Acta Pdiatrica ISSN 08035253

REGULAR ARTICLE

Prevalence and risk factors of suppurative complications in children with

pneumonia
Patrice Franois (pfrancois@chu-grenoble.fr)1, Amlie Desrumaux2, Christine Cans3, Isabelle Pin2, Patricia Pavese4, Jos Labarre1
1.Quality of Care Unit, Centre Hospitalier Universitaire de Grenoble, Grenoble, France
2.Department of Pediatrics, Centre Hospitalier Universitaire de Grenoble, Grenoble, France
3.Medical Information Department, Centre Hospitalier Universitaire de Grenoble, Grenoble, France
4.Department of Infectious Diseases, Centre Hospitalier Universitaire de Grenoble, Grenoble, France

Keywords
Children, Community-acquired pneumonia,
Epidemiology, Lung abscess, Pleural empyema
Correspondence
Patrice Francois, Unite devaluation medicale, Pavillon Taillefer, CHU de Grenoble, BP 217; 38043 Grenoble Cedex 09, France
Tel: +33 4 76 76 87 87 |
Fax: +33 4 76 76 88 31 |
Email: pfrancois@chu-grenoble.fr
Received
13 October 2009; revised 1 December 2009;
accepted 22 January 2010.
DOI:10.1111/j.1651-2227.2010.01734.x

Abstract
Aim: To identify the baseline characteristics associated with suppurative complications in children with community-acquired primary pneumonia.
Methods: A retrospective study included all children from 28 days to 15 years old, who presented with community-acquired pneumonia at two French hospitals from 1995 to 2003. Complicated pneumonia was defined by the presence of empyema and or lung abscess.
Results: Of 767 children with community-acquired pneumonia, 90 had suppurative complications: 83 cases of pleural empyema and seven cases of lung abscess. The mean prevalence of complicated pneumonia was 3% during the 19951998 period, and then steadily increased following a
linear trend to reach 23% in 2003. Children with complicated pneumonia were older and had a longer
symptomatic period preceding hospitalization. They were more likely to receive antibiotics, especially
aminopenicillins (p < 0.01), and nonsteroidal anti-inflammatory drugs, especially ibuprofen
(p < 0.001). In multivariable analysis, ibuprofen was the only preadmission therapy that was
independently associated with complicated pneumonia [adjusted OR = 2.57 (1.514.35)].
Conclusion: This study confirms an association between the use of prehospital ibruprofen and suppurative
pneumonic complications.

INTRODUCTION
An increase in the incidence of pleural empyema was
reported for the first time in a retrospective cohort study of
50 children hospitalized in Ohio between 1988 and 1994,
showing that 34% of cases had occurred in the last
12 months (1). The same phenomenon was then reported
by many studies in the United States (25) and in Europe
(610). Between 1995 and 2003, the rate of pleural empyema steadily rose from 14 to 26 per million paediatric hospital admissions in the UK (7). Consistently, Tan et al.
showed an increase in the prevalence of parapneumonic
empyema from 22% in 1994 to 53% in 1999 amongst pneumonia cases caused by Streptococcus pneumoniae in eight
American hospitals (3). Many authors also reported an
increase in pneumonia associated lung abscesses and cavitations (3,11,12).
The reasons for this increase in the prevalence of suppurating complications in children with pneumonia have not
been clearly identified. In a retrospective study of 540 children presenting with pneumonia between 1993 and 1999
(2), Byington et al. found that suppurating complications
were associated with age, recent chicken pox, infection with
S. pneumoniae (especially serotype 1), and therapy with
antibiotics and non-steroidal anti-inflammatory drugs
(NSAIDs) prior to hospital admission.

The aim of this study was to identify the baseline characteristics associated with suppurative complications in
children hospitalized with community-acquired pneumonia. More specifically, we focused on therapy with antibiotics, glucocorticoids and NSAIDs prior to hospital
admission.

PATIENTS AND METHODS

Study design
We conducted a retrospective study of children who were
admitted with community-acquired pneumonia in two hospitals in Ise`re, France, between January 1, 1995 and December 31, 2003.
Setting
The population of the Ise`re district in France included
220 800 children under 15 years of age. Two hospitals in
the area admitted paediatric emergencies: the Grenoble
University Hospital, with 86 paediatric beds, and the Voiron
General Hospital, with 14 paediatric beds.
Population
Computerized administrative hospital discharge abstracts
with an International Classification of Diseases, 10th

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866

861

Suppurative complications of pneumonia

Francois et al.

revision (ICD-10) diagnosis code consistent with the

occurrence of uncomplicated or complicated pneumonia
(J14, J15, J16, J17, J18, J85, J86, J90) were identified for
children aged from 28 days to 15 years. Hospitalization
was defined as direct admission to an inpatient hospital
unit, transfer from an emergency department to an inpatient hospital unit, or admission to an emergency department with discharge to any setting more than 24 h after
initial presentation.
The medical records of all identified patients were
reviewed. Two physicians independently validated each
case of pneumonia as definite, probable or absent using predefined criteria. Pneumonia was defined by a new radiographic pulmonary infiltrate associated with at least one of
three clinical symptoms including fever above 38C, cough
or thoracic pain. Patients were excluded if they had hospitalacquired pneumonia or lower respiratory tract infection
secondary to an inherent illness (asthma, bronchiolitis,
severe heart disease, severe neurological disease and swallowing disorder, sickle-cell disease, primary or secondary
immunodeficiency, foreign body inhalation).
Uncomplicated vs. complicated pneumonia
Complicated pneumonia was defined as pleural effusion
more than 1 cm thick on ultrasound or by presence of lung
abscess or cavitation on presentation or during hospital
stay. Conversely, uncomplicated pneumonia was defined as
pneumonia without evidence of pleural effusion, abscess or
cavitation from hospital admission to discharge.
Data collection
Two physicians collected the following baseline data using a
structured chart review: patient demographics, fever duration, number of medical consultations or visits, therapy with
antibiotic, glucocorticoids and NSAIDs before presentation,
physical examination findings, relevant laboratory test
results and chest radiography findings from the time of presentation. As a result of the observational nature of our
study, laboratory tests were ordered at the discretion of physicians in charge of the patients.
Statistical analysis
Categorical variables were expressed as frequencies and
percentages, and continuous variables as median and interquartile range (IQR, 25th and 75th percentiles). In univariable analysis, we compared the characteristics for patients
with complicated and uncomplicated pneumonia, using the
v2 or Fisher exact tests when appropriate for categorical
variables and the Kruskal-Wallis test for continuous variables. Differences in the prevalence of complicated pneumonia according to study year were analysed using v2 for
trends.
Multivariable logistic regression was used to estimate the
odds ratios of complicated pneumonia associated with the
use of antibiotics, glucocorticoids or NSAIDs prior to hospital admission after adjusting for demographic (age, gender), study year, fever duration, and number of medical
consultations.

862

In a supplementary analysis, we determined whether

there was a correlation between the annual volume of complicated pneumonia and sales of ibuprofen, using the Spearman test. Annual sales of the paediatric form of ibuprofen in
France during the period of the study were provided by the
AFSSAPS (French Agency for Safety of Health Care Products). p-values less than 0.05 were considered statistically
significant. Data were analysed using Statview (SAS Institute Inc., Cary, NC, USA).

RESULTS
Patients
From 1995 to 2003, 1184 hospital stays were recorded with
an ICD-10 discharge diagnosis consistent with the occurrence of pneumonia (Fig. 1). The medical record was not
found or was incomplete for 69 children. After chart review,
348 children were excluded because of age <28 days
(n = 10), the lack of clinical or radiological inclusion criterion (bronchiolitis or bronchitis, n = 116), respiratory tract
infection complicating an inherent pathology (asthma,
cystic fibrosis, heart disease, neurological disease, immunodeficiency, foreign body inhalation, n = 178) and hospitalacquired pneumonia (n = 44). Finally, 767 children with
primary community-acquired pneumonia were analysed.
The median age of all children was 3.1 years (IQR, 1.4
5.7), 55.6% were male patients, and 92.4% lived in the Ise`re
district. Our study sample consisted of 677 children (88.3%)
with uncomplicated pneumonia and 90 children (11.7%)
with suppurative complication: 70 cases of pleural empyema, 13 cases of pleural empyema associated with lung
abscess, and 7 cases of lung abscess without pleural involvement. Two of these children had a fatal outcome related to
group A streptococcus pleural empyema. One underwent
septic shock and cardiopulmonary arrest at the arrival in
the emergency department, the other had malignant hyperthermia with multiple organ failure and died in the intensive
care unit 6 h after hospital admission.
Trends in suppurative complications
From 1995 to 1998, the prevalence of complicated pneumonia was stable, ranging from two to four cases per year
(2.8% of hospitalized pneumonia cases). Since 1999, we
have observed an increase in the annual number of cases,
reaching 34 cases in 2003 (23.1% of hospitalized pneumonia cases, Fig. 2). A model consisting of a constant phase
from 1995 to 1998 followed by a linear growth phase from
1999 to 2003 (p < 0.001) adequately fitted the data. When
we considered only children living in Ise`re, the populational
incidence of suppurative complications proceeded from 0.5
to 13 per 100 000 children between 1995 and 2003.
Microbiological data
In the complicated pneumonia group, the microbiological
diagnosis was positively established in 19 cases (21%), with
bacterial isolation in blood culture (ten cases) and or in
pleural effusion (11 cases) (Table 1). These bacteria comprised 14 S. pneumoniae, two group A streptococcus, one

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866

Francois et al.

Suppurative complications of pneumonia

Hospital discharge abstracts with ICD-10

diagnosis codes of pneumonia
(n = 1184)

Missing or incomplete medical record (n = 69)

Excluded (n = 348)
Secondary lower respiratory tract infection (n = 178)
No clinical / radiological inclusion criteria (n = 116)
Hospital-acquired pneumonia (n = 44)
Age <28 days (n = 10)

Analyzed
(n = 767)

Uncomplicated pneumonia (n = 677)

Complicated pneumonia (n = 90)

Pleural empyema (n = 70)
Pleural empyema with lung abscess (n = 13)
Lung abscess without pleural involvment (n = 7)

Figure 1 Study population.

40

35

5
20
4
15

10

Million bottles

25
N cases

Community-acquired pneumonia

30

5
0

Table 1 Microbiological data for children with complicated and uncomplicated

community-acquired pneumonia

1
1995

1996

1997

1998

1999

2000

2001

2002

2003

Figure 2 Progression in the annual number of cases of complicated pneumonia and paediatric-form ibuprofen sales from 1995 to 2003. number of complicated pneumonia. ibuprofen sales (million bottles).

Fusobacterium necrophorum, one F. nucleatum and one

Staphylococcus aureus. Eleven strains of S. pneumoniae
were penicillin-sensitive and three showed intermediate
sensitivity. The serotype was known for eight strains: five
belonged to group 1 and the other three to groups 5, 9 and
14. In the uncomplicated pneumonia group, 13 blood cultures were positive. One was a group A streptococcus and
12 were S. pneumoniae, ten of which were sensitive to

Definite diagnosis*
Streptococcus pneumoniae
Streptococcus pyogenes A
Fusobacterium sp.
Staphylococcus aureus

Uncomplicated
(n = 677)

Complicated
(n = 90)

12 (BS)
1 (BS)

14 (9 BS, 7 PE)
2 (PE)
2 (1 BS, 1 PE)
1 (PE)

*Bacteria isolation in blood strain (BS) or pleural effusion (PE).

penicillin and two showing intermediate sensitivity. The

serotype was established in nine strains: four belonged to
group 1, two to group 19, and the other three belonged to
groups 4, 6 and 14.
Baseline characteristics associated with suppurative
complications
Children with complicated pneumonia were older, were
more likely to be enrolled after 1998 and to be hospitalized
after a longer preadmission period with clinical symptoms
(Table 2). They also underwent more medical examinations
and received antibiotics more often in the preadmission

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866

863

Suppurative complications of pneumonia

Francois et al.

Table 2 Comparison of baseline characteristics for children with complicated

and uncomplicated community-acquired pneumonia

Characteristics
Demographics
Male gender, n (%)
Age (years)*
Hospitalization after
1998, n (%)
Preadmission data
Fever duration*
No. medical
examinations*
Antibiotics, n (%)
Amino-penicillin
Cephalosporin
Macrolide
Others
Anti-inflammatory
treatment, n (%)
Ibuprofen
Aspirin
Glucocorticoids
Other anti-inflammatory

Community-acquired pneumonia

Table 3 Unadjusted and adjusted odds ratios (OR) and 95% confidence interval (95% CI) of complicated pneumonia associated with preadmission
treatments

Uncomplicated
(n = 677)

Preadmission
treatments

373 (55.1)
3.0
(1.35.6)
402 (59.4)

3 (25)
2 (12)

Complicated
(n = 90)

54 (60.0)
4.1
(2.06.6)
80 (88.9)

6 (410)
2 (23)

0.43
0.02
<0.001

<0.001
<0.001

270 (39.9)
137 (20.2)
112 (16.5)
56 (8.3)
14 (2.0)
203 (29.9)

46 (51.1)
31 (34.4)
19 (21.1)
12 (13.3)
3 (3.3)
42 (46.6)

0.05
<0.01
0.29
0.16
0.44
<0.01

97 (14.3)
50 (7.4)
47 (6.9)
14 (2.1)

33 (36.7)
5 (5.5)
8 (8.9)
4 (4.4)

<0.001
0.17
0.20
0.30

*Median (25th and 75th percentiles).

Missing values in retrospective chart review were substituted by null values for
fever duration before admission (n = 38), number of medical visits before
admission (n = 88), preadmission antibiotic treatment (n = 31), and preadmission use of anti-inflammatory drugs (n = 336).

period. The use of amino-penicillins was more frequent in

children with complicated pneumonia, whereas no difference was found for other antibiotics.
Complicated pneumonia was associated with a more frequent use of anti-inflammatory drugs (46.6% vs. 29.9%,
p < 0.01) and specially ibuprofen (36.7% vs. 14.3%,
p < 0.001). No significant difference in the use of aspirin,
corticoids or other NSAIDs [niflumic acid (16 cases), ketoprofen or tiaprofenic acid] was found between children with
complicated and uncomplicated pneumonia. Importantly,
no significant first-order interaction was found between the
use of ibuprofen and the study period: the unadjusted odds
ratios of complicated pneumonia associated with ibuprofen
were 2.43 and 2.48 for the 19951998 and 19992003 study
periods respectively. In multivariable analysis adjusting for
demographic and contextual variables, ibuprofen was the
only treatment medication used before hospital admission
that remained associated with an increased odds ratio of
complicated pneumonia [2.57, 95% confidence interval
(1.514.35); p < 0.001] (Table 3).

DISCUSSION
Our study shows an increase in suppurative complications
of pneumonia (pleural empyema and lung abscesses)
between 1995 and 2003 in two hospitals in France. The
prevalence of suppurative complications amongst children

864

Antibiotics
Amino-penicillin
Cephalosporin
Macrolide
Other antibiotics
Anti-inflammatory
Ibuprofen
Aspirin
Glucocorticoids
Other

Univariable analysis
OR (95% CI)

Multivariable analysis
p

2.07 (1.293.32) <0.01

1.35 (0.782.33) 0.28
1.71 (0.883.32) 0.12
1.63 (0.465.79) 0.45

aOR (95% CI)

1.57 (0.912.72)
1.24 (0.672.30)
1.26 (0.582.73)
2.19 (0.539.14)

0.11
0.49
0.56
0.28

3.35 (2.085.38) <0.0001 2.57 (1.514.35) <0.001

0.74 (0.291.90) 0.53
1.72 (0.694.99) 0.31
1.31 (0.602.87) 0.50
1.41 (0.583.41) 0.44
2.20 (0.716.8)
0.17
2.41 (0.688.56) 0.17

aOR: odds ratios were adjusted for gender, age, study year, fever duration, and
number of medical consultations.

hospitalized with community-acquired pneumonia rose

from 2.8% in the 19951998 period to 23.1% in 2003, following a linear growth pattern since 1999. The corresponding annual populational incidence of complicated
pneumonia increased from 0.5 to 13 cases per 100 000 children between 1995 and 2003.
Our findings are consistent with previous studies conducted in Western countries (110). North-American and
British studies showed that this trend appeared in the middle of the 1990s, from 1994 to 1996. Our study suggests that
this trend is more recent in France, beginning in 1999. It
should be noted that two studies in the United States
showed that the pleural effusion incidence reached a maximum in 2000 and has decreased since 20012002 (4,5). The
authors attribute this decrease to paediatric vaccination
with 7-valent pneumococcal conjugate vaccine (13,14).
However, this hypothesis was contradicted by a study
reported by Byington et al. concerning all children hospitalized for invasive pneumococcal infection in Utah, from
1997 to 2003 (15,16). Our study does not contribute information here because in France vaccination with Prevnar
(Wyeth Lederle Vaccines Sa, Philadelphie, PA, USA) has
been recommended only since March 2002.
A noteworthy finding of our study was the significant
relationship between the use of ibuprofen and complicated
pneumonia. This was in accordance with the observation of
a comparable independent association reported by Byington et al. [adjusted OR = 4 (2.56.5)]. Other studies also
suggest that ibuprofen may be involved in necrotizing fasciitis caused by streptococcus A bacteria (17,18). In a controlled study of children with chickenpox, Zerr et al.
showed a significant relationship between ibuprofen administration and necrotizing fasciitis occurrence (19). Lesko
et al. reported a significant increase for invasive contagion
with streptococcus A bacteria in children who received ibuprofen but not in children who received acetaminophen
(20). In addition, the increase in complicated pneumonia
incidence observed in our study was highly correlated with
the changes in annual sales of paediatric forms of ibuprofen
(drinkable solution) in France (r = 0.94; p < 0.001) (Fig. 2),

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866

Francois et al.

which is an estimate of the exposure of French children to

this drug.
As a result of the methodological characteristics of our
study, the apparent association between the preadmission
use of ibuprofen and complicated pneumonia should be
interpreted with caution. As a result of our retrospective data
collection method, we cannot ensure that ibuprofen administration predated the complication onset and a protopathic
bias cannot be excluded. Indeed, children with severe pneumonia might have a higher fever in the initial stages and
develop pleural pain (10) thus making the use of analgesic
and antipyretic drugs more likely. However, this bias could
relate to all anti-inflammatory drugs including aspirin and
glucocorticoids. This was not the case and argues in favour
of a specific effect of ibuprofen. A prospective study powered
to address this specific question is needed.
Owing to the low number of isolated bacteria, our study
cannot document the main hypothesis described in the literature, based on the possible change in bacteria distribution
favouring more virulent strains. In isolating S. pneumoniae
in blood or pleural effusion, Tan et al. found that serotype 1
was more frequent in complicated pneumonia (24.4% vs.
3.6%) (3). The same predominance of serotype 1 pneumococcus in complicated pneumonia was shown in studies
reported by Byington et al. in the U.S. and Eastham et al. in
the UK (2,8). In addition, we know that serotype 1 is more
often associated with invasive pneumococcal infections
(21,22). We note as well that there was no association
between penicillin-resistant S. pneumoniae and suppurating
complications (2,3,8). In our study, we found only one
infection attributable to S. aureus but three caused by streptococcus A bacteria, two of which resulted in death, which
could confirm the specific aggressiveness of this bacterium
(23). We note that the penicillin-sensitive S. pneumoniae
ratio is the same in the two groups and that serotype 1 is
also more frequent in complicated forms than in the simple
form. Yet, the modifications of distribution bacteria species
and strain involved in pneumonia and empyema vary
depending on the study, even within a country, making it
extremely difficult to isolate a consistent explanation for the
increase in empyema incidence.

CONCLUSION
Consistent with previous studies, we found that medical
treatments with NSAIDs, especially ibuprofen, prior to hospital admission are associated with increased odds of suppurative complications amongst children hospitalized with
community-acquired pneumonia. Although further study is
warranted to confirm this hypothesis, we recommend caution in the use of ibuprofen for children with suspected
pneumonia.

ACKNOWLEDGEMENTS
The authors thank Dr. Philippe Cavalie, AFSSAPS (French
Agency for Safety of Health Care Products), for his help in
drug sales data collection.

Suppurative complications of pneumonia

COMPETING INTERESTS
All authors certify that they have no commercial associations that might pose a conflict of interests in connection
with this article.

References
1. Hardie W, Bokulic R, Garcia VF, Reising SF, Christie CD.
Pneumococcal pleural empyemas in children. Clin Infect Dis
1996; 22: 105763.
2. Byington CL, Spencer LY, Johnson TA, Pavia AT, Allen D,
Mason EO, et al. An epidemiological investigation of a sustained high rate of pediatric parapneumonic empyema: risk
factors and microbiological associations. Clin Infect Dis
2002; 34: 43440.
3. Tan TQ, Mason EO, Wald ER, Barson WJ, Schutze GE, Bradley JS, et al. Clinical characteristics of children with complicated pneumonia caused by Streptococcus pneumoniae.
Pediatrics 2002; 110: 16.
4. Buckingham SC, King MD, Miller ML. Incidence and etiologies
of complicated parapneumonic effusions in children, 1996 to
2001. Pediatr Infect Dis J 2003; 22: 499504.
5. Schultz KