bs_bs_banner

ORIGINAL ARTICLE

Systemic corticosteroids for community-acquired pneumonia:

Reasons for use and lack of benefit on outcome

EVA POLVERINO,1 CATIA CILLÓNIZ,1 POVILAS DAMBRAVA,2 ALBERT GABARRÚS,1 MIQUEL FERRER,1
CARLOS AGUSTÍ,1 ELENA PRINA,4 BEATRIZ MONTULL,3 ROSARIO MENENDEZ,3
MICHAEL S. NIEDERMAN5,6 AND ANTONI TORRES1

1
Respiratory Disease Department, Hospital Clínic i Provincial de Barcelona, IDIBAPS, CIBER de Enfermedades Respiratorias
(Ciberes), 2Respiratory and Sleep Disorders Unit, Hospital of Vendrell, Barcelona, 3Respiratory Disease Department, Hospital
La Fe, Valencia, Spain, 4Emergency Care Department, Policlinic – University of Milan, Milan, Italy, 5Department of Medicine,
Winthrop University Hospital, Mineola and 6SUNY, Stony Brook, New York, USA

ABSTRACT
Background and objective: Although the benefits
of systemic corticosteroids in community-acquired
pneumonia (CAP) are not clear, their use is frequent in
clinical practice. We described the frequency of this
practice, patients’ characteristics and its clinical
impact.
Methods: We investigated all adult CAP patients
visited between June 1997 and January 2008 (n = 3257).
Results: Two hundred and sixty patients received sys-
temic corticosteroids (8%) with a mean daily dose of 45
(33) mg (median, 36 mg/day). Patients receiving corti-
costeroids were older (74 (13) vs 65 (19) years), had
more comorbidities (respiratory, 59% vs 38%, cardiac,
29% vs 16%, etc.), higher Pneumonia Severity Index
(Fine IV–V, 76% vs 50%) and had received inhaled cor-
ticosteroids (36% vs 15%) and previous antibiotics
(31% vs 23%) more frequently (P < 0.01, each). Signifi-
cant predictors of corticosteroid administration were:
chronic obstructive pulmonary disease (odds ratio
(OR), 1.91), fever (OR, 0.59), expectoration (OR, 1.59),
creatinine (+1 mg/dL, OR, 0.92), SaO2 ⱖ 92% (OR, 0.46),
C-reactive protein (+5 mg/dL; OR, 0.92) and cardiac
failure (OR, 1.76). Mortality (6% vs 7%; P = 0.43) and
time to clinical stability (4 (3–6) vs 5 (3–7) days;
P = 0.11) did not differ between the two groups, while
length of hospital stay was longer for the steroid group
(9 (6–14) vs 6 (3–9) days; P < 0.01).
Conclusions: The main reasons for administering
systemic steroids were the presence of chronic respira-
tory comorbidity or severe clinical presentation, but
therapy did not influence mortality or clinical stability;
by contrast, steroid administration was associated with
prolonged length of stay. Nevertheless the steroid
group did not show an increased mortality as it was
expected according to the initial Pneumonia Severity
Correspondence: Antoni Torres, Department of Pneumology,
Hospital Clinic, Villarroel 170, Barcelona, Spain. Email: atorres@
clinic.ub.es
Received 29 May 2012; invited to revise 26 June 2012; revised
9 August 2012; accepted 25 August 2012 (Associate Editor:
Marcos Restrepo).
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
SUMMARY AT A GLANCE
Despite the lack of a clear benefit, systemic corti-
costeroids are frequently administered in CAP. We
investigated clinical reasons for steroid prescrip-
tion and the impact on major outcomes, showing
no correlation with mortality or clinical stability
independent of severity and cumulative steroid
dosage, but an increase in length of hospital stay.
Index score. Influence of steroids on outcomes of CAP
need to be further investigated through randomized
clinical trial.
Key words: adjunctive therapy, clinical stability, community-
acquired pneumonia, mortality, systemic corticosteroid.
Abbreviations: CAP, community-acquired pneumonia; COPD,
chronic obstructive pulmonary disease; IQR, interquartile range;
LOS, length of stay; OR, odds ratio; PSI, Pneumonia Severity
Index.
INTRODUCTION
The role of corticosteroids as an adjunctive therapy
for community-acquired pneumonia (CAP) is still
under debate. Their potential beneficial use is based
on the rationale of their anti-inflammatory effect and
the consequent influence on pneumonia severity and
outcomes. Indeed, it has been shown that the admin-
istration of systemic corticosteroids is associated with
reduced pulmonary inflammation in patients with
bacterial pneumonia.1
Some clinical trials testing the usefulness of sys-
temic steroids as adjunctive therapy of severe CAP
have described some benefits by decreasing mortality
or length of stay (LOS),2–6 while others show no
Respirology (2013) 18, 263–271
doi: 10.1111/resp.12013
264
benefits in the same end-points or even an increase in
related side-effects (e.g. hyperglycaemia, superinfec-
tions).7,8 A recent meta-analysis does not support the
recommendation of corticosteroids as a standard of
care for patients with severe CAP, due to the lack of
strong evidence.9
Despite providing a clear benefit, the use of corti-
costeroids in CAP is relatively frequent in practice,
possibly in relation to patients’ comorbidities (respi-
ratory diseases, cardiac or renal failure for example)
or to clinical presentation (wheezing or rhonchi).
The aim of this study was to investigate the clinical
conditions associated with corticosteroids prescrip-
tion in CAP and, secondly, to analyse their impact on
CAP outcomes.
METHODS
Study participants
We performed a prospective observational study on
adult patients admitted to the Hospital Clinic, Bar-
celona, Spain, with CAP10 between June 1997 and
January 2008. Patients were excluded from the study
if one of the following criteria applied: severe immu-
nosuppression (HIV, immunosuppressants including
chronic systemic corticosteroids, current or previous
alcohol intake >80 g/day, active neoplasia), active
tuberculosis and patients with a confirmed alterna-
tive diagnosis at follow up. Nursing home patients
were not excluded from the analysis. Patients not
hospitalized were visited after 7–15 days in the out-
patient clinic.
Data collection
Data on age, gender, smoking and alcohol habits
(80 g/day), previous antibiotic therapy, prior influ-
enza and pneumococcal vaccinations, comorbidities
such as chronic obstructive pulmonary disease
(COPD), and cardiac, liver, renal or central nervous
system disorders were collected. Main clinical signs
and symptoms (cough, expectoration, fever, respira-
tory and heart rates, etc.) were recorded at admis-
sion, as well as analytical data (leucocyte count,
C-reactive protein, arterial blood gases, etc.), Pneu-
monia Severity Index (PSI) and CURB-65 (Confusion,
Blood Urea, Respiratory Rate, Blood Pressure, 65
years old) scores,11,12 and empirical antimicrobial
treatment. Surviving patients underwent radiological
and serological follow up after 30 days.
The prevalence of pulmonary (pleural effusion,
empyema, atelectasis, cavitations and respiratory
distress) and extrapulmonary complications of
pneumonia (septic shock, acute renal failure, endo-
carditis, meningitis, cardiac arrhythmias, syndrome
of inappropriate antidiuretic hormone secretion,
diarrhoea, myocardial infarction, cardiac arrest, pan-
creatitis) were also recorded, as well as 30-day mor-
tality rate and clinical stability. Clinical stability was
defined as the number of days from admission when
the patient met all of the following criteria: (i) tem-
perature <37.8°C; (ii) heart rate <100 bpm; (iii) respi-
ratory rate <24 breaths/min; (iv) systolic blood
Respirology (2013) 18, 263–271
E Polverino et al.
pressure >90 mm Hg; and (v) recovered respiratory
insufficiency (arterial oxygen saturation > 90% or
arterial oxygen tension (PaO2) > 60 mm Hg on room
air or basal arterial oxygen saturation similar to
usual values in stable conditions).
The prospective collection of clinical data was
approved by the Institutional Review Board. Patients’
identification remained anonymous and informed
consent was considered unnecessary due to the
observational nature of the study. All reported data
are the result of the clinical routine activity and all
tests and procedures were ordered by the attending
physicians, not involved in this study.
Systemic therapy with corticosteroids
Therapy with systemic corticosteroids (oral or intra-
venous) was at the discretion of the physician and not
randomized. We recorded daily doses of methyl-
prednisolone or equivalent doses of other corticoster-
oids (prednisone, hydrocortisone, etc.) during the
first 7 days after admission and the cumulative 7-day
dose. Patients receiving corticosteroid treatment for
more than 7 days were also recorded.
Microbiological investigations
Routine sampling to determine the microbial aetiol-
ogy of pneumonia included sputum, urine for Strep-
tococcus pneumoniae and Legionella pneumophila
antigens, blood samples (culture and serology tests)
and nasopharyngeal swabs. Pleural puncture, tra-
cheobronchial aspirates and bronchoalveolar lavage,
when available, were collected for Gram and Ziehl–
Nielsen stains and for cultures for bacterial, fungal
and mycobacterial pathogens.
Statistical analysis
Categorical variables were reported as frequencies
and percentages, while continuous variables were
reported as mean (standard deviation) or median and
interquartile range (IQR) for data that were not
normally distributed (Kolmogorov–Smirnov test).
Categorical variables were compared using the chi-
square test or Fisher’s exact test where appropriate.
Continuous variables were compared using the t-test
or the non-parametric Mann–Whitney U according
to data distribution. Correlation analyses were per-
formed using the Spearman’s rank correlation. Uni-
variate and multivariate logistic regression analyses
were performed to identify variables predictive of (i)
corticosteroid administration; (ii) mortality; or (iii)
LOS (dependent variables). Variables that showed a
significant result in the univariate analysis (P < 0.1)
were included in the corresponding multivariate
logistic regression backward stepwise model. Multi-
variate analyses for mortality and LOS were adjusted
for PSI, multilobar infiltration and previous antibiotic
therapy.
The Hosmer–Lemeshow goodness-of-fit test was
performed to assess the overall fit of the models [10].
All tests were two-tailed and significance was estab-
lished at 5% (SPSS version 16.0 for Windows, SPSS
Inc., Chicago, IL, USA).
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
Corticosteroids in CAP 265

RESULTS during the first 7 days. In 68% of these patients (n,

A total number of 4549 patients with CAP were seen in
the hospital during the study period, 3257 were
included in the analysis according to inclusion/
exclusion criteria (Fig. 1). Systemic corticosteroids
were administered in 260 patients (8% of total) in
addition to standard antimicrobial therapy. The mean
daily dose of methyl-prednisolone or equivalents was
of 45 (30) mg (median, 36 mg/day; IQR, 27–51 mg)
176) corticosteroids were administered for more than
1 week.
Main demographic characteristics of the popula-
tions (steroid and non-steroid groups) are shown in
Table 1.
Systemic corticosteroids were preferentially admin-
istered in older patients with more comorbidities,
aspiration risk and higher PSI (Tables 1,2). In particu-
lar, corticosteroid-treated patients showed a higher

4549 patients visited the

Hospital for suspicion of
pneumonia

Recruited patients: Excluded patients:

3257 1292

260 patients (8%) treated with 2997 (92%) patients not

systemic steroids at admission receiving systemic steroids

176 (68%) patients received

Figure 1 Patients’ profile for study steroids for >7 days
recruitment.

Table 1 Main demographic characteristics of the study populations according to steroid administration

Non-steroid Steroid P-value

Number, % 2997 (92) 260 (8) —

Age (years), mean (SD) 65 (20) 74 (13) <0.01
Males, n (%) 1690 (58) 158 (61) 0.33
Smoking habit <0.01
Active smokers, n (%) 707 (24) 39 (15) <0.01
Ex-smokers, n (%) 788 (28) 112 (43) <0.01
Non-smokers, n (%) 1429 (49) 107 (42) <0.01
Mild alcohol consumption (<80 g/day), n (%) 320 (11) 34 (13) 0.28
Pneumococcal vaccine, n (%) 307 (15) 66 (28) <0.01
Influenza vaccine, n (%) 898 (42) 166 (70) <0.01
No. of comorbidities ⱖ 1, n (%) 1853 (62) 215 (83) <0.01
Neurologic disease, n (%) 533 (18) 59 (22) 0.78
Respiratory disease, n (%) 1117 (38) 151 (59) <0.01
Cardiac failure, n (%) 469 (16) 74 (29) <0.01
Chronic renal failure, n (%) 174 (6) 22 (9) 0.14
Diabetes mellitus, n (%) 469 (16) 68 (27) <0.01
Chronic hepatic disease, n (%) 100 (3) 7 (3) 0.54
Evidence of frequent aspiration, n (%) 322 (11) 51 (20) <0.01
Previous antibiotic therapy, n (%) 662 (23) 79 (31) <0.01
Autonomy for daily activities 0.08
Total, n (%) 1435 (79) 181 (73) —
Partial, n (%) 237 (13) 42 (17) —
None, n (%) 138 (8) 25 (10) —
Previous pneumonia episode, n (%) 353 (12) 69 (27) <0.01
Protonic pump inhibitors use, n (%) 419 (15) 30 (12) 0.30
Inhaled corticosteroids use, n (%) 437 (15) 92 (36) <0.01
Statins, n (%) 87 (11) 1 (33) 0.30

SD, standard deviation.

© 2012 The Authors Respirology (2013) 18, 263–271

Respirology © 2012 Asian Pacific Society of Respirology
266 E Polverino et al.

Table 2 Clinical presentation data

Clinical presentation Non-steroid group Steroid group P-value

Fever, n (%) 2478 (85) 190 (74) <0.01

Chills, n (%) 1484 (51) 125 (49) 0.60
Cough, n (%) 2345 (81) 220 (87) 0.01
Sputum, n (%) 167 (58) 172 (70) <0.01
Pleuritic pain, n (%) 1204 (42) 92 (37) 0.16
Dyspnoea, n (%) 1886 (65) 217 (84) <0.01
Digestive symptoms, n (%) 115 (10) 0 (0) 0.39
Mental status alteration, n (%) 521 (18) 70 (27) <0.01
Length of symptoms before admission (days), median (IQR) 4 (3–7) 5 (3–7) 0.69
Respiratory rate (breaths/min), median (IQR) 24 (20–32) 28 (24–35) <0.01
Heart rate (bpm), median (IQR) 96 (84–108) 100 (85–110) 0.32
Systolic blood pressure (mm Hg), median (IQR) 130 (113–150) 140 (120–153) <0.01
Diastolic blood pressure (mm Hg), median (IQR) 71 (65–80) 75 (63–82) 0.51
Temperature (°C), median (IQR) 37.7 (36.7–38.3) 37.5 (36.5–38.2) 0.06
Analytical data
C-reactive protein (mg/dL), median (IQR) 17.6 (8.1–27.3) 14.0 (6.2–25.3) 0.04
Creatinine (mg/dL), median (IQR) 1.0 (0.8–1.3) 1.2 (0.9–1.6) <0.01
Leucocyte count (109 cell/L), median (IQR) 12.6 (9.0–17.3) 13.9 (9.8–18.2) 0.08
Neutrophils (%), median (IQR) 83 (76–87) 82 (76–88) 0.94
Platelet count (109 cell/L), median (IQR) 232 (181–297) 264 (205–340) <0.01
Na (mmol/L), median (IQR) 136 (133–139) 136 (133–139) 0.49
K (mmol/L), median (IQR) 4.0 (3.6–4.3) 4.0 (3.7–4.4) 0.08
Haematocrit (%), median (IQR) 40 (36–43) 41 (37–44) <0.01
Glycaemia (mg/dL), median (IQR) 122 (104–154) 132 (108–175) <0.01
PaO2 (mm Hg), median (IQR) 62 (54–72) 59 (52–65) 0.09
PaCO2 (mm Hg), median (IQR) 34.5 (30.3–38.4) 36.1 (31.5–42.4) <0.01
PaO2/FiO2, median (IQR) 291 (248–333) 267 (238–300) 0.08
pH, median (IQR) 7.45 (7.41–7.48) 7.45 (7.41–7.48) 0.05
HCO3 (mmol/L), median (IQR) 24.2 (21.7–27.0) 24.8 (22.0–27.2) 0.19
PSI classes I–III, n (%) 1406 (50) 62 (24) <0.01
PSI classes IV–V, n (%) 1397 (50) 198 (76) —
Radiological data
Cavitation, n (%) 26 (1) 2 (1) 0.99
Atelectasis, n (%) 131 (5) 9 (4) 0.54
Pleural effusion 404 (14) 32 (13) 0.56
Multilobar infiltration, n (%) 344 (25) 65 (33) 0.02
Distress, n (%) 17 (2) 3 (2) 0.74
Empyema, n (%) 17 (2) 1 (1) 0.50
Bacteraemia, n (%) 298 (10) 27 (10) 0.91
Acute renal failure, n (%) 109 (8) 28 (14) <0.01
Shock, n (%) 56 (4) 8 (4) 1.0
Cardiac arrhythmias, n (%) 43 (3) 16 (8) <0.01

FiO2, fraction of inspiratory oxygen; HCO3, bicarbonate; IQR, interquartile range; PaCO2, arterial carbon dioxide tension; PaO2, arterial
oxygen tension; PSI, Pneumonia Severity Index.

percentage of respiratory comorbidities (any respira-
tory disease, 59% vs 38%; COPD, 27% vs 15%; asthma,
11% vs 4%; P < 0.01, each), more cardiac failure and
more diabetes mellitus (Table 1). PSI score was not
different between COPD patients who received
steroid and COPD subjects of the non-steroid group
(COPD patients receiving steroids, PSI IV–V: 77%;
COPD patients not receiving steroids, PSI IV–V: 71%;
P = 0.340). The frequency of COPD in the group of
diabetic patients who received steroids was double
that among diabetic patients of the non-steroid group
(35% vs 17%, P < 0.01). A higher percentage of
patients from the steroid group had received antibi-
otic therapy prior to hospital visit and were chroni-
cally using inhaled corticosteroids (ICS) (Table 1).
Respirology (2013) 18, 263–271
Patients receiving steroids were more symptomatic
(i.e. dyspnoea, cough, sputum) and more severe (mul-
tilobar infiltration, acute renal failure, arterial hypox-
aemia) than patients of non-steroid group (Table 2).
Moreover, patients from the steroid group also
showed more respiratory (42% vs 34%, P = 0.021) and
extrapulmonary complications (28% vs 17%, P < 0.01)
at admission.
Of the respiratory complications observed on
admission, cavitations and atelectasis were uncom-
mon (1–5%) and similarly distributed among the two
groups. Pleural effusion was equally distributed
between the two groups, while multilobar infiltration
was more frequent among patients in the steroid
group (Table 2). The most frequent extrapulmonary
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
Corticosteroids in CAP
Table 3 Microbial aetiology of community-acquired pneumonia according to steroid therapy
Non-steroid group
Pathogen n (% of all patients)
Streptococcus pneumoniae 506 (17%)
Atypical pathogens 120 (4%)
Legionella pneumophila 94 (3%)
Haemophilus influenzae 43 (1%)
Pseudomonas aeruginosa 30 (1%)
Staphylococcus aureus 23 (1%)
Others <1%
Polymicrobial infection 126 (4%)
Atypical pathogens: Mycoplasma pneumoniae, Klebsiella pneumoniae, Chlamydophilia psittaci, Chlamydophila pneumoniae.
Others: Moraxella catarrhalis, Enterobacteriaceae, etc. Polymicrobial infection: bacteria plus virus or bacteria or atypical pathogen.
complications were acute renal failure, septic shock
and cardiac arrhythmias. The other extrapulmonary
complications (such as diarrhoea) were reported in
less than 1.5% of groups each and differences
between the two study groups were not significant.
Microbiological findings
Microbial aetiology could be determined in 1284
(39%) patients (overall population) and more fre-
quently among patients in the steroid group (n, 117
(45%) vs n, 1167 (39%), P = 0.02). No significant differ-
ences were observed in the composition of microbial
aetiology between the two groups (Table 3).
Predictors of steroid administration
Factors significantly associated with the administra-
tion of systemic corticosteroids are shown in Table 4.
The multivariate logistic regression analysis
showed an association of corticosteroid administra-
tion with COPD (odds ratio (OR), 1.91), fever (OR,
0.59), expectoration (OR, 1.56), arterial hypoxaemia
(arterial oxygen saturation ⱖ 92%; OR, 0.46),
C-reactive protein (+5 mg/dL; OR, 0.92), creatinine
(+1 mg/dL; OR, 1.19) and cardiac failure (OR, 1.76).
Main outcomes
Patients receiving corticosteroids were more fre-
quently hospitalized (90% vs 76%; P < 0.01) and
received more mechanical ventilation (MV), but had a
similar rate of admission to intensive care unit
(Table 5).
Moreover, they had longer LOS, but experienced
similar duration of MV in comparison with the non-
steroid group.
The 30-day mortality rate for patients in the steroid
group was similar to those in the non-steroid group,
as was clinical stability (Table 5).
Cumulative doses of systemic corticosteroids
and main outcomes
The median cumulative dose of systemic corticoster-
oids administered during the first 7 days of hospitali-
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
267
Steroid group
n (% of all patients) P-value
38 (15%) 0.53
10 (4%) 0.74
7 (3%) 0.55
5 (2%) 0.37
5 (2%) 0.24
3(1%) 0.23
<1% —
15 (6%) 0.09
zation was 243 mg of methyl-prednisolone (IQR, 140–
290 mg) or a corresponding equivalent dose of other
corticosteroids. The cumulative dose did not show
any association with mortality, the need for MV, or
any respiratory (pleural effusion, etc.) or extrapulmo-
nary (renal failure, etc.) complication (P > 0.1, each),
with the only expected exception of length of hospital
stay (r = 0.259, P < 0.01).
Contemporarily, the administration of systemic
corticosteroids for more than 7 days was clearly
associated with prolonged LOS (median (IQR), 13
(9–18) days vs 6 (5–11) days; P < 0.01), delayed clini-
cal stability (median (IQR), 6.5 (3–10) days vs 4
(3–5.5); P < 0.01), with increased use of MV (24% vs
3% of patients; P < 0.01) and with increased fre-
quency of extrapulmonary complications (37% vs
22% of patients; P = 0.021), shock (7% vs 2%) and
cardiac arrhythmia (12% vs 5%). In contrast, pro-
longed administration of systemic corticosteroids
(>7 days) did not show any association with mortal-
ity (steroids >7 days, 6.1% vs steroids ⱕ7 days, 7.6%;
P = 0.654).
High-risk pneumonia according to PSI
and outcomes
We also separately analysed the patients in high PSI
risk classes (IV–V) in order to evaluate a potentially
different impact of systemic corticosteroids on out-
comes in patients with more severe pneumonia. We
therefore investigated a total of 1592 patients (steroid
group 198 (12%); non-steroid group, 1394 (88%)).
Even in this subset of high-risk patients, we did not
observe any differences in intensive care unit admis-
sion rates (steroid group, 12 (6%); non-steroid group,
100 (7%); P = 0.14) or mortality (steroid group, 17
(9%); non-steroid group, 135 (11%); P = 0.41) or days
to clinical stability (steroid group, 4.5 (3–8) days; non-
steroid group, 5 (3–8) days; P = 0.49) between the two
groups. In addition, as with the overall population,
patients from the steroid group had longer LOS
(steroid group, 10 (6–15) days; non-steroid group, 7
(5–11); P < 0.01).
Respirology (2013) 18, 263–271
268 E Polverino et al.

