Efficacy of Corticosteroids in Community-acquired
A Randomized Double-Blinded Clinical Trial
Dominic Snijders1, Johannes M. A. Daniels2, Casper S. de Graaff1, Tjip S. van der Werf3, and Wim G. Boersma1
1Department of Pulmonary Diseases, Medical Center Alkmaar, Alkmaar, The Netherlands;2Department of Pulmonology, VU University Medical
Center, Amsterdam, The Netherlands; and3Departments of Internal Medicine and Pulmonary Diseases and Tuberculosis, University Medical Center
Groningen, University of Groningen, the Netherlands
Rationale: Some studieshave shown a beneficial effect of corticoste-
roids in patients with community-acquired pneumonia (CAP), pos-
sibly by diminishing local and systemic antiinflammatory host
Objectives: To assess the efficacy of adjunctive prednisolone treat-
ment in patients hospitalized with CAP.
Methods: Hospitalized patients, clinically and radiologically diag-
randomized to receive 40 mg prednisolone for 7 days or placebo,
along with antibiotics. Primary outcome was clinical cure at Day 7.
to clinical stability, defervescence, and C-reactive protein. Disease
severity was scored using CURB-65 (a severity index for community-
acquired pneumonia evaluating Confusion, blood Urea nitrogen,
Respiratory rate, Blood pressure, and age 65 or older) and Pneumo-
nia Severity Index.
Measurements and Main Results: We enrolled 213 patients. Fifty-four
(25.4%) patients had a CURB-65 score greater than 2, and 93
(43.7%) patients were in Pneumonia Severity Index class IV-V.
Clinical cure at Days 7 and 30 was 84/104 (80.8%) and 69/104
(66.3%) in the prednisolone group and 93/109 (85.3%) and 84/109
(77.1%) in the placebo group (P 5 0.38 and P 5 0.08). Patients on
prednisolone had faster defervescence and faster decline in serum
C-reactive protein levels compared with placebo. Subanalysis of
patients with severe pneumonia did not show differences in clinical
the prednisolone group (20 patients, 19.2%) than in the placebo
different between the two groups.
Conclusions: Prednisolone (at 40 mg) once daily for a week does not
improve outcome in hospitalized patients with CAP. A benefit in
more severely ill patients cannot be excluded. Because of its
association with increased late failure and lack of efficacy prednis-
olone should not be recommended as routine adjunctive treatment
Keywords: community-acquired pneumonia; corticosteroids; infection
Community-acquired pneumonia (CAP) is a leading cause of
morbidity and mortality worldwide (1). Despite the develop-
ments in antibiotic therapy, no substantial progress has been
made in the last decades (2). Additional therapeutic interven-
tions along with antibiotics may help to improve outcome in
patients with CAP.
Corticosteroids have been evaluated in the past decades for
treatment of sepsis and septic shock. Earlier studies before 1990
showed no impact on mortality, whereas studies using current
definitions of sepsis and septic shock showed survival benefit
when corticosteroids were administered at a low dose for a
prolonged period of time (3). In contrast, the Corticosteroid
Therapy of Septic Shock study failed to show mortality re-
duction in patients with sepsis (4). Only faster reversal of shock
Corticosteroids in CAP might be effective in reducing excess
systemic and pulmonary inflammation, which might translate to
improved outcome (5, 6). One earlier study found a beneficial
effect of hydrocortisone on the clinical course in patients
with pneumococcal pneumonia (7), but a more recent study
(8) showed a marked improvement in PaO2/FIO2and also
a survival advantage in patients with severe CAP admitted to
the intensive care unit (ICU) when treated with corticosteroids.
The study was ended after inclusion of 46 patients because
of the benefit found in the interim analysis. This beneficial effect
of steroids on severe CAP was also found in a retrospective
Our hypothesis was that adjunctive treatment with cortico-
steroids along with antibiotic treatment may improve outcome
in patients with CAP. We conducted a clinical randomized
placebo-controlled trial in hospitalized patients with CAP.
Primary end point was clinical outcome at Day 7. Secondary
end points were clinical outcome at Day 30, length of stay
(LOS), time to clinical stability (TTCS), 30-day mortality,
defervescence, and serum C-reactive protein levels (CRP).
AT A GLANCE COMMENTARY
Scientific Knowledge on the Subject
Corticosteroids are used in patients with sepsis or septic
shock, of which a large part is due to pneumonia. The use
of corticosteroids along with antibiotics in patients with
community-acquired pneumonia (CAP) may lead to a di-
minished local and systemic antiinflammatory response.
What This Study Adds to the Field
Prednisolone as an adjunctive treatment along with anti-
biotics does not improve outcome in hospitalized patients
with CAP and may lead to more late failure in patients
with nonsevere CAP.
(Received in original form May 29, 2009; accepted in final form February 1, 2010)
Supported by an unrestricted grant from Astra Zeneca.
Some of the results of these studies have been previously reported in the form of
an abstract at the 2009 annual ATS conference (Snijders D, de Graaff CS, van der
Werf TS, Boersma WG. Clinical efficacy of prednisolone in patients with
community-acquired pneumonia, a randomized clinical trial [abstract A6116].
Presented at the 2009 ATS International Conference, May 15–20.).
Correspondence and requests for reprints should be addressed to Dominic
Snijders, M.D., Department of Pulmonary Diseases, Medical Centre Alkmaar,
Wilhelminalaan 15, 1812 JD Alkmaar, The Netherlands. E-mail: firstname.lastname@example.org
This article has an online supplement, which is accessible from this issue’s table of
contents at www.atsjournals.org
Am J Respir Crit Care Med
Originally Published in Press as DOI: 10.1164/rccm.200905-0808OC on February 4, 2010
Internet address: www.atsjournals.org
Vol 181. pp 975–982, 2010
Patients were prospectively enrolled between August 2005 and July
2008 at the Medical Centre Alkmaar, a 900-bed teaching hospital in the
Netherlands. The study protocol was approved by the local medical
Patients were eligible if they met the following criteria: (1) Written
informed consent obtained. (2) Clinical symptoms suggestive of CAP:
cough (with or without sputum), fever (.38.58C), pleuritic chest pain,
or dyspnea. (3) New consolidations on chest radiograph. (4) Age 18
years or older.
Patients were excluded from the study if one of the following
criteria applied: Presence of severe immunosuppression (HIV infec-
tion, use of immunosuppressants), malignancy, pregnancy or breast-
feeding, use of macrolides for more than 24 hours, use of prednisone
15 mg or more for more than 24 hours, any condition requiring
corticosteroids, any likely infection other than CAP, obstruction
pneumonia (e.g., from lung cancer), pneumonia that developed within
8 days after hospital discharge, and indications that patients were
unable and/or unlikely to comprehend and/or follow the protocol.
Subgroup analysis of patients with severe CAP (CURB-65 [severity
index for community-acquired pneumonia evaluating Confusion, blood
Urea nitrogen, Respiratory rate, Blood pressure, and age 65 or older]
score . 2 or PSI class IV and V) was planned (10, 11).
Patients were double-blinded randomized to receive 40 mg of prednis-
olone once daily or placebo for a total of 7 days, administered in the
same way as the antibiotics (intravenous or oral). When patients were
switched from intravenous to oral antibiotics the study drug was also
switched. Randomization was based on a one-on-one allocation by
means of prenumbered containers containing seven vials for intrave-
nous administrations and seven capsules. The allocation sequence was
computer generated and was kept in a safe at the hospital pharmacy
throughout the course of the study.
All patients were treated with antibiotics according to national
guidelines (12). In all patients urinary antigen testing for Legionella
pneumophila was performed. In general, patients with mild to moder-
ate severe CAP (CURB-65 , 3 or PSI I–III) were treated with
amoxicillin. Patients with moderate to severe CAP, with (a suspicion
of) atypical pathogens or with an intolerance to amoxicillin were
started on moxifloxacin. Alteration of antibiotic treatment was allowed
but the use of macrolides was discouraged because of their immuno-
modulating effect. Duration of antibiotic treatment was entirely left to
the discretion of the medical team in charge, as was the decision
whether or not to switch from intravenous to oral treatment. There
were no criteria for hospital discharge and the investigators did not
influence decisions concerning discharge.
Standard laboratory assessment was performed on presentation and
included renal and liver functions, electrolytes, glucose, hematology,
and CRP (Beckman Coulter Inc., Fullerton, CA). Arterial blood gas
analysis was performed as clinically indicated. Serum samples were
drawn each day of hospitalization until Day 7 and on Day 14 for
assessment of CRP levels.
Clinical outcome at Day 7 and Day 30 was defined as: Cure—resolution
or improvement of symptoms and clinical signs related to pneumonia
without the need for additional or alternative therapy. Failure—
persistence or progression of all signs and symptoms that developed
during the acute disease episode after randomization, or the development
of a new pulmonary or extrapulmonary infection, or the deterioration
of chest radiography after randomization, or death due to pneumonia,
Figure 1. Study flow chart.
976 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINEVOL 1812010
or the inability to complete the study owing to adverse events.
Indeterminate—patients receiving less than 80% of the study drug
for reasons other than clinical failure, a concomitant infection outside
the respiratory tract requiring antibiotic treatment, loss to follow-up, or
death unrelated to the primary diagnosis (13).
An early failure was defined as lack of resolution of signs and
symptoms of pneumonia within 72 hours of treatment and persistence
or progression thereafter. A late failure was defined as a recurrence of
signs and symptoms of pneumonia after 72 hours of admission after an
initially beneficial response to treatment.
Time to clinical stability was assessed by using the criteria defined
by Halm and colleagues (14). In short, patients were clinically stable if
all four of the following criteria were met: improvement of cough and
shortness of breath, temperature less than 37.88C for at least 8 hours,
white blood cell count normalizing, and adequate oral intake and
gastrointestinal absorption. Because of the possibility of elevated WBC
by prednisolone use, the criterion of normalizing WBC was replaced by
declining serum CRP levels. Defervescence was defined as a tempera-
ture less than 37.58C.
At admission, a sputum specimen was ordered for Gram stain, semi-
quantitative culture, and Streptococcus pneumoniae antigen. If possi-
ble, two sets of blood cultures were drawn before the start of antibiotic
therapy. Urine was collected for antigen testing for S. pneumoniae and
L. pneumophila serogroup 1 (enzyme immunoassay, Binax-NOW;
Binax, Portland, ME). Pleural fluid if present was examined by Gram
stain, culture, and S. pneumoniae antigen test.
Blood samples for serology (Serion ELISA classic; Virion GmbH,
Wu ¨rzburg, Germany) were obtained on Day 1 and Day 14 of the study
for detection of antibodies to Mycoplasma pneumoniae, Chlamydo-
phila pneumoniae, L. pneumophila serogroup 1–7, influenza A and
B virus, parainfluenza virus 1–3, respiratory syncytial virus, and
adenovirus. A fourfold increase in antibody titer was considered as
A sample size was calculated based on published data of an earlier trial
(15) in which 75 patients out of a total of 220 patients with CAP
received steroid treatment along with antibiotic treatment. Clinical
success was 93.3% in patients with steroids and 75.9% in patients
without steroids. We calculated that 92 patients were needed in both
arms to detect a difference of 15% between steroid and placebo
treatment at Day 7 with a power of 80% and an a level of 0.05.
The data were summarized as frequencies or percentages for
categorical variables and as means and standard deviations for contin-
uous variables. Differences between the treatment groups were com-
pared by the chi-square or Fisher exact test for categorical variables
TABLE 1. DEMOGRAPHIC FEATURES OF THE INTENTION-TO-TREAT GROUP OF HOSPITALIZED
PATIENTS WITH COMMUNITY-ACQUIRED PNEUMONIA (N 5 213)
(n 5 104)
(n 5 109)P Value
Chronic heart disease
Ischemic heart disease
Clinical signs and symptoms
Systolic blood pressure, mm Hg
Heart rate, bpm
Respiratory rate, breaths/min
C-reactive protein, mg/L
WBC, 3 109/L
Pneumonia Severity Index
63.0 6 17.9
64.0 6 18.7
38.3 6 1.2
127.8 6 24.1
100.6 6 20.6
26.5 6 7.4
258.5 6 154.0
14.8 6 6.8
38.5 6 1.0
129.3 6 24.5
97.0 6 18.7
25.5 6 7.2
214.5 6 144.2
15.3 6 7.2
Definition of abbreviations: COPD 5 chronic obstructive pulmonary disease; CURB-65 5 severity index for community-acquired
pneumonia evaluating Confusion, blood Urea nitrogen, Respiratory rate, Blood pressure, and age 65 or older; ICU 5 intensive
care unit; WBC 5 white blood cell count.
Data are presented as n (%) or mean 6 SD.
TABLE 2. ANTIMICROBIAL TREATMENT IN THE TWO
Prednisolone GroupPlacebo GroupP Value
Amoxicillin and acyclovir
Ciprofloxacin and cefuroxime
Data are presented as n (%).
Snijders, Daniels, de Graaff, et al.: Steroids in CAP977
and a two-sample t test or Mann–Whitney test for continuous variables.
The Kaplan-Meier method was used to analyze time from admission to
discharge and TTCS. Differences in LOS and TTCS between treatment
groups were compared by a log-rank test. Hazard and odds ratios are
reported with 95% confidence intervals. Statistical Package for Social
Sciences (SPSS) version 16.0 was used for data management and
statistical analysis. A planned interim analysis after 125 included
patients with regard to mortality, LOS, and clinical outcome at Days
7 and 30 showed no significant differences and the study was continued
A total of 213 patients were enrolled in the study. Mean age was
63.5 6 18.2 years and 124 (57.9) patients were male. Study flow
chart is shown in Figure 1. There was an imbalance between the
two study groups with respect to CRP levels on admission and
prevalence of chronic heart disease (Table 1). Patients with
severe pneumonia were evenly distributed among the two
groups (CURB-65 score > 3: 28 [13.1%] vs. 26 [12.2%] patients;
P 5 0.61; PSI IV–V: 48 [46.2%] vs. 45 [41.3%] patients; P 5
0.47). Reasons for exclusion are listed in Table E1 in the online
Antimicrobial treatment was similar in both study arms (Table
2). Twenty-six (25.0%) patients in the prednisolone group and
25 (22.0%) patients in the placebo group were using antibiotics
before admission (Table E2).
Primary and secondary outcome parameters are shown in Table
3. No differences in clinical outcome at Day 7 were found
between patients in the prednisolone and placebo groups
(80.8% vs. 85.3%; P 5 0.38). Kaplan-Meier plots of LOS and
TTCS are shown in Figures 2 and 3. A total of 37 (17.4%)
patients did not complete their course of study medication.
Reasons for not completing the course of study medication were:
death in 10 patients, overruling decisions by attending physician
to prescribe corticosteroids in 14 patients (3 patients with COPD
and 1 patient with asthma), withdrawal of informed consent in 5
patients, and postrandomization exclusion in 8 patients. There
were no differences between the prednisolone group and the
placebo group in the need for additional corticosteroids (six
[5.8%] vs. 8 [7.3%] patients; P 5 0.64). Of the 14 patients who
were given additional corticosteroids, 7 (50.0%) were admitted
to the ICU. Resolution of fever was faster in the prednisolone
group (Figure 4). Median (6interquartile range [IQR]) day of
defervescence was Day 2 6 1 day in the prednisolone group and
Day 3 6 2 days in the placebo group (P , 0.01).
The decline in CRP levels was faster in the prednisolone
group up until Day 7 (Figure 5). At Day 14, patients in the
with patients in the placebo group (41.73 6 64.98 vs. 22.05 6
53.32 mg/L; P , 0.01).
In patients with nonsevere CAP late failures occurred more
often in the prednisolone group than in the placebo group
(CURB-65 0–2; 15/76 vs. 5/87 patients; P , 0.01 and PSI classes
I–III; 10/56 vs. 4/64 patients; P 5 0.05).
In the prednisolone group, five (4.8%) patients with late
failure needed an additional course of antibiotics, seven (6.7%)
patients needed another or a prolonged course of prednisolone,
and six (5.8%) patients developed a pleural effusion or empy-
ema necessitating additional therapy. In the placebo group, two
(1.8%) patients needed additional antibiotics, three (2.8%)
patients needed a course of prednisolone, and one patient
developed a pleural effusion requiring additional therapy.
Subanalysis of primary and secondary outcome parameters
in patients with severe CAP (CURB-65 . 2 or PSI class 4–5)
TABLE 3. CLINICAL OUTCOME BY INTENTION-TO-TREAT AND PER-PROTOCOL ANALYSIS
OutcomePrednisolone Group Placebo GroupP Value
Odds Ratio or Mean
Difference (95% CI)
Intention to treat
Clinical cure at Day 7
Clinical cure at Day 30
Clinical cure at Day 7
Clinical cure at Day 30
10.0 6 12.0
4.9 6 6.8
10.6 6 12.8
4.9 6 5.2
20.56 (24.00 to 2.8)
0.03 (21.6 to 1.71)
10.0 6 12.1
5.0 6 7.0
10.4 6 13.1
4.9 6 5.3
20.40 (24.01 to 3.22)
0.12 (21.68 to 1.92)
Definition of abbreviations: CI 5 confidence interval; LOS 5 length of stay; TTCS 5 time to clinical stability.
All data are presented as n (%) or mean 6 SD.
Figure 2. Kaplan-Meier curves showing the effect of the intervention
on length of stay in the intention-to-treat population. Solid line 5
prednisolone; dashed line 5 placebo. Log rank, 0.84; hazard ratio, 1.15;
95% confidence interval, 0.81–1.55; P 5 0.36.
978 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINEVOL 1812010
are shown in Table 4. Subanalysis of mechanically ventilated
patients showed no differences in the primary and secondary
outcome parameters (Table E3).
An etiological diagnosis for CAP was made in 118 (55.4%)
patients; S. pneumoniae (78 [36.6%]) was the most frequently
found causative microorganism. Distribution of the pathogens
between the two groups is shown in Table 5. Patients with
pneumococcal pneumonia in the prednisolone group had a lower
clinical cure rate than patients with pneumococcal pneumonia in
the placebo group. The clinical cure rate in the prednisolone
group at Day 7 and Day 30 was 29 (69.0%) and 20 (47.6%). In
the placebo group the clinical cure rate was 31(86.1%) and 28
(77.8%) (P 5 0.08 and P 5 0.01). Patients with pneumococcal
pneumonia and prednisolone treatment also had a significantly
higher number of late failures (11 [26.2%] vs. 2 [5.6%]; P 5 0.02).
Patients in the prednisolone group who had no pathogen
identified had a shorter TTCS than patients in the placebo group
(3 [IQR, 2] vs. 4 [IQR, 2] d; P 5 0.01), a shorter LOS (5.5 [IQR,
3] vs. 7 [IQR 7] d; P 5 0.03), and faster defervescence (2 [IQR, 1]
vs. 3 [IQR, 3] d; P , 0.01). No other differences were observed
with respect to etiology and clinical outcome.
Hyperglycemia with the need for additional therapy during
admission occurred in five (2.3%) patients in the prednisolone
group and two (0.9%) patients in the placebo group (P 5 0.27).
Confusion during admission was noted in four (1.9%) patients
in the prednisolone group and in three (1.4%) patients in the
placebo group (P 5 0.72). A superinfection occurred in 10
(2.1%) patients in the prednisolone group and in 4 (1.9%)
patients in the placebo group (P 5 0.10). One patient in the
placebo group developed a fungal infection after he was treated
with hydrocortisone for septic shock in the ICU after clinical
failure. Another patient in the placebo group was diagnosed
with pulmonary embolism 10 days after hospital admission. A
total of 63 (60.6%) patients in the prednisolone group and 72
(66.1%) in the placebo group did not have any treatment-
related adverse event (P 5 0.41).
This is the first randomized double-blinded placebo-controlled
trial of corticosteroids in hospitalized patients with CAP. We
found no beneficial effects of adjunctive corticosteroids in
patients hospitalized with CAP; clinical cure was equal in both
groups at Day 7. A trend toward a higher clinical cure rate in
the placebo group was observed. The overall clinical cure rate
(83% at Day 7 and 71.8% at Day 30) is in concordance with
other studies (15). Our findings contrast with the findings in
other recent studies. In experimental studies a benefit has been
found with the combination of hydrocortisone and antibiotics
(16, 17). Both studies demonstrated a reduction in inflammatory
cytokines. The use of ciprofloxacin and hydrocortisone in
a piglet model of Pseudomonas pneumonia also decreased
bacterial burden more than ciprofloxacin-treated or untreated
piglets. In the only randomized double-blinded clinical trial
published to date, evaluating corticosteroids in patients with
CAP admitted to the ICU, a marked reduction in mortality and
LOS was found (8). The study included a small number of
patients because the study was ended after interim analysis
showed reduced mortality and improved oxygenation in pa-
tients treated with corticosteroids. A Spanish retrospective
study also found a reduced mortality in patients treated
with corticosteroids (9). Both these studies only included
patients with severe CAP, who are more likely to benefit from
Figure 3. Kaplan-Meier curves showing the effect of the intervention
on time to clinical stability in the intention-to-treat population. Solid
line 5 prednisolone; dashed line 5 placebo. Log rank, 0.60; hazard
ratio, 1.14; 95% confidence interval, 0.82–1.59; P 5 0.44.
Figure 4. Defervescence in patients with community-acquired pneu-
monia treated with prednisolone or placebo. Diamonds 5 pednisolone;
squares 5 placebo. Data are presented as mean 6 SD. * P , 0.01.
Day 14 for patients in both study groups. Diamonds 5 pednisolone;
squares 5 placebo. Data are presented as mean 6 SD. * P 5 0.03;fP ,
Serum C-reactive protein levels during treatment and at
Snijders, Daniels, de Graaff, et al.: Steroids in CAP 979
corticosteroids. A possible rationale for the use of corticoste-
roids is the existence of relative adrenal insufficiency in severe
CAP (18). However, patients included in the Corticosteroid
Therapy of Septic Shock study had sepsis or septic shock, with
one-third of the patients suffering from a pulmonary infection
(4). There were no better outcomes in patients who were
nonresponders to a corticotrophin test. In our study, subanalysis
of patients with severe CAP did not show a beneficial effect of
corticosteroids, although our definition of severe CAP was
based on the CURB-65 or the PSI, not the modified American
Thoracic Society criteria as used by Confalonieri and colleagues
(8, 19). Also, the absolute numbers of patients who needed
mechanical ventilation was low. Gotoh and colleagues (20)
examined adrenal insufficiency in 64 patients hospitalized with
CAP and found a low incidence of adrenal insufficiency in this
population. Only 14% fulfilled the criteria for adrenal insuffi-
ciency. Adrenal insufficiency is probably not clinically relevant
in patients with nonsevere CAP.
Symptom resolution, reduction of LOS, and reduction of
intravenous antibiotic therapy are also important clinical goals
in the treatment of patients with CAP. In a study by Mikami
and colleagues (21), the authors concluded that corticosteroids
in patients with CAP hastens symptom resolution and shortens
the duration of treatment with intravenous antibiotics. No effect
on LOS was observed, but this study had a low number of
included patients and the open-label design. The antiinflamma-
tory effects of prednisolone did not lead to a shorter LOS or
TTCS in our study, despite the observed faster defervescence
and decline in CRP in patients treated with prednisolone. The
more than twofold increase of late failures in the prednisolone
group raises questions about the occurrence of a rebound of
inflammation after initial suppression by corticosteroids. Sub-
clinical inflammation is found in a majority of patients with
severe CAP. In a large study elevated IL-6 levels were found in
clinically stable patients with severe CAP on the day of
discharge. Furthermore, higher IL-6 levels were correlated with
a higher mortality in the subsequent 3 months (22). The
assumption of rebound of inflammation in our study is strength-
ened by the higher CRP levels in patients in the prednisolone
group after 2 weeks, after an initially faster decline in the first
week. This rebound phenomenon is also observed in the study
by Garcia-Vidal and colleagues (9). Nonsurvivors on cortico-
steroid therapy died later than nonsurvivors without corticoste-
roids (13.8 vs. 7.1 d, respectively). The effect of a delayed
inflammatory response due to withdrawal of corticosteroids can
cause a prolongation of the time between admission and death
in the nonsurvivors with corticosteroids. In light of the presence
of subclinical inflammation in patients with CAP on hospital
discharge, a tapering of corticosteroids might protect patients
against the rebound of inflammation (22). Another possible
explanation for the higher incidence of late failure may be
nosocomial infection, leading to additional treatment. Never-
theless, similar to this trial, recent metaanalysis found no
evidence of increased risk for nosocomial infections in cortico-
steroid-treated patients (3). In our study, the inclusion of
patients with nonsevere pneumonia may have resulted in
a higher rate of late failures instead of increased late mortality.
Late failures may lead to new or prolonged courses of antibi-
otics or corticosteroids and subsequently to a higher risk of
adverse events. The corticosteroid treatment–related adverse
events in our study, however, were low and did not differ from
placebo. The time to clinical stability was similar between
the two groups, with a mean of 5 days. As the TTCS can be
used as a decision tool for safely switching antibiotics from
TABLE 4. SUBANALYSIS OF CLINICAL OUTCOME OF SEVERE PNEUMONIA AS DEFINED BY CURB-65 SCORE GREATER
THAN 2 OR PNEUMONIA SEVERITY INDEX CLASS LESS THAN III BY INTENTION-TO-TREAT ANALYSIS
Prednisolone GroupPlacebo GroupP Value
Odds Ratio or Mean
Difference (95% CI)
Clinical cure at Day 7
Clinical cure at Day 30
Pneumonia severity index classes 4–5
Clinical cure at Day 7
Clinical cure at Day 30
19.4 6 20.2
9.9 6 11.2
20.4 6 22.7
8.7 6 9.8
21.00 (213.24 to 11.24)
1.13 (25.18 to 7.43)
15.0 6 16.4
7.7 6 9.8
16.4 6 18.4
6.8 6 7.6
21.38 (28.85 to 6.09)
1.96 (23.07 to 4.71)
Definition of abbreviations: CI 5 confidence interval; CURB-65 5 severity index for community-acquired pneumonia evaluating Confusion, blood Urea nitrogen,
Respiratory rate, Blood pressure, and age 65 or older; LOS 5 length of stay; TTCS 5 time to clinical stability.
All data are presented as n (%) or mean 6 SD.
TABLE 5. CAUSE OF COMMUNITY-ACQUIRED PNEUMONIA
IN 231 PATIENTS
(n 5 104)
(n 5 109)
Data are presented as n (%).
* Mixed infection with S. pneumoniae.
†One mixed infection influenza A/Haemophilus influenzae, two mixed infection
influenza A/S. pneumoniae.
‡One mixed infection M. pneumoniae/S. pneumoniae.
980AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINEVOL 1812010
intravenous to oral administration, the effect of prednisolone on
the duration of intravenous antibiotic therapy will be limited
(23). CAP can be caused by different pathogens and the effect
of prednisolone on the different pathogens can also be different.
Patients with pneumococcal pneumonia treated with corticoste-
roids had a higher clinical failure rate in our study. No effects on
outcome with respect to other pathogens were noted, although
the absolute numbers in our study may have been be too small
to detect such differences.
Some limitations may apply to our study. First, no assess-
ment of adrenal function was performed in these patients, so no
data regarding the presence of relative adrenal insufficiency are
known. Second, the use of clinical cure as primary outcome is
a subjective outcome parameter, prone to bias. But in our
opinion this reflects daily clinical practice and because of the
randomized design the introduction of bias is minimized.
Furthermore, the exclusion criteria of the need for cortico-
steroid therapy may have led to an underrepresentation of
patients with COPD. The simultaneous occurrence of bronchial
obstruction and CAP in patients with COPD necessitates the
use of systemic corticosteroids in these patients (24). Although
the question of whether mortality by CAP is higher in patients
with COPD is still a subject for debate, the possible exclusion of
these patients may have created a selection bias (25, 26). The
percentage of patients with COPD in our study was 20.2%,
whereas in a previous study in our hospital 36.6% of the
patients with CAP had COPD (15). Our findings should not
be extended to patients with CAP and COPD, as generalizabil-
ity is limited.
A fourth possible limitation is that our study may have been
underpowered to detect a clinically significant difference in this
population, which also included a large proportion of nonsevere
pneumonia. But even in this case, the number needed to treat
to detect a favorable outcome with prednisolone will not be
clinically relevant. Also the effect of prednisolone can be
diminished by a late administration. As the clinical cure rate
is rather high in nonsevere pneumonia, the administration of
prednisolone 24 hours after admission is not likely to have
a significant effect on the clinical course. As all patients were
included within 24 hours of admission, we do not believe this
limits our findings.
The Dutch guidelines concerning antibiotic therapy in
patients with CAP differ from the Infectious Diseases Society
of America/American Thoracic Society guidelines (12, 27). The
low antibiotic resistance in the Netherlands (see http://
the conflicting data in the literature concerning combination
therapy support the antibiotics used in this study conducted in
the Netherlands (28, 29)
As the last limitation, we used prednisolone in a once-daily
dosage, for practical reasons, which may not be sufficient for
establishing effective serum levels during 24 hours, although the
pharmacodynamic effects are known to last beyond the time
frame indicated by pharmacokinetic parameters. This limits the
comparison with studies using hydrocortisone by continuous
In conclusion, prednisolone at 40 mg once daily for 1 week
does not improve outcome in hospitalized patients with CAP. A
benefit in more severely ill patients cannot be excluded.
Because of its association with increased late failure in patients
with nonsevere CAP and lack of benefit, prednisolone should
not be recommended as routine adjunct treatment in CAP.
Conflict of Interest Statement: D.S. received up to $1,000 from Brahms for an ATS
Travel Grant 2009. D.S.’s dependent received $5,001–$10,000 from AstraZeneca
as an institutional unrestricted research grant. J.M.A.D. received $50,001–
$100,000 from GlaxoSmithKline as an unrestricted collaborative research grant.
C.S.D.G. received up to $1,000 from Boehringer Ingelheim for a lecture for
family doctors/pulmonary function tests. T.S.V.D.W. received up to $1,000 from
Pfizer for serving on an advisory board (antifungals) and up to $1,000 from Pfizer
and up to $1,000 from Novartis in lecture fees. W.G.B. received up to $1,000
from Brahms to advise on planning a new study and up to $1,000 from Ooms
Allergy for serving on an advisory board; $1,001–$5,000 from Brahms AG for
lectures on LRTI and PCT; and $1,001–$5,000 from Sanofi-Aventis as an expert
witness, serving as a national coordinator Cassiopea study (PE).
Acknowledgment: The authors thank the medical and nursing staff at the
Medical Centre Alkmaar who participated in this study, Tjeerd van der Ploeg
for statistical advice, the staff at the medical microbiology department, and the
staff at the department for clinical chemistry.
1. Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weisfeld LA,
Kapoor WN. Prognosis and outcomes of patients with community-
acquired pneumonia. A meta-analysis. JAMA 1996;275:134–141.
2. Dı ´az LA, Mortensen EM, Anzueto A, Restrepo MI. Novel targets in
the management of pneumonia. Ther Adv Respir Dis. 2008;2:387–
3. Annane D, Bellissant E, Bollaert PE, Briegel J, Confalonieri M, De
Gaudio R, Keh D, Kupfer Y, Oppert M, Meduri GU. Corticosteroids
in the treatment of severe sepsis and septic shock in adults. JAMA
4. Sprung CL,Annane D, Keh D, Moreno R, Singer M, Freivogel K, Weiss
YG, Benbenishty J, Kalenka A, Forst H, et al. Hydrocortisone therapy
for patients with septic shock. N Engl J Med 2008;358:111–124.
5. Puren AJ, Feldman C, Savage N, Becker PJ, Smith C. Patterns of
cytokine expression in community-acquired pneumonia. Chest 1995;
6. Monto ´n C, Torres A, El-Ebiary M, Filella X, Xaubet A, de la Bellacasa
JP. Cytokine expression in severe pneumonia: a bronchoalveolar
lavage study. Crit Care Med 1999;27:1745–1753.
7. Wagner HN, Bennet IL, Lasagna L, Cluff LE, Rosenthal MB, Mirick
GS. The effect of hydrocortisone upon the course of pneumococcal
pneumonia treated with penicillin. Bull Johns Hopkins Hosp 1956;98:
8. Confalonieri M, Urbino R, Potena A, Piattella M, Parigi P, Puccio G,
Della Porta R, Giorgio C, Blasi F, Umberger R, et al. Hydrocortisone
infusion for severe community-acquired pneumonia: a preliminary
randomized study. Am J Respir Crit Care Med 2005;171:242–248.
9. Garcia-Vidal C, Calbo E, Pascual V, Ferrer C, Quintana S, Garau J.
Effects of systemic steroids in patients with severe community-
acquired pneumonia. Eur Respir J 2007;30:951–956.
10. Lim WS, van der Eerden MM, Laing R, Boersma WG, Karalus N, Town
GI, Lewis SA, Macfarlane JT. Defining community acquired pneu-
monia severity on presentation to hospital: an international derivation
and validation study. Thorax 2003;58:377–382.
11. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, Singer DE,
Coley CM, Marrie TJ, Kapoor WN, et al. A prediction rule to identify
low-risk patients with community-acquired pneumonia. N Engl J Med
12. Schouten JA, Prins JM, Bonten M, Degener JE, Janknegt R, Hollander
JM, Jonkers R, Wijnands W, Verheij T, Sachs A, Kullberg BJ.
(Optimizing the antibiotics policy in The Netherlands. VIII. Revised
SWAB guidelines for antimicrobial therapy in adults with community-
acquired pneumonia). Ned Tijdschr Geneeskd 2005;49:2495–2500.
13. Chow AW, Hall CB, Klein JO, Kammer RB, Meyer RD, Remington JS.
Evaluation of new anti-infective drugs for the treatment of respira-
tory tract infections. Infectious Diseases Society of America and the
Food and Drug Administration. Clin Infect Dis 1992;15:62–88.
14. Halm EA, Fine MJ, Marrie TJ, Coley CM, Kapoor WN, Obrosky DS,
Singer DE. Time to clinical stability in patients hospitalized with
community-acquired pneumonia: implications for practice guidelines.
15. van der Eerden MM, Vlaspolder F, de Graaff CS, Groot T, Bronsveld
W, Jansen HM, Boersma WG. Comparison between pathogen di-
rected antibiotic treatment and empirical broad spectrum antibiotic
treatment in patients with community acquired pneumonia: a pro-
spective randomised study. Thorax 2005;60:672–678.
16. Li Y, Cui X, Li X, Solomon SB, Danner RL, Banks SM, Fitz Y, Annane
D, Natanson C, Eichacker PQ. Risk of death does not alter the efficacy
of hydrocortisone therapy in a mouse E. coli pneumonia model: risk
and corticosteroids in sepsis. Intensive Care Med 2008;34:568–577.
17. Sibila O, Luna CM, Agustı ´ C, Baquero S, Gando S, Patro ´n JR, Morato
JG, Absi R, Bassi N, Torres A. Effects of glucocorticoids in ventilated
piglets with severe pneumonia. Eur Respir J 2008;32:1037–1046.
Snijders, Daniels, de Graaff, et al.: Steroids in CAP981
18. Salluh JI, Verdeal JC, Mello GW, Arau ´jo LV, Martins GA, de Sousa Download full-text
Santino M, Soares M. Cortisol levels in patients with severe commu-
nity-acquired pneumonia. Intensive Care Med 2006;32:595–598.
19. Ewig S, Ruiz M, Mensa J, Marcos MA, Martinez JA, Arancibia F,
ment of severity criteria. Am J Respir Crit Care Med 1998;158:1102–1108.
20. Gotoh S, Nishimura N, Takahashi O, Shiratsuka H, Horinouchi H, Ono
H, Uchiyama N, Chohnabayashi N. Adrenal function in patients with
community-acquired pneumonia. Eur Respir J 2008;31:1268–1273.
21. Mikami K, Suzuki M, Kitagawa H, Kawakami M, Hirota N, Yamaguchi
H, Narumoto O, Kichikawa Y, Kawai M, Tashimo H. Efficacy of
corticosteroids in the treatment of community-acquired pneumonia
requiring hospitalization. Lung 2007;185:249–255.
22. Yende S, D’Angelo G, Kellum JA, Weissfeld L, Fine J, Welch RD, Kong
L, Carter M, Angus DC; GenIMS Investigators. Inflammatory markers
at hospital discharge predict subsequent mortality after pneumonia and
sepsis. Am J Respir Crit Care Med 2008;177:1242–1247.
23. Ramirez JA, Vargas S, Ritter GW, Brier ME, Wright A, Smith S,
Newman D, Burke J, Mushtaq M, Huang A. Early switch from
intravenous to oral antibiotics and early hospital discharge: a pro-
spective observational study of 200 consecutive patients with com-
munity-acquired pneumonia. Arch Intern Med 1999;159:2449–2454.
24. Niewoehner DE, Erbland ML, Deupree RH, Collins D, Gross NJ, Light
RW, Anderson P, Morgan NA. Effect of systemic glucocorticoids on
exacerbations of chronic obstructive pulmonary disease. N Engl J
25. Restrepo MI, Mortensen EM, Pugh JA, Anzueto A. COPD is associated
with increased mortality in patients with community-acquired pneu-
monia. Eur Respir J 2006;28:346–351.
26. Snijders D, van der Eerden MM, de Graaff CS, Boersma WG. The
influence of COPD on mortality and severity scoring in community-
acquired pneumonia. Respiration 2010;79:46–53.
27. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD,
Dean NC, Dowell SF, File TM Jr, Musher DM, Niederman MS,
et al. Infectious Diseases Society of America/American Thoracic
Society consensus guidelines on the management of community-
acquired pneumonia in adults. Clin Infect Dis 2007;44:S27–S72.
28. Robenshtok E, Shefet D, Gafter-Gvili A, Paul M, Vidal L, Leibovici L.
Empiric antibiotic coverage of atypical pathogens for community
acquired pneumonia in hospitalized adults. Cochrane Database Syst
Rev 2008;1:CD004418 10.1002/14651858.CD004418.pub3.
29. Feldman C, Anderson R. Therapy for pneumococcal bacteremia: mono-
therapy or combination therapy? Curr Opin Infect Dis 2009;22:137–
982 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINEVOL 1812010