Clinical Microbiology and Infection 22 (2016) 814.e1e814.

e7

Contents lists available at ScienceDirect

Clinical Microbiology and Infection

journal homepage: www.clinicalmicrobiologyandinfection.com

Original article

Betamethasone and dexamethasone in adult community-acquired

bacterial meningitis: a quality registry study from 1995 to 2014
M. Glimåker 1, 2, *, M. Brink 3, P. Naucler 1, 2, J. Sjo
€ lin 4
1)
Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institutet, Stockholm, Sweden
2)
Department of Infectious Diseases, Karolinska University Hospital, Stockholm, Sweden
3)
Institute of Biomedicine, Department of Infectious Diseases, Sahlgrenska University Hospital, Gothenburg, Sweden
4)
Section of Infectious Diseases, Department Medical Sciences, Uppsala University, Uppsala, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Acute bacterial meningitis (ABM) is a highly lethal disease. Available data support the use of cortico-
Received 26 April 2016 steroids in high-income countries, but the effect on mortality is still controversial. The effects of corti-
Received in revised form costeroids on mortality and sequelae were evaluated in the national Swedish quality registry. In total,
21 June 2016
during 1995e2014 1746 adults with ABM were included, of whom 989 were treated with corticosteroids
Accepted 26 June 2016
Available online 9 July 2016
(betamethasone, n ¼ 766; dexamethasone, n ¼ 248; methylprednisolone, n ¼ 2), 498 were not given
corticosteroids and in 259 patients data for corticosteroids were missing. Fatal outcome was observed in
Editor: Professor L. Leibovici 8.9% of the patients in the corticosteroid-treated group vs. 17.9% in the non-corticosteroid-treated group
(p <0.001), resulting in an odds ratio (OR) of 0.57 with a 95% confidence interval (CI) of 0.40e0.81
Keywords: adjusted for age, sex, mental status, and door-to-antibiotic time. In patients with meningitis caused by
Bacterial meningitis S. pneumoniae, mortality was 10.2% in the corticosteroid-treated group and 21.3% in the non-
Betamethasone corticosteroid-treated group (p <0.001) with an adjusted OR of 0.50 (95% CI 0.31e0.80). In ABM pa-
Corticosteroid treatment tients with non-pneumococcal aetiology the adjusted OR was 0.71 (95% CI 0.40e1.26). Lower mortality
Dexamethasone
was observed in the corticosteroid-treated group with impaired mental status, whereas no significant
Mortality
difference was found in patients with unaffected mental status. The adjusted ORs for betamethasone and
dexamethasone were 0.49 (95% CI 0.28e0.84) and 0.61 (95% CI 0.37e1.01), respectively. Corticosteroid
treatment decreases mortality in ABM and should be administered initially with antibiotics in adult ABM
patients with impaired mental status regardless of presumed aetiology. Betamethasone seems to be at
least as effective as dexamethasone. M. Glimåker, CMI 2016;22:814.e1e814.e7
© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All
rights reserved.

Introduction hamper the diffusion of antibiotics into the cerebrospinal fluid

Acute bacterial meningitis (ABM) is a life-threatening disease
[1,2] in which early diagnosis and timely and effective antimicrobial
therapy are the major determinants for outcome [3,4]. Adjunctive
therapy with corticosteroids has been supported by animal studies
showing reduced inflammation in the central nervous system
(CNS) by dexamethasone [5]. On the other hand, there are experi-
mental data showing that corticosteroids may potentiate ischemic
injury to neurons and reduce meningeal permeability that could
* Corresponding author. Martin Glimåker, Department of Infectious Diseases,
Karolinska University Hospital, 171 76 Stockholm, Sweden.
E-mail address: martin.glimaker@karolinska.se (M. Glimåker).
(CSF) [6,7].
Clinical benefit of adjunctive corticosteroid treatment has been
established for meningitis in children caused by H. influenzae, with
reduced risk of hearing deficits [8]. In a large European placebo-
controlled multicentre trial of adults with ABM, administration of
dexamethasone reduced the risk of death and adverse outcomes
[9]. The beneficial effect of dexamethasone was most pronounced
in pneumococcal meningitis with reduced mortality from 34% to
14%. There are, however, other controlled trials, mostly from low-
income countries, that have not been able to confirm the efficacy
of adjuvant corticosteroids [10e13]. Several important factors such
as general health status, HIV-prevalence, disease severity, stan-
dards of diagnostics, and general care differ between high- and
low-income countries, making comparisons difficult of results of
ABM trials from different parts of the world.

http://dx.doi.org/10.1016/j.cmi.2016.06.019
1198-743X/© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7 814.e2

A recent meta-analysis of 25 trials investigating the effects of Materials and methods

corticosteroids in ABM found that corticosteroids reduced hearing
loss, irrespective of causative bacteria, and reduced mortality in
pneumococcal meningitis, without increasing severe adverse
events [14]. There was no benefit of corticosteroid therapy in low-
income countries. The investigators concluded that the present
evidence merits the use of corticosteroids in adults and children in
high-income countries, although the strength of the evidence was
suboptimal [15,16].
A sufficiently large randomized study to definitively answer
the question of whether corticosteroids have a place in the
treatment of ABM is unlikely to be performed in the foreseeable
future [17]. A recent study indicates beneficial effects of dexa-
methasone [1], but there is still a need for implementing studies.
The first Swedish national guidelines for the management of
bacterial CNS infections were published in 2004 by the Swedish
Association of Infectious Diseases, recommending adjuvant treat-
ment with dexamethasone to be initiated before or together with
the first dose of antibiotics, and continued for 2e4 days in all
adults with ABM. Because, traditionally, betamethasone was used
in most cases of brain oedema of non-ABM aetiology and dexa-
methasone has not been licensed in Sweden since 2002, beta-
methasone was recommended if dexamethasone was not
available. Betamethasone is structurally very similar to dexa-
methasone, differing only by the configuration of a methyl group
in position 16 [18]. The two substances have a high CNS pene-
tration and similar anti-inflammatory potency [19]. Dexametha-
sone was available in some clinics until 2007 despite not licensed
in Sweden since 2002. Thus, the two corticosteroids were used in
parallel until 2007, but since then, dexamethasone has not been
available in Sweden and therefore betamethasone has become the
sole adjuvant corticosteroid for the treatment of ABM.
The primary aim of this retrospective quality registry study was
to investigate the effects of adjuvant corticosteroids on mortality in
adult ABM. Secondary aims were to evaluate the effects on hearing
impairment and neurological deficits at follow-up and to compare
outcomes depending on whether dexamethasone or betametha-
sone was used.
The study comprised adult ABM patients (age 16 years) who
were registered in the national Swedish quality registry for ABM
during the period January 1995 to December 2014. Using conven-
tional diagnostic criteria, ABM diagnoses were set by local spe-
cialists in infectious diseases at each of the 32 Swedish infectious
disease clinics. The diagnoses were based on clinical criteria, CSF
analysis, and microbiological tests on blood and CSF as previously
described [3]. ABM was defined as community-acquired if the pa-
tients had not been hospitalized or had operations on the CNS
within 30 days before admission.
In the registry, sex, age, aetiology, mental status on admission,
corticosteroid and antibiotic treatment, door-to-antibiotic time, in-
hospital mortality, neurological sequelae, and hearing deficits are
routinely recorded. Mental status on admission was recorded in
956 patients, as the reaction level scale (RLS) [20] in 895 patients
and as the Glasgow coma scale (GCS) in an additional 61 patients
(Table 1). In the latter GCS was converted to RLS for standardized
comparison [21]. Adequate antibiotic treatment was defined as
intravenous b-lactam antibiotics for which the isolated bacteria
were sensitive according to susceptibility testing at local labora-
tories and administered in meningitis dosages. In patients with
unknown aetiology third-generation cephalosporin ± ampicillin or
meropenem was considered adequate. Neurological and hearing
deficits were registered at the follow-up 2e6 months after
discharge. Neurological sequelae were specified as disabling
headache, cognitive dysfunction/dementia, vertigo, or fatigue
causing limitations of daily activity, epileptic seizures, ataxia, or
persistent neurological deficits. Hearing disability was defined by
the patient as new onset of impairment, and audiometry was per-
formed when appropriate.
Adequate corticosteroid treatment was defined as dexametha-
sone 10 mg or betamethasone 8 mg every 6 hours intravenously,
initiated within 1 hour from the start of antibiotics. Two patients
who were given high doses (1 g) of methylprednisolone were also
considered adequately treated with corticosteroids. The duration of
corticosteroid treatment was not noted in the registry but the

Table 1
Characteristics of patients during the different study periods by groups with different corticosteroid treatment

Beta-methasone; Dexa-methasone; Corticosteroid No corticosteroid treatment; No data for corticosteroid

1995e2007 1995e2007 treatmenta; 1995e2014 (n ¼ 498) treatment; 1995e2014
(n ¼ 278) (n ¼ 243) 1995e2014 (n ¼ 259)
(n ¼ 989)

Median age (interquartile range) 58 (42e68) 59 (43e71) 58 (42e68) 65 (51e75) 59 (39e70)

Males/femalesb 147/131 113/128 486/501 236/259 84/76
RLS on admission; number with available data 110 40 611 232 113
RLS 1 (%) 29 (26.4) 5 (12.5) 211 (34.6) 103 (44.4) 36 (31.9)
RLS 2e3 (%) 53 (48.2) 25 (62.5) 290 (47.5) 101 (43.6) 48 (42.5)
RLS 4e8 (%) 28 (25.5) 10 (25.0) 110 (18.0) 28 (12.1) 29 (25.7)
Time from admission to start of antibiotics;
number with available data 244 206 856 394 182
1 hour (%) 88 (36.1) 61 (29.6) 324 (37.9) 75 (19.0) 53 (29.1)
>1 hour (%) 156 (63.9) 145 (70.4) 532 (62.1) 319 (81.0) 129 (70.9)
Aetiology
S. pneumoniae (%) 163 (58.6) 141 (58.0) 547 (55.3) 221 (44.4) 125 (48.3)
N. meningitidis (%) 31 (11.2) 28 (11.5) 118 (11.9) 46 (9.2) 34 (13.1)
Other/unknown bacteriac (%) 84 (30.2) 74 (30.5) 324 (32.8) 231 (46.4) 100 (38.6)

RLS ¼ reaction level scale. The RLS is as follows: 1, mentally alert; 2, drowsy or confused, responsive to light stimulation; 3, very drowsy or confused, responsive to strong
stimulation; 4, unconscious, localizes but does not ward off pain; 5, unconscious, withdrawing movements on pain; 6, unconscious, stereotype flexion on pain; 7, unconscious,
stereotype extension on pain; 8, unconscious, no response [20]. Conversion of Glasgow coma scale (GCS) to RLS is as follows: GCS 14e15 ¼ RLS 1, GCS 12e13 ¼ RLS 2, GCS
10e11 ¼ RLS 3, GCS 8e9 ¼ RLS 4, GCS 6 ¼ RLS 5, GCS 5 ¼ RLS 6, GCS 4 ¼ RLS 7, and GCS 3 ¼ RLS 8 [21].
a
Treatment with betamethasone (n ¼ 744), dexamethasone (n ¼ 243), or methylprednisolone (n ¼ 2).
b
The sex of the patient was not noted in 104 patients (two with dexamethasone treatment, three without corticosteroids, and 99 with no data for corticosteroids).
c
Streptococci (n ¼ 115), H. influenzae (n ¼ 98), L. monocytogenes (n ¼ 77), S. aureus (n ¼ 75), Enterobacteriacae (n ¼ 33), other bacteria (n ¼ 33), and unknown (n ¼ 224).
814.e3 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7
guideline recommendation is 2e4 days, which is in line with the
first published corticosteroid studies [8,9].
In the first Swedish guidelines in 2004, neuro-intensive care, as
previously described [22], was recommended in severe cases with
coma. In the guideline revision in 2009, impaired mental status and
new onset seizures were deleted as contraindications to perform
early lumbar puncture without a previous CT-scan [3]. Since 2008,
data on intensive care (70.1% of patients), mechanical ventilation
(31.7%), and neuro-intensive care (20.8) have been recorded in the
registry.
Statistics
The main analysis on mortality was performed on the whole
patient material during the 20-year period (1995e2014). The
comparison between betamethasone and dexamethasone was
restricted to patients from 1995e2007 (Fig. 1). A two-tailed Chi2-
test was used for categorical data. Missing data (age 0.6%, gender
6.0%, RLS on admission 45.2%, time from admission to start of
treatment 18.0%, and corticosteroid treatment 14.8%) were imputed
using chained equations as described by Royston et al., generating
50 imputation sets [23]. Logistic regression analyses were per-
formed with odds ratios (ORs) and confidence intervals (CIs)
calculated using Rubin's method [24]. Multivariate analyses were
adjusted for sex, age (five categories: age 45, 45e54, 55e64,
65e74, and 75), mental status on admission (three categories: RLS
1, 2e3, and 4) and time to adequate antibiotic treatment (three
categories: 1 h, 1e2 h, and >2 h). Stratified analyses were per-
formed for different aetiologies and calendar periods according to
changes in the Swedish treatment guidelines (2004 and 2009),
mental status on admission, and door-to-antibiotic time. STATA,
version 13.0, was used for statistical analysis.
Ethics
The study was approved by the ethics committee at Karolinska
Institutet (Dnr 2015/756e32). The patients were informed that
clinical data were registered in the registry, which could be applied
anonymously for research purposes.
Fig. 1. Annual proportion (%) of adult patients with community-acquired bacterial meningitis treated with corticosteroids (betamethasone or dexamethasone) administered
concomitant with the start of antibiotics in 1995 to 2014. The annual number (n) of patients is shown on the x-axis. In addition, two patients, one in 1996 and one in 1997, were
treated with methylprednisolone.
Results
During the 20-year period, 1832 adults with community-
acquired ABM were registered, representing about 70% of all
adult cases with ABM in Sweden according to the Swedish National
Board of Health and Welfare (personal communication). Seventy
patients without mortality data, 14 treated with low doses of hy-
drocortisone for septic shock, and two with uncertainty about type
of corticosteroid treatment were excluded. Hence, 1746 patients
were included in the final analysis, of whom 989 were adequately
treated with corticosteroids, betamethasone n ¼ 744, dexametha-
sone n ¼ 243, and methylprednisolone n ¼ 2 (Table 1). Cortico-
steroids were not given to 498 patients, and in 259 patients data for
corticosteroid treatment were missing. Patients with missing
corticosteroid data did not differ notably in demographic charac-
teristics from those with corticosteroid data (Table 1).
The overall mortality during hospital stay was 214 (12.3%) of the
1746 cases and varied during the 20-year study period as follows:
1995e1999, 62/418 (14.8%); 2000e2004, 65/414 (15.7%);
2005e2009, 49/369 (13.3%); and 2010e2014, 38/545 (7.0%).
Corticosteroids vs. no corticosteroids
The use of betamethasone and dexamethasone during the study
period is shown in Fig. 1. Mortality in the total material with data on
corticosteroid treatment and mortality (n ¼ 1487) was significantly
lower in the corticosteroid-treated group (8.9%) than in the non-
corticosteroid-treated group (17.9%, Fig. 2a). Stratified mortality
analyses in corticosteroid treated vs. non-corticosteroid treated
patients according to time periods with different guidelines were as
follows: 41/368 (11.1%) vs. 58/281 (20.6%) in 1995e2004; 28/252
(11.1%) vs. 15/72 (20.8%) in 2005e2009; and 19/369 (5.1%) vs. 16/
145 (11.0%) in 2010e2014.
Without adjustment for confounders, lower mortalities were
seen not only in the patients with S. pneumoniae (10.2% vs. 21.3%,
p <0.001), but also in the patients with other specified bacteria (27/
330 ¼ 8.2% vs. 33/212 ¼ 15.6%, p <0.01) and in those with unknown
aetiology (5/112 ¼ 4.5% vs. 9/65 ¼ 13.8%, p <0.05). Among the pa-
tients with specified bacterial aetiology, no effect of corticosteroids
 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7 814.e4

Table 2
Multivariate analyses of corticosteroid treatment in relation to mortality (patients
not treated with corticosteroids were used as reference)

Adjusted odds ratioa (95% CI)

No corticosteroid Corticosteroid
treatment treatment

All 1 (ref) 0.57 (0.40e0.81)

Microbiological agent
S. pneumoniae 1 (ref) 0.50 (0.31e0.80)
Other bacteria or unknownb 1 (ref) 0.71 (0.40e1.26)
Time periodc
1995e2004 1 (ref) 0.53 (0.33e0.87)
2005e2009 1 (ref) 0.57 (0.25e1.31)
2010e2014 1 (ref) 0.66 (0.30e1.48)
Reaction level scale
1 1 (ref) 0.78 (0.28e2.21)
2e3 1 (ref) 0.52 (0.30e0.89)
4 1 (ref) 0.54 (0.28e1.07)
Time to given antibioticsd
1 hour 1 (ref) 0.40 (0.18e0.88)
>1 hour 1 (ref) 0.60 (0.40e0.89)
Corticosteroid typee
Betamethasone 1 (ref) 0.49 (0.28e0.84)
Dexamethasone 1 (ref) 0.61 (0.37e1.01)
a
Adjustment for sex, age, time to start of antibiotic treatment, and reaction level
scale, except in stratified analyses according to reaction level scale where adjust-
ment was made for sex, age, and time to start of antibiotics, and in analyses ac-
cording to time to start of antibiotics where adjustment was made for sex, age, and
reaction level scale.
b
629 with known aetiology: N. meningitidis (n ¼ 198), Streptococci (n ¼ 115).
H. influenzae (n ¼ 98), L. monocytogenes (n ¼ 77), S. aureus (n ¼ 75), Enterobacteriacae
(n ¼ 33), other bacteria (n ¼ 33), and 224 patients with unknown aetiology.
c
Periods defined according to when the Swedish treatment guidelines for bac-
terial meningitis were updated.
d
Time from arrival to hospital to administration of adequate antibiotics.
e
Analyses restricted to 1994e2007 and excluded two patients that received
methylprednisolone.

Favourable outcome in terms of survival and recovery without

Fig. 2. Mortality during hospital stay (a) and sequelae (hearing, neurological deficits,
or both) among survivors at 2e6-month follow-up (b) in adult patients with
community-acquired bacterial meningitis of different aetiologies treated in 1995e2014
with or without corticosteroids.
sequelae was observed in similar or higher proportions in
corticosteroid-treated compared with non-corticosteroid-treated
patients in all the major aetiological groups (S. pneumoniae: 45.2%
vs. 35.1%, p <0.05; N. meningitidis; 68.6% vs. 58.1%; H. influenzae:
52.6% vs. 52.2%; streptococci/staphylococci: 52.6% vs. 46.5%), except
in cases with L. monocytogenes (40.0% vs. 48.5%). Apart from
pneumococcal meningitis, these differences were not statistically
significant.

was observed in those with L. monocytogenes (8/33 ¼ 24.2% vs. 8/
37 ¼ 21.6%, p ¼ 0.78).
The corticosteroid-treated patients were younger and started on
antibiotic treatment earlier but were more comatose than the non-
corticosteroid-treated patients (Table 1). However, mortality
adjusted for sex, age, door-to-antibiotic time, and mental status on
admission still resulted in an OR of 0.57 (95% CI 0.40e0.81) in the
total material (Table 2). In cases with pneumococcal meningitis, the
adjusted OR for death was 0.50 (95% CI 0.31e0.80). However, after
adjustment for confounders, the reduction in mortality in patients
with non-pneumococcal aetiology did not reach statistical signifi-
cance, with an adjusted OR of 0.71 (95% CI 0.40e1.26). Reduction in
adjusted mortality was virtually similar over different time periods
(Table 2).
Among the 1310 survivors, data were available for hearing and
neurological sequelae at follow-up in 1166 patients with similar
outcomes irrespective of whether corticosteroids were given
(Fig. 2b). Hearing deficit was noted in 24.8% of survivors in the
corticosteroid treated-group vs. 21.6% (p ¼ 0.24) in those without
corticosteroids.
Corticosteroids vs. no corticosteroids in relation to mental status
Data for mental status on admission and corticosteroid
treatment were available in 843 patients. The effect of cortico-
steroids on mortality in relation to mental status at admission is
depicted in Fig. 3. The adjusted analysis showed decreased
mortality with corticosteroid treatment for the group with
moderately impaired mental status (RLS 2e3) and a trend to
lower mortality in the comatose patients (RLS 4e8), whereas
there was no notable effect in patients without impaired mental
status (RLS 1) (Table 2).
Corticosteroids vs. no corticosteroids in relation to time to treatment
Data for corticosteroid treatment and door-to-antibiotic time
were reported in 1250 patients. After adjustment for potential
confounders, significantly lower mortality was observed in the
corticosteroid group treated with antibiotics <1 hour from admis-
sion (OR 0.40, 95% CI 0.18e0.88), as well as in those treated later OR
0.60, 95% CI 0.40e0.89).
814.e5 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7
Fig. 3. Outcomes related to mental status on admission. Mortality during hospital stay (a) and sequelae (hearing, neurological deficits, or both) at 2e6-month follow-up (b) in adult
patients treated in 1995e2014 with or without corticosteroids. RLS, reaction level scale.
Betamethasone vs. dexamethasone
During 1995e2007, data for type of corticosteroid and mortality
were available in 521 patients. Mortality was 13.2% and 9.7%
(p ¼ 0.2) in dexamethasone- and betamethasone-treated patients,
respectively. Adjusted analysis revealed that the effect of betame-
thasone was similar or better than that of dexamethasone (Table 2).
In the surviving patients, hearing or neurological sequelae were
more often observed in dexamethasone-treated than in
betamethasone-treated patients: 109/198 (55.1%) vs. 106/238
(44.5%, p <0.05).
Discussion
Despite modern antibiotic treatment and intensive care, mor-
tality and morbidity still remain high in ABM [1]. Because increased
intracranial pressure (ICP) is the underlying mechanism for most
deaths in ABM, different measures to counteract elevated ICP
should have the greatest potential to improve outcome. The path-
ophysiological mechanisms resulting in increased ICP in ABM are
multifactorial. Release of bacterial components in the subarachnoid
space leads to an inflammatory response that contributes to
increased permeability of the blood-brain barrier causing cerebral
extracellular oedema, impaired CSF-absorption with increased CSF-
volume, a cytotoxic intracellular brain oedema, and increased ce-
rebral blood flow (hyperaemia); all adding to elevated ICP [25].
Glycerol causes an osmotic pressure gradient that can counteract
brain oedema, but the results from clinical studies are conflicting
[26]. For severe cases of ABM with impending cerebral herniation,
neuro-intensive care with ICP monitoring and CSF drainage may be
beneficial [22,27].
High-dose corticosteroid treatment is the most studied adjuvant
ABM-treatment with widespread adoption in clinical practice.
Corticosteroids have been shown to reduce mortality and neuro-
logical sequelae in randomized controlled trials in Europe but have
failed to do so in several large studies performed in Africa, South
America, and Asia [8e13]. Aggregate data from available high-
quality studies support the recommendation to use corticoste-
roids in high-income countries based on a decrease of neurological
sequelae [14e16]. However, a reduction in mortality could not be
established [14].
To the authors' knowledge, the present quality registry study
over 20 years is the largest investigation of the efficacy of cortico-
steroids in community-acquired ABM in adults. The risk of fatal
outcome was halved in patients with corticosteroid treatment in
comparison with those without such treatment (8.9% vs. 17.9%).
Despite falling overall mortality over time, the decreased mortality
in the corticosteroid-treated patients remained virtually the same.
Lower age and earlier antibiotic treatment were observed in the
corticosteroid-treated patients, but these patients had worse
mental status on admission, resulting in an almost identical
reduction in mortality in the adjusted analysis. These results agree
with those from the randomized placebo-controlled study by de
Gans et al., which included 301 patients [9]. As in that study, the
present study found that the beneficial effect of corticosteroids was
most evident in patients with pneumococcal meningitis, and that
corticosteroids do not seem to increase sequelae in survivors.
Although not reaching statistical significance in the adjusted
analysis, the present results indicate that corticosteroids may be
beneficial rather than harmful in adults with ABM caused by
N. meningitidis, other bacteria, or with unknown aetiology (Fig. 2a,
Table 2). These results are also in agreement with earlier studies
[1,9,17,28] in which trends towards a positive effect were seen in
patients with ABM caused by bacteria other than S. pneumoniae. In
a recent prospective cohort study the effect of dexamethasone
decreased mortality in patients with pneumococcal and non-
pneumococcal aetiology [1]. However, the study was not
designed to specifically study corticosteroid treatment and the
proportion of patients not treated with corticosteroids constituted
only 11%. In the present study L. monocytogenes was the only
pathogen for which a trend towards negative outcome was
observed in the corticosteroid-treated group. Thus, no support was
found for corticosteroid treatment in Listeria meningitis. However,
the power of the study was not sufficient to prove or disprove a
positive or negative effect in different specific aetiologies, apart
from S. pneumoniae.
The frequency of sequelae was similar whether corticosteroids
were given or not, indicating that the reduced mortality did not
 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7
result in increased sequelae in survivors at the follow-up
2e6 months after discharge, a finding consistent with those re-
ported in short- and long-term follow-up in adult ABM [1,29].
The present results suggest that corticosteroids are beneficial in
patients with impaired mental status (RLS 2), whereas a similar
low mortality of about 3% was observed in mentally unaffected
cases (RLS 1), regardless of corticosteroid treatment. These results
are in accordance with the findings of de Gans et al. [9]. Based on
160 patients with GCS 12e14 in that study and 314 with RLS 1 in the
present study, these results indicate that the beneficial effect is
limited in mentally alert patients, and consequently, that the sup-
port for adjunctive corticosteroid treatment in this patient group
may be questioned.
The findings of the present study, in conjunction with others,
strongly support empirical corticosteroid treatment in mentally
affected patients with ABM, and that treatment should not be
discontinued in cases in which S. pneumoniae is not found. How-
ever, corticosteroid treatment should be considered in all ABM
cases, not only in cases with impaired mental status, because timely
adequate treatment is pivotal for a favourable outcome [3,4],
deterioration may occur rapidly after presentation and treatment
start at admission [30,31], and corticosteroids are shown to be
effective only if given together with the first doses of antibiotics.
Unambiguous clinical guidelines are also important in a fulminant
disease such as ABM. The importance of timely management of
corticosteroids has not yet been fully assessed other than that this
may be beneficial when administered together with the start of
antibiotics. This study indicates that corticosteroids retain their
effect even in patients in whom antibiotic treatment has been
delayed for various reasons.
The optimal duration of corticosteroid treatment has not been
thoroughly investigated. A 4-day regimen is usually recommended
simply because this has been practiced in most studies [14]. How-
ever, in a study on children 2-day treatment was as effective as
4 days [32]. With adequate antibiotic treatment, the elevated ICP
most probably peaks very early in the course of a disease [22,27]
and declines after 24e48 hours. Thus, a shorter duration than
4 days may be appropriate, at least in cases with rapid clinical
improvement. The results of the present study also indicate that
corticosteroids should probably be withdrawn if L. monocytogenes
are detected as the causative bacteria. In this study the duration of
corticosteroid treatment was not defined, but 2e4 days has been
the recommendation in Sweden since 2004.
Most studies have involved dexamethasone, and the effect of
other corticosteroids in ABM is considered to be an important
unanswered issue [14]. Dexamethasone has a high penetration into
CSF and long half-life [33] and has therefore been regarded as the
glucocorticoid of choice for ABM-treatment. Betamethasone is a
diastereoisomer of dexamethasone with a slightly higher anti-
inflammatory potency, comparable CSF penetration, and some-
what longer half-life in CSF [34]. With the dosage chosen, it can thus
be assumed that betamethasone and dexamethasone would have
equivalent effects in ABM. To the authors' knowledge, this is the first
study to compare the clinical efficacy of these different corticoste-
roids in adults with ABM. As in the prevention of neonatal respira-
tory morbidity, where both betamethasone and dexamethasone
have shown beneficial effects [35,36], the results of the present
study strongly indicate that betamethasone equals dexamethasone
with respect to mortality and sequelae in ABM. Increased use of
corticosteroids, i.e. betamethasone, could have contributed to the
decreased mortality (7%) observed during 2010e2014. However,
earlier antibiotic treatment [3] and increased use of neuro-intensive
care [22] were probably also important factors.
The main strength of this study, namely being the evaluation of
corticosteroids for ABM in real clinical practice, is also the study's
814.e6
major limitation, i.e. its retrospective non-interventional design
with lack of randomization and control of the original data. A risk of
confounding by indication must be considered but adjustment for
important confounding variables was performed. The large amount
of missing data on mental status is a drawback, but imputation
analyses were used to compensate for this. Furthermore, the
number of corticosteroid-treated patients increased over time,
especially after the publication of the study by de Gans et al. in 2002
[9], along with a decrease in mortality, probably caused by imple-
mentation of advanced neuro-intensive care and new recommen-
dations aiming at prompt lumbar puncture associated with earlier
treatment [3,22]. However, regardless of the time period, the
adjusted OR for a treatment effect remained in the interval of 0.53
to 0.66. Finally, adjustment for immunocompromised states and
important clinical measures of disease severity, such as septic shock
and seizures, would have been desirable but these data were not
available in the registry.
An important strength of the study is the large sample size that
enabled consideration of several important confounders, such as
age, aetiology, calendar time, mental status, and time to treatment
by stratified and multivariate analyses, which thereby consolidate
the results. The outcome was consistent and generally pointed at
favourable effects of corticosteroids vs. no corticosteroids in these
analyses, implicating a high external validity of the present
findings.
Conclusion
The present study increases the strength of evidence indicating
that corticosteroid treatment decreases ABM mortality in resource-
rich countries and should be administered initially together with
the start of antibiotics in mentally affected adult patients with
community acquired ABM, regardless of presumed aetiological
agent. The results further indicate that betamethasone is equally as
effective as dexamethasone. The value and importance of registry
studies, especially in clinical situations where randomized trials are
unlikely to take place, is demonstrated.
Transparency declaration
No conflict of interest. The Swedish Quality Registry for bacterial
meningitis is funded by the Swedish Association of Local Author-
ities and Regions.
References
[1] Bijlsma MW, Brouwer MC, Kasanmoentalib ES, Kloek AT, Lucas MJ, Tanck MW,
et al. Community-acquired bacterial meningitis in adults in the Netherlands,
2006e14: a prospective cohort study. Lancet Infect Dis 2016;16:339e47.
[2] van de Beek D, de Gans J, Tunkel AR, Wijdicks EFM. Community-acquired
bacterial meningitis in adults. New Engl J Med 2006;354:44e53.
[3] Glimåker M, Johansson B, Grindborg O, € Bottai M, Lindquist L, Sjolin J. Adult
bacterial meningitis: earlier treatment and improved outcome following
guideline revision promoting prompt lumbar puncture. Clin Infect Dis
2015;60(8):1162e9.
[4] Koster-Rasmussen R, Korshin A, Meyer CN. Antibiotic treatment delay and
outcome in acute bacterial meningitis. J Infect 2008;57:449e54.
[5] Tauber MG, Khayam-Bashi H, Sande MA. Effects of ampicillin and corticoste-
roids on brain water content, cerebrospinal fluid pressure, and cerebrospinal
fluid lactate levels in experimental pneumococcal meningitis. J Infect Dis
1985;151:528e34.
[6] Lutsar I, Friedland IR, Jafri HS, Wubbel L, Ahmed A, Trujillo M, et al. Factors
influencing the anti-inflammatory effect of dexamethasone therapy in
experimental pneumococcal meningitis. J Antimicrob Chemother 2003;52:
651e5.
[7] Sapolsky RM, Pulsinelli WA. Glucocorticoids potentiate ischemic injury to
neurons: therapeutic implications. Science 1985;229:1397e400.
[8] McIntyre PB, Berkey CS, King SM, Schaad UB, Kilpi T, Kanra GY, et al. Dexa-
methasone as adjunctive therapy in bacterial meningitis. A meta-analysis of
randomized clinical trials since 1988. JAMA 1997;278:925e31.
814.e7 M. Glimåker et al. / Clinical Microbiology and Infection 22 (2016) 814.e1e814.e7
[9] de Gans J, van de Beek D. European Dexamethasone in Adulthood Bacterial
Meningitis Study I. Dexamethasone in adults with bacterial meningitis. New
Engl J Med 2002;347:1549e56.
[10] Molyneux EM, Walsh AL, Forsyth H, Tembo M, Mwenechanya J, Kayira K, et al.
Dexamethasone treatment in childhood bacterial meningitis in Malawi: a
randomised controlled trial. Lancet 2002;360:211e8.
[11] Nguyen TH, Tran TH, Thwaites G, Ly VC, Dinh XS, Ho Dang TN, et al. Dexa-
methasone in vietnamese adolescents and adults with bacterial meningitis.
N Engl J Med 2007;357:2431e40.
[12] Peltola H, Roine I, Ferna ndez J, Zavala I, Ayala SG, Mata AG, et al. Adjuvant
glycerol and/or dexamethasone to improve the outcomes of childhood bac-
terial meningitis: a prospective, randomized, double-blind, placebo-
controlled trial. Clin Infect Dis 2007;45:1277e86.
[13] Scarborough M, Gordon SB, Whitty CJ, French N, Njalale Y, Chitani A, et al.
Corticosteroids for bacterial meningitis in adults in sub-Saharan Africa. N Engl
J Med 2007;357:2441e50.
[14] Brouwer MC, McIntyre P, Prasad K, van de Beek D. Corticosteroids for acute
bacterial meningitis. Cochrane Database Syst Rev 2015:CD004405.
[15] van de Beek D, Cabellos C, Dzupova O, Esposito S, Klein M, Kloek AT, et al.
ESCMID guideline: diagnosis and treatment of acute bacterial meningitis. Clin
Microbiol Infect 2016;22(Suppl. 3):S37e62.
[16] McGill F, Heyderman RS, Michael BD, Defres S, Beeching NJ, Borrow R, et al.
The UK joint specialist societies guideline on the diagnosis and management
of acute meningitis and meningococcal sepsis in immunocompetent adults.
J Infect 2016;72:405e38.
[17] Heckenberg SG, Brouwer MC, van der Ende A, van de Beek D. Adjunctive
dexamethasone in adults with meningococcal meningitis. Neurology
2012;79:1563e9.
[18] Bar-Lev MR, Maayan-Metzger A, Matok I, Heyman Z, Sivan E, Kuint J. Short-
term outcomes in low birth weight infants following antenatal exposure to
betamethasone versus dexamethasone. Obstet Gynecol 2004;104:484e8.
[19] Trenque. Adrenal corticoid steroids. Drug facts and comparisons. St. Louis,
Mo.: Facts and Comparisons; 1997. p. 122e8.
[20] Starmark JE, Stalhammar D, Holmgren E. The Reaction Level Scale (RLS85).
Manual and guidelines. Acta Neurochir 1988;91:12e20.
[21] Starmark JE, Stalhammar D, Holmgren E, Rosander B. A comparison of the
Glasgow Coma Scale and the Reaction Level Scale (RLS85). J Neurosurg
1988;69:699e706.
[22] Glimåker M, Johansson B, Halldorsdottir H, Wanecek M, Elmi-Terander A,
Ghatan PH, et al. Neuro-intensive treatment targeting intracranial hyperten-
sion improves outcome in severe bacterial meningitis: an intervention-
control study. PLoS One 2014;9:e91976.
[23] Sauerbrei W, Royston P, Schumacher M. Comments on 'Performance of using
multiple stepwise algorithms for variable selection' by Ryan E. Wiegand. Statistics
in Medicine 2010;29:1647e59. Stat Med 2011;30:892e4; author reply 894e5.
[24] Rubin DB. Multiple imputation for nonresponse in surveys. John Wiley &
Sons; 1987.
[25] Scheld WM, Koedel U, Nathan B, Pfister HW. Pathophysiology of bacterial
meningitis: mechanism(s) of neuronal injury. J Infect Dis 2002;186(Suppl. 2):
S225e33.
[26] Wall EC, Ajdukiewicz KM, Heyderman RS, Garner P. Osmotic therapies added
to antibiotics for acute bacterial meningitis. Cochrane Database Syst Rev
2013;3:CD008806.
[27] Lindvall P, Ahlm C, Ericsson M, Gothefors L, Naredi S, Koskinen LO. Reducing
intracranial pressure may increase survival among patients with bacterial
meningitis. Clin Infect Dis 2004;38:384e90.
[28] Costerus JM, Brouwer MC, van der Ende A, van de Beek D. Repeat lumbar
puncture in adults with bacterial meningitis. Clin Microbiol Infect 2016;22:
428e33.
[29] Weisfelt M, Hoogman M, van de Beek D, de Gans J, Dreschler WA,
Schmand BA. Dexamethasone and long-term outcome in adults with bacterial
meningitis. Ann Neurol 2006;60:456e68.
[30] Aronin SI, Peduzzi P, Quagliarello VJ. Community-acquired bacterial menin-
gitis: risk stratification for adverse clinical outcome and effect of antibiotic
timing. Ann Intern Med 1998;129:862e9.
[31] Radetsky M. Fulminant bacterial meningitis. Pediatr Infect Dis J 2014;33:
204e7.
[32] Syrogiannopoulos GA, Lourida AN, Theodoridou MC, Pappas IG, Babilis GC,
Economidis JJ, et al. Dexamethasone therapy for bacterial meningitis in chil-
dren: 2- versus 4-day regimen. J Infect Dis 1994;169:853e8.
[33] Balis FM, Lester CM, Chrousos GP, Heideman RL, Poplack DG. Differences in
cerebrospinal fluid penetration of corticosteroids: possible relationship to the
prevention of meningeal leukemia. Journal Clin Oncol 1987;5:202e7.
[34] Trenque T, Lamiable D, Vistelle R, Millart H, Leperre A, Choisy H. Comparative
pharmacokinetics of two diastereoisomers dexamethasone and betametha-
sone in plasma and cerebrospinal fluid in rabbits. Fundam Clin Pharmacol
1994;8:430e6.
[35] Nada AM, Shafeek MM, El Maraghy MA, Nageeb AH, Salah El Din AS,
Awad MH. Antenatal corticosteroid administration before elective caesarean
section at term to prevent neonatal respiratory morbidity: a randomized
controlled trial. Eur J Obstet Gynecol Reprod Biol 2016;199:88e91.
[36] Gyamfi-Bannerman C, Thom EA, Blackwell SC, Tita AT, Reddy UM, Saade GR,
et al. Antenatal betamethasone for women at risk for late preterm delivery.
N Engl J Med 2016;374:1311e20.