Eur Respir J 1998; 12: 113–115
Printed in UK - all rights reserved
Copyright ©ERS Journals Ltd 1998
European Respiratory Journal
ISSN 0903 - 1936
Aetiology and outcome of severe community-acquired
pneumonia in Singapore
Y-K. Tan, K-L. Khoo, S-P . Chin, Y-Y. Ong
Severe community-acquired pneumonia (SCAP) occurs
in approximately 18–36% of all community-acquired pneu-
monia (CAP) [1–3] and is associated with a mortality of
up to 58% . This condition has considerable economi-
cal impact because of the need for intensive care monitor-
ing and mechanical ventilation, the loss of productivity
through prolonged hospitalization and the loss of life. The
most common organisms responsible for SCAP in the
West are Streptococcus pneumoniae, gram-negative bacilli
and Legionella pneumophila. Haemophilus influenzae and
Staphylococcus aureus are less common causes of SCAP
[4, 5]. Guidelines [6–8] now exist for the initial empi-
rical management of SCAP, but there may be variation in
the aetiological agents in different geographical areas
and, therefore, provision for this variation may be neces-
sary when treating patients in different geographical re-
There is very little literature on the aetiology of SCAP
in Asia [9, 10]. The present study aimed to look at
whether there is a difference between the aetiological
agents in SCAP in an Asian country and those in the West,
where most published guidelines originate. If the aetiolog-
ical agents in the Asian community were found to be dif-
ferent from those in the West, such information should be
added to the recommendations for initial antimicrobial
agents from the consensus of the American, British and
Canadian Thoracic Societies.
Subjects and methods
This was a retrospective study of 57 consecutive pati-
ents admitted to the intensive care unit (ICU) of Singapore
General Hospital from January 1989 to May 1993 with the
diagnosis of SCAP. A diagnosis of CAP was made when
there was an acute lower respiratory tract infection with an
onset before admission to hospital, associated with clini-
cal and radiological evidence of pulmonary consolidation
within 48 h of admission. Patients >12 yrs of age and those
without a history of recent hospitalization within a month
were eligible. Patients were excluded if they had pulmonary
oedema, aspiration pneumonia, underlying immunocom-
promised states, such as malignancy, organ transplantation,
infection with the human immunodeficiency virus (HIV),
underlying chronic obstructive pulmonary disease (COPD),
bronchiectasis, or restrictive lung disease. Although COPD
is a risk factor for CAP, not all such patients with SCAP
were routinely admitted to the ICU because of limited re-
sources, and they were thus excluded from the analysis.
Aetiology and outcome of severe community-acquired pneumonia in Singapore. Y-K. Tan,
K-L. Khoo, S-P. Chin, Y-Y. Ong. ©ERS Journals Ltd 1998.
ABSTRACT: The aim of this study was to determine the aetiology and outcome of
severe community-acquired pneumonia, and to assess whether the existing guidelines
for initial antimicrobial therapy are being applied.
The records of 57 consecutive nonimmunocompromised patients admitted to the
medical intensive care unit (ICU) between January 1989 and May 1993 with this
diagnosis were reviewed. The microbiological data, chest radiographic changes and
outcome were analysed.
Nine (16%) of the 57 patients had pulmonary tuberculosis. When these patients
were excluded from further analysis, a microbiological diagnosis was made in 41
(72%) cases. The most commonest pathogens were Burkholderia pseudomallei (n=10),
Klebsiella spp. (n=5) and Staphylococcus aureus (n=5), Mycoplasma pneumoniae (n=4)
and Streptococcus pneumoniae (n=2) were less common. This microbiological spec-
trum was quite different from that in the West, where the incidence of S. pneumoniae
was higher. Also, when pulmonary tuberculosis was excluded, the mortality (67%)
was much higher than that in other series. This was attributed to the high incidence
of unrecognized B. pseudomallei infection, which is associated with a very high mor-
tality in the region under study.
In addition to applying published guidelines on severe community-acquired pneu-
monia, the endemicity of certain organisms such as Mycobacterium tuberculosis and
Burkholderia pseudomallei in different geographical regions needs to be considered
when choosing initial empirical antimicrobial therapy.
Eur Respir J 1998; 12: 113–115.
Dept of Respiratory and Critical Care Med-
icine, Singapore General Hospital, Singa-
Correspondence: Y-K. Tan
Dept of Respiratory and Critical Care Med-
Singapore General Hospital
Fax: 65 2271736
Keywords: Burkholderia pseudomallei
initial empirical antimicrobial therapy
severe community-acquired pneumonia
Received: September 2 1997
Accepted after revision March 9 1998
Y-K. TAN ET AL.
Patients who were admitted from chronic care facilities
such as nursing homes were also excluded. ICU admis-
sions were for mechanical ventilation, potential ventila-
tory support, or intensive medical and nursing care and
The data recorded included age and sex, history of smok-
ing, diabetes mellitus and alcohol misuse, chest radiogra-
phic changes, results of microbiological investigations, and
outcome. The microbiological data were based primarily
on sputum and/or endotracheal aspirate Gram-stain and cul-
ture, acid-fast bacilli smear and mycobacterial cultures,
blood cultures, and serology for Mycoplasma and Legio-
nella, taken within 24 h of admission to hospital. Pleural
fluid cultures, bronchoscopy and bronchoalveolar lavage
(BAL) were also performed if indicated.
A definitive diagnosis of the aetiological agent was
made when either the blood cultures or the pleural fluid
cultures were positive, or when the acid-fast bacilli smear
or mycobacterial cultures in sputum were positive. A pro-
bable diagnosis of the aetiological agent was made when
there was: 1) a four-fold rise in mycoplasma serology or a
single titre of 1:128 in the presence of cold agglutinins; or
2) a four-fold rise in Legionella serology or a single titre
of 1:256 or higher. A probable diagnosis was also made
when sputum, endotracheal aspirate and BAL cultures were
A total of 57 patients were identified during the study
period (table 1). Nine patients had active pulmonary
tuberculosis based on acid-fast bacilli smear or mycobac-
terial cultures from sputum. After excluding these pati-
ents, 48 patients remained (35 males and 13 females) with
SCAP. Their mean (±SD) age was 53±18 yrs (range 12–86
yrs). All but two required mechanical ventilation and their
mean length of stay in the ICU was 9.1±9.0 days. Twenty
(42%) were present smokers of exsmokers, 14 (29%) had
diabetes mellitus and six (12%) had a history of alcohol
After excluding pulmonary tuberculosis, the top three
causative pathogens in this series were B. pseudomallei
(n=10), S. aureus (n=4) and Klebsiella spp. (n=4). Five of
the patients with B. pseudomallei infection (50%) had dia-
betes mellitus and one had a history of alcohol misuse.
The overall mortality was 61%. If pulmonary tuberculosis
was excluded, the mortality was 67%.
In this study of patients with SCAP admitted to the
ICU, 9 patients had pneumonia due to M. tuberculosis.
Although rare in the West, this agent has also been re-
ported in a few other Asian countries [9, 10]. This aetio-
logical agent should thus still be considered in patients
presenting with CAP in countries where tuberculosis is
endemic. Ten patients had pneumonia due to B. pseudo-
mallei. This infection is endemic in many south-east
Asian countries [11, 12] and northern Australia . This
organism was found by another centre in Singapore to
cause 7% of SCAP cases . All patients who had this
infection died. The antibiotics of choice are a combination
of intravenous ceftazidime and oral doxycline .
Regarding the remaining pathogens, the numbers were too
small to make any significant conclusion.
The present series showed that the aetiological agent
could be identified in a large percentage of patients (74%).
This was because a large number of patients with SCAP
were bacteraemic and it emphasizes the need for routine
blood cultures in patients with SCAP. Some prospective
studies have made no aetiological diagnosis in 31–61% of
patients [15, 16].
The mortality of patients with SCAP was much higher
in the present series than in other studies and was as high
as the mortality (63%) in another centre in Singapore .
This was attributed to the very high mortality associated
with unsuspected B. pseudomallei infection.
The findings in the present series were further strength-
ened by similar findings in another retrospective study by
a different centre in Singapore . In addition to the rec-
ommendations by the American, Canadian and British
Thoracic Societies, the endemicity of certain organisms in
different geographical region needs to be considered. For
instance, in south-east Asian countries and northern Aus-
tralia, it is important to target initial empirical antimicro-
bials for severe community-acquired pneumonia against
B. pseudomallei. ANSTEY et al.  found that a fatal out-
come was strongly associated with inappropriate initial
antibiotic therapy and identified appropriately tailored the-
rapeutic regimens for the Northern Territory of Australia.
A useful antibiotic regimen for patients with severe-com-
munity-acquired pneumonia in Singapore would be a
combination of intravenous ceftazidime (high dose), clo-
xacillin and oral doxycycline. This combination of antimi-
crobials would cover the more usual pathogens in severe
community-acquired pneumonia, Staphylococcus aureus,
atypical pneumonia (Mycoplasma pneumonia, Legionella
pneumophila and Chlamydia spp.) and Burkholderia pseu-
Table 1. – Aetiological agents and outcome
No. of patients
(acid-fast bacilli smear)
(nine from blood culture, one
from pleural fluid and sputum)
(four from blood, culture, one
from sputum and ETT culture)
(four from blood culture,
one from BAL culture)
(four from serology)
(three from blood culture,
one from BAL culture)
(two from serology)
(one from ETT culture, one
from sputum and BAL culture)
(one from blood culture)
ETT: endotracheal tube; BAL: bronchoalveolar lavage.
SEVERE COMMUNITY-ACQUIRED PNEUMONIA
1. Marrie TJ, Durant H, Yates L. Community-acquired pneu-
monia requiring hospitalisation: 5-year prospective study.
Rev Infect Dis 1989; 11: 586–599.
Sorensen J, Forsberg P, Hakanson E, et al. A new diagno-
stic approach to the patient with severe pneumonia. Scand
J Infect Dis 1989; 21: 33–41.
Woodhead MA, Macfarlane JT, Rodgers FG, Laverick A,
Pilkington R, Macrae AD. Aetiology and outcome of sev-
ere community-acquired pneumonia. J Infect 1985; 10:
Hirani NA, Macfarlane JT. Impact of management guide-
lines on the outcome of severe community acquired pneu-
monia. Thorax 1997; 52: 17–21.
Moire P, Verchen JB, Chevret S, Chastang C, Gajdos P.
Severe community-acquired pneumonia: etiology, epidemio-
logy, and prognosis factors. Chest 1994; 105: 1487–1495.
Niederman MS, Bass JB, Campbell GD, et al. Guidelines
for the initial management of adults with community-
acquired pneumonia: diagnosis, assessment and severity
and initial antimicrobial therapy. Am Rev Respir Dis 1993;
British Thoracic Society. Guidelines for the management
of community-acquired pneumonia in adults admitted to
hospital. Br J Hosp Med 1993; 49: 346–350.
Mandell LA, Niederman M. The Canadian Community-
Acquired Pneumonia Consensus Group. Antimicrobial
treatment of community-acquired pneumonia in adults: a
conference report. Can J Infect Dis 1993; 4: 25–28.
9. Dahmash NS, Chowdhury MN. Re-evaluation of pneu-
monia requiring admission to an intensive care unit: a
prospective study. Thorax 1994; 490: 71–76.
Lee HK, Hui KP, Tan WC, Lim TK. Severe community
acquired pneumonia in Singapore. Singapore Med J 1996;
Boonsawa W, Boonma P, Tangdajahiran T, Paupermpoon-
sir S, Wongpratoom W, Romphryk A. Community-acqui-
red pneumonia in adults at Svinagarind Hospital. J Med
Assoc Thai 1990; 73: 345–351.
Tan AL, Ang BSP, Ong YY. Melioidosis. Epidemiology
and antibiogram of cases in Singapore. Singapore Med J
1990; 31: 335–337.
Curried B, Howard D, Nguyen VT, Withnall K, Merianos
A. The 1990–91 outbreak of meliodosis in the Northern
Territory of Australia: clinical aspects. SE Asian J Trop
Med Publ Health 1993; 23: 436–443.
Suputtamongkol Y. Diagnosis and treatment of meliodo-
sis. Med Prog 1996; 23: 19–22.
Rello J, Rodriguez R, Jubert P, Alvarez B. Severe com-
munity-acquired pneumonia in the elderly: epidemiology
and prognosis. Study group for severe community-acqui-
red pneumonia. Clin Infect Dis 1996; 23: 723–728.
Feldman C, Ross SA, Goolam MD, Omar J, Smith C. The
etiology of severe community-acquired pneumonia and
its impact on initial, empiric, antimicrobial chemotherapy.
Respir Med 1995; 89: 187–192.
Anstey NM, Currie BJ, Withnall KM. Community-acqui-
red Acinetobacter pneumonia in the Northern Territory of
Australia. Clin Infect Dis 1992; 14: 83–91.