]
8.76 0.92 [0.22, 3.85]
20.78 1.80 [1.13, 2.88]
100.00 1.86 [1.10, 3.15]
1 10 100 1000
Fever increases risk
post-ischemic brain. It has been found that hypo-
thermia may interrupt this cycle, and conversely,
hyperthermia would enhance or reactivate this
process (29). The explanation for the possible
neuroprotective effect of hypothermia in acute
stroke is not completely understood but most
studies suggested that it might be due to slowing
down of the apoptosis and necrosis in the penum-
bra. In ischemic stoke, the time window for
reperfusion to be established is likely to be widened
by such a measure, but no such benefit is expected
for hemorrhagic stroke (8). Georgilis et al. (14)
have studied the causes of fever in acute stroke.
Fever occurred in 37.6% of 330 patients; more
often in hemorrhagic stroke with mass effect,
transtentorial herniation and intraventicular
blood as also in large ischemic strokes. Patients
with fever had lower scores on admission on the
Glasgow Coma Scale. In their study 22.7% of
patients with fever had an infectious cause; the
development of fever was associated with prior use
of an invasive technique especially urinary cathe-
terization (14). This means that majority of
patients with stroke develop fever due to pro-
inflammatory cytokines and not due to infectious
cause.
To summarize, our updated meta-analysis shows
that fever is associated with doubling of odds of
short-term mortality in acute ischemic stroke and
the association is independent of age and stroke
severity
References
1. Carlberg B, Asplund K, Hagg E. Course of blood pressure
in different subsets of patients after acute stroke. Cere-
brovasc Dis 1991;1:281–7.
407
Prasad & Krishnan
2. Busto R, Dietrich WD, Globus MY, Valdes I, Scheinberg
P, Ginsberg MD. Small differences in intraischaemic brain
temperature critically determine the extent of ischaemic
injury. J Cereb Blood Flow Metab 1987;7:729–38.
3. Brandstater ME. Prognostication in stroke rehabilitation.
In: Chino N, Melvin JL, eds. Functional evaluation of
stroke patients. Tokyo: Springer-Verlag, 1996;93–102.
4. Kim Y, Busto R, Dietrich WD, Kraydieh S, Ginsberg MD.
Delayed postischemic hyperthermia in awake rats worsens
the histopathological outcome of transient focal cerebral
ischemia. Stroke 1996;27:2274–81.
5. Dietrich W, Busto R, Valde¢s I, Loor Y. Effects of nor-
mothermic versus mild hyperthermic forebrain ischemia in
rats. Stroke 1990;21:1318–25.
6. Kammersgaard LP, Jørgensen HS, Rungby JA et al.
Admission body temperature predicts long-term mortality
after acute stroke: The Copenhagen Stroke Study. Stroke
2002;33:1759–62.
7. Reith J, Jorgensen HS, Pedersen PM et al. Body temper-
ature in acute stroke: relation to stroke severity, infarct
size, mortality and outcome. Lancet 1996;347:422–5.
8. Azzimondi G, Bassein L, Nonino F, Fiorani L, Vignatelli L.
Fever in acute stroke worsens prognosis: a prospective
study. Stroke 1995;26:2040–3.
9. Castillo J, Martinez F, Leira R, Prieto JM, Lema M, Noya
M. Mortality and morbidity of acute cerebral infarction
related to temperature and basal analytic parameters.
Cerebrovasc Dis 1994;4:56–71.
10. Terent A, Andersson B. Prognosis for patients with cere-
brovascular stroke and transient ischaemic attacks. Ups J
Med Sci 1981;86:63–74.
11. Huo CN, Zhao YX. Relationship between fever and mor-
tality in patients with acute ischemic cerebrovascular
stroke [in Chinese]. Chin J Nurs 1997;32:640–2.
12. Hanchaiphiboolkul S. Body temperature and mortality in
acute cerebral infarction. J Med Assoc Thai 2005;88:26–
31.
13. Roy MK, Ray A. Effect of body temperature on mor-
tality of acute stroke. J Assoc Physicians India 2004;52:
959–61.
14. Georgilis K, Plomaritoglou A, Dafni U, Bassiakos Y,
Vemmos K. Aetiology of fever in patients with acute stroke.
J Int Med 1999;246:203–9.
15. Jorgensen HS, Reith J, Pedersen PM, Nakayama H, Olsen
TS. Body temperature and outcome in stroke patients.
Lancet 1996;348:193. Letter.
408
16. Castillo J, Davalos A, Marrugat J, Noya M. Timing for
fever-related brain damage in acute ischemic stroke. Stroke
1998;29:2455–60.
17. Sharma JC, Ross IN. Antipyretic therapy in acute stroke.
Lancet 1998;352:740–1.
18. McWalter R, McMahon A, Fraser H, Bruce V, Hendrick
S. Does body temperature on admission predict long-term
outcome after an acute stroke? Cerebrovasc Dis 1998;
8(Suppl 4):11.
19. Boysen G, Christensen H. Stroke severity determines body
temperature in acute stroke. Stroke 2001;32:413.
20. Sharma JC, Fletcher S, Vassallo M, Ross I. What influ-
ences outcome of stroke-pyrexia or dysphagia? Int J Clin
Pract 2001;55:17–20.
21. Wong AA, Davis JP, Schluter PJ, Henderson RD, OÕSul-
livan JD, Read SJ. The effect of admission physiological
variables on 30 day outcome after stroke. J Clin Neurosci
2005;12:905–10.
22. Prasad K, Krishnan PR. Fever is associated with doubling
of odds of short term mortality in ischemic stroke. Int J
Stroke 2008;3(Suppl 1):418.
23. Hajat C, Hajat S, Sharma P. Effects of poststroke pyrexia
on stroke outcome: a meta-analysis of studies in patients.
Stroke 2000;31:410–4.
24. Review Manager (RevMan) [Computer program]. Version
4.2 for Windows. Copenhagen: The Nordic Cochrane
Centre, The Cochrane Collaboration, 2003.
25. Greer DM, Funk SE, Reaven NL, Ouzounelli M, Uman
GC. Impact of fever on outcome in patients with stroke
and neurologic injury: a comprehensive meta-analysis.
Stroke 2008;39:3029–35.
26. Dippel DWJ, van Breda EJ et al. Effect of paracetamol
(acetaminophen) on body temperature in acute ischemic
stroke: a double-blind, randomized Phase II Clinical Trial.
Stroke 2001;32:1607–12.
27. den Hertog HM, van der Worp HB, van Gemert HM et al.
The Paracetamol (Acetaminophen) In Stroke (PAIS) trial:
a multicentre, randomised, placebo-controlled, phase III
trial. Lancet Neurol 2009;8:434–40.
28. Den Hertog HM, van der Worp HB, Tseng MC, Dippel
DW. Temperature-lowering therapy for acute stroke.
Stroke 2009;40:e481–2.
29. Coimbra C, Drake M, Boris-Moller F, Wieloch T. Long-
lasting neuroprotective effect of postischemic hypothermia
and treatment with an anti-inflammatory ⁄ antipyretic drug.
Stroke 1996;27:1578–85.