Table 4 Significant factors associated with initial administration of systemic corticosteroids in univariate and multi-
variate logistic regression analyses

Univariate Multivariate

Variable OR 95% CI P-value OR 95% CI P-value

Age (+5 year) 1,16 1,115–1,212 <0.01 — — —

Smoking habit — — <0.01 — — —
Active smokers 0.74 0.51–1.08 0.11 — — —
Ex-smokers 1.89 1.43–2.51 <0.01 — — —
Non-smokers 1 — — — — —
Inhaled corticosteroids 3.2 2.45–4.26 <0.01 — — —
Chronic obstructive respiratory disease 2.12 1.64–2.74 <0.01 1.91 1.34–2.74 <0.01
Diabetes mellitus 1.86 1.38–2.49 <0.01 — — —
Previous antibiotic in the last month 1.50 1.14–1.99 <0.01 — — —
Previous pneumonia 2.63 1.95–3.54 <0.01 — — —
Aspiration evidence 1.97 1.42–2.74 <0.01 — — —
Fever 0.50 0.37–0.67 <0.01 0.59 0.40–0.88 <0.01
Cough 1.65 1.12–2.41 <0.01 — — —
Expectoration 1.66 1.25–2.20 <0.01 1.56 1.07–2.26 0.02
Leucocyte count (+10 ¥ 109 cell/L) 1.13 0.99–1.30 0.08 — — —
Creatinine (+1 mg/dL) 1.20 1.07–1.35 <0.01 1.19 1.02–1.39 0.03
C-RP (+5 mg/dL) 0.94 0.88–1.00 <0.05 0.92 0.86–0.99 0.03
Platelet count (+10 ¥ 109 cell /L) 1.23 1.01–1.04 <0.01 — — —
PaO2/FiO2 (+10 unit) 0.97 0.95–0.98 <0.01 — — —
Arterial oxygen saturation ⱖ92% 0.50 0.38–0.66 <0.01 0.46 0.33–0.65 <0.01
PSI classes IV–V 3.21 2.39–4.32 <0.01 — — —
Multilobar infiltration 1.82 1.38–2.40 <0.01 — — —
Cardiac failure 2.86 1.57–2.78 <0.01 1.76 1.21–2.58 <0.01
Renal failure 2.13 1.36–3.34 <0.01 — — —
Cardiac arrhythmia 3.08 1.70–5.61 <0.01 — — —

Chronic obstructive respiratory disease includes asthma, chronic bronchitis and emphysema. The variables subjected to a univariate
analysis were: age, gender, smoking, alcohol consumption, previous antibiotic, inhaled corticosteroids, previous pneumonia, fever,
shivers, cough, expectoration, chronic pulmonary disease, chronic cardiovascular disease, diabetes mellitus, neurological disease,
chronic renal failure, chronic liver disease, creatinine, C-reactive protein level, leucocytes, platelets, Na, PO2/FiO2, arterial oxygen
saturation, PSI, multilobar infiltration, pulmonary complications and extrapulmonary complications. ‘+10 ¥ 109 cell/L’ indicates the
increase by ten 109 cell/L. ‘+1 mg/dL’ indicates the increase by one mg/dL. ‘+1 ¥ 109 cell/L’ indicates the increase by one 109 cell/L. ‘+1 unit’
indicates the increase by one unit.
CI, confidence interval; C-RP, C-reactive protein; OR, odds ratio; PaO2/FiO2, arterial oxygen tension/inspiratory oxygen fraction; PSI,
Pneumonia Severity Index.

Table 5 Main outcomes

Non-steroid group Steroid group P-value

Site of care <0.01

Outpatients, n (%) 696 (24) 27 (10)
Ward admission, n (%) 2006 (69) 218 (84)
ICU admission, n (%) 228 (8) 15 (6)
Days of mechanical ventilation, median (IQR) 5 (2–8) 3 (2.0–6.5) 0.90
Length of hospital stay (days), median (IQR) 6 (3–9) 9 (6–14) <0.01
Length of ICU stay (days), median (IQR) 5 (3–9) 6 (4–9) 0.85
Days to clinical stability, median (IQR) 4 (3–6) 5 (3–7) 0.11
Mechanical ventilation, n (%) 176 (6.5) 25 (10.8) <0.02
1-month mortality, n (%) 154 (6) 18 (7) 0.43

ICU, intensive care unit; IQR, interquartile range.

Respirology (2013) 18, 263–271 © 2012 The Authors

Respirology © 2012 Asian Pacific Society of Respirology
Corticosteroids in CAP 269

Predictors of mortality and LOS Both multivariate analyses (mortality and LOS)
Univariate and multivariate logistic regression analy- were adjusted for PSI, multilobar infiltration and pre-
ses revealed the following risk factors for mortality: vious antibiotic therapy.
age, aspiration risk, altered mental status on admis-
sion, increased respiratory rate, chills, renal failure,
bacteraemia, MV (Table 6). Administration of sys-
temic corticosteroids did not show any statistical asso- DISCUSSION
ciation with mortality (univariate and multivariate
analysis). The utility of administering corticosteroids in CAP is
Univariate and multivariate logistic regression still an unresolved issue in the literature. Although
analyses revealed the following risk factors for animal13 and observational studies on inflammation
prolonged LOS: administration of systemic cor- suggest a potential benefit1 by decreasing lung bacte-
ticosteroids, age, aspiration evidence, PSI, dyspnoea, rial burden and systemic inflammation,14–16 evidence
increased respiratory rate; increased C-reactive in humans is unclear. As confirmation, current guide-
protein, multilobar infiltration, pleural effusion, acute lines for CAP management do not suggest the use of
renal failure, cardiac arrhythmia, MV (Table 7). systemic corticosteroids, independently of comor-
bidities and severity.10,17
To date, few randomized trials2,4,6,8 have included
different degrees of CAP severity. The only trial
Table 6 Multivariate logistic regression analyses of showing reduced mortality was performed in patients
mortality admitted to the intensive care unit and the trial was
halted prematurely due to 0% mortality in the steroid
95% CI treatment arm.2 In contrast, an observational pro-
spective study by Salluh et al.18 on severe CAP patients
OR Lower Upper P-value described that the main indications for systemic
steroid were bronchospasm (52.5%) and septic shock
Age (+5 years) 1.20 1.06 1.35 0.003 (36%), but found no differences in mortality in
Aspiration evidence 4.51 2.50 8.13 <0.001 comparison with patients not receiving corticoster-
Altered mental status 2.35 1.38 4.01 0.002 oids. Nevertheless, patients with CAP often receive
Chills 0.36 0.20 0.64 <0.01 systemic corticosteroids.
Respiratory rate (+5 bpm) 1.33 1.13 1.57 <0.001 The present work being an observational study
Creatinine, (+1 mg/dL) 1.26 1.021 1.56 0.032 describing the current practice in CAP cannot be
Bacteraemia 2.45 1.23 4.90 0.011 compared with clinical trials investigating systemic
Mechanical Ventilation 9.30 4.89 17.71 <0.01 steroids as adjunctive therapy of CAP. However, to our
knowledge, this is the first study in the literature to
CI, confidence interval; OR, odds ratio.
analyse the clinical features of CAP patients associ-
ated with the administration of systemic corticoster-
oids that excluded patients receiving chronic
Table 7 Multivariate logistic regression analyses of outpatient corticosteroid therapy. It is also a compre-
length of hospital stay hensive analysis of consecutive patients in our hospi-
tal setting diagnosed with CAP by research staff not
95% CI
involved in the patients treatment.
OR Lower Upper P-value
The main findings of this study were:
1 Systemic corticosteroids were administered at the
Steroid administration 2.01 1.29 3.14 <0.01 discretion of the managing physician for perceived
Age (+5 years) 1.08 1,03 1.13 <0.01 need, in 8% of all patients with CAP in our series. The
Aspiration evidence 1.89 1.25 2.85 <0.01 main reasons for administering steroids were the
Dyspnoea 1.41 1.04 1.92 0.03 presence of a chronic respiratory disease or high
Respiratory rate (+5 bpm) 1.18 1.06 1.32 <0.01 pneumonia severity. Importantly, none of the patients
C-RP (+5 mg/dL) 1.09 1.03 1.16 <0.01 were on chronic outpatient corticosteroid therapy,
PSI classes IV–V 1.54 1.05 2.25 0.02 and thus the use of steroids was not just a continua-
Multilobar infiltration 1.89 1.33 2.67 <0.01 tion of a pre-existing therapy. In addition, we found a
Pleural effusion 3.62 1.91 6.86 <0.01 series of predictors associated with the administra-
Extrapulmonary <0.01 tion of corticosteroids that confirmed a higher sever-
complications ity population overall when compared with patients
No 1.00 — — — who did not receive these drugs.
Shock 1.51 0.60 3.80 0.39 2 We could not find any association of steroid admin-
Renal failure 1.48 0.91 2.41 0.12 istration with lower mortality or shorter time to clini-
Cardiac arrhythmia 5.98 2.13 16.75 <0.01 cal stability. In contrast, it was associated with
Others 3.38 1.17 9.74 0.02 prolonged LOS (>7 days). Moreover, patients who
Mechanical ventilation 7.84 3.19 19.26 <0.01 received steroids for longer time (>7 days) showed
worse clinical course (LOS, clinical stability, MV)
CI, confidence interval; C-RP, C-reactive protein; OR, odds and more systemic complications (shock, cardiac
ratio; PSI, Pneumonia Severity Index. arrhythmia). These results remained similar when we
© 2012 The Authors Respirology (2013) 18, 263–271
Respirology © 2012 Asian Pacific Society of Respirology
270
analysed the more severe patients (elevated PSI)
separately, and independently of cumulative corticos-
teroids doses.
3 The analysis of predictors of LOS clearly indicated a
number of parameters of CAP severity (PSI, multilo-
bar infiltration, pleural effusion, MV, etc.) along with
the administration of systemic steroids.
For all these reasons, the administration of systemic
steroids does not seem to be clearly beneficial in our
population.
However, there are a few points that are arguable
and this study has some limitations.
1 Patients receiving steroids had an elevated PSI;
however, the mortality was not higher in this group as
might have been expected. This could be explained by
the fact that patients receiving corticosteroids had
received ICS therapy more frequently and this may
play an undetermined influence on modulation
of local inflammatory response19–21 or microbial
growth.22,23 Another factor that may influence the
main outcomes in the steroid group is the higher
prevalence of previous antibiotic use in these
patients. Despite the practical difficulties involved in
determining whether the previous antibiotic was
appropriate according to microbial aetiology of pneu-
monia, different studies have reported that prior anti-
microbial treatment is a protective factor for severe
CAP (OR, 0.37; 95% confidence interval, 0.17–0.79;
P = 0.009).24–26 An early reduction in bacterial load or a
coexisting anti-inflammatory effect (i.e. macrolides)
may have a protective effect on CAP severity and
explain why patients in the steroid group did not
show increased mortality despite a higher PSI.
However, all these factors should have been control-
led for in the multivariate analyses.
2 The older age and more severe clinical presenta-
tion in the group (PSI score, multilobar infiltration,
acute renal failure) could also explain the longer LOS
measured.
3 Some limitations need to be addressed. The data
collection did not include the side-effects possibly
related to systemic corticosteroids (hyperglycaemia,
myopathy, etc.).
In this study, the patients who received systemic
corticosteroids were mostly affected by asthma or
COPD, had previously received ICS and other antibi-
otic cycles in a higher percentage, and showed worse
clinical conditions on admission. While steroid
administration did not influence mortality rate or
time to clinical stability (independently of the cumu-
lative dosage of corticosteroids), on the other hand, it
was associated with prolonged LOS. However,
because steroids were mostly prescribed to patients
with some chronic respiratory disease and higher PSI
and resulted in no excess mortality in comparison
with control group, this also may suggest that the use
of systemic steroids is not contraindicated in those
patients whose pneumonia seems associated with an
exacerbation of an underlying chronic respiratory
disease at medical evaluation.
In addition, despite the lack of recommendation or
protocol for the use of corticosteroids in CAP, this
practice was shown to be still relatively frequent with
the physician having to make the decision to use
Respirology (2013) 18, 263–271
E Polverino et al.
these drugs at his own discretion. Only a very well-
performed, randomized clinical trail focusing on very
severe patients will definitively clarify if corticoster-
oids have some utility in CAP despite the possible
collateral effects and help draw recommendations
and protocols to guide clinical practice.
Acknowledgements
Financial support was received from 2009SGR911, CIBERES
(Initiative of ISCIII, CB06/06/0028), IDIBAPS.
REFERENCES
1 Monton C, Ewig S, Torres A et al. Role of glucocorticoids on
inflammatory response in nonimmunosuppressed patients with
pneumonia: a pilot study. Eur. Respir. J. 1999; 14: 218–20.
2 Confalonieri M, Urbino R, Potena A et al. Hydrocortisone infu-
sion for severe community-acquired pneumonia: a preliminary
randomized study. Am. J. Respir. Crit. Care Med. 2005; 171: 242–8.
3 Garcia-Vidal C, Calbo E, Pascual V et al. Effects of systemic ster-
oids in patients with severe community-acquired pneumonia.
Eur. Respir. J. 2007; 30: 951–6.
4 Meijvis SC, Hardeman H, Remmelts HH et al. Dexamethasone
and length of hospital stay in patients with community-acquired
pneumonia: a randomised, double-blind, placebo-controlled
trial. Lancet 2011; 377: 2023–30.
5 Mikami K, Suzuki M, Kitagawa H et al. Efficacy of corticosteroids
in the treatment of community-acquired pneumonia requiring
hospitalization. Lung 2007; 185: 249–55.
6 Fernandez-Serrano S, Dorca J, Garcia-Vidal C et al. Effect of cor-
ticosteroids on the clinical course of community-acquired pneu-
monia: a randomized controlled trial. Crit. Care 2011; 15: R96.
7 Marik P, Kraus P, Sribante J et al. Hydrocortisone and tumor
necrosis factor in severe community-acquired pneumonia. A
randomized controlled study. Chest 1993; 104: 389–92.
8 Snijders D, Daniels JM, de Graaff CS et al. Efficacy of corticoster-
oids in community-acquired pneumonia: a randomized double-
blinded clinical trial. Am. J. Respir. Crit. Care Med. 2010; 181:
975–82.
9 Salluh JI, Povoa P, Soares M et al. The role of corticosteroids in
severe community-acquired pneumonia: a systematic review.
Crit. Care 2008; 12: R76.
10 Lim WS, Baudouin SV, George RC et al. BTS guidelines for the
management of community acquired pneumonia in adults:
update 2009. Thorax 2009; 64(Suppl. 3): iii1–55.
11 Fine MJ, Auble TE, Yealy DM et al. A prediction rule to identify
low-risk patients with community-acquired pneumonia. N. Engl.
J. Med. 1997; 336: 243–50.
12 Lim WS, van der Eerden MM, Laing R et al. Defining community
acquired pneumonia severity on presentation to hospital: an
international derivation and validation study. Thorax 2003; 58:
377–82.
13 Sibila O, Luna CM, Agusti C et al. Effects of glucocorticoids in
ventilated piglets with severe pneumonia. Eur. Respir. J. 2008; 32:
1037–46.
14 Menendez R, Cavalcanti M, Reyes S et al. Markers of treatment
failure in hospitalised community acquired pneumonia. Thorax
2008; 63: 447–52.
15 Menendez R, Martinez R, Reyes S et al. Biomarkers improve mor-
tality prediction by prognostic scales in community-acquired
pneumonia. Thorax 2009; 64: 587–91.
16 Martinez R, Menendez R, Reyes S et al. Factors associated with
inflammatory cytokine patterns in community-acquired pneu-
monia. Eur. Respir. J. 2011; 37: 393–9.
17 Mandell LA, Wunderink RG, Anzueto A et al. Infectious Diseases
Society of America/American Thoracic Society consensus
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
Corticosteroids in CAP
guidelines on the management of community-acquired pneu-
monia in adults. Clin. Infect. Dis. 2007; 44(Suppl. 2): S27–S72.
18 Salluh JI, Soares M, Coelho LM et al. Impact of systemic corticos-
teroids on the clinical course and outcomes of patients with
severe community-acquired pneumonia: a cohort study. J. Crit.
Care 2011; 26: 193–200.
19 Gutierrez P, Closa D, Piner R et al. Macrophage activation in
exacerbated chronic obstructive pulmonary disease with and
without community-acquired pneumonia. Eur. Respir. J. 2010;
36: 285–91.
20 Malo de Molina R, Mortensen EM, Restrepo MI et al. Inhaled
corticosteroid use is associated with lower mortality for subjects
with COPD and hospitalised with pneumonia. Eur. Respir. J. 2010;
36: 751–7.
21 Chen D, Restrepo MI, Fine MJ et al. Observational study
of inhaled corticosteroids on outcomes for COPD patients
with pneumonia. Am. J. Respir. Crit. Care Med. 2011; 184: 312–
6.
© 2012 The Authors
Respirology © 2012 Asian Pacific Society of Respirology
271
22 Calverley PM, Anderson JA, Celli B et al. Salmeterol and flutica-
sone propionate and survival in chronic obstructive pulmonary
disease. N. Engl. J. Med. 2007; 356: 775–89.
23 Drummond MB, Dasenbrook EC, Pitz MW et al. Inhaled corti-
costeroids in patients with stable chronic obstructive pulmonary
disease: a systematic review and meta-analysis. JAMA 2008; 300:
2407–16.
24 Ruiz M, Ewig S, Torres A et al. Severe community-acquired pneu-
monia. Risk factors and follow-up epidemiology. Am. J. Respir.
Crit. Care Med. 1999; 160: 923–9.
25 Menendez R, Torres A, Rodriguez CF et al. Reaching stability in
community-acquired pneumonia: the effects of the severity of
disease, treatment, and the characteristics of patients. Clin.
Infect. Dis. 2004; 39: 1783–90.
26 Johnson D, Carriere KC, Jin Y et al. Appropriate antibiotic utili-
zation in seniors prior to hospitalization for community-
acquired pneumonia is associated with decreased in-hospital
mortality. J. Clin. Pharm. Ther. 2004; 29: 231–9.
Respirology (2013) 18, 263–271
Copyright of Respirology is the property of Wiley-Blackwell and its content may not be copied or emailed to
multiple sites or posted to a listserv without the copyright holder's express written permission. However, users
may print, download, or email articles for individual use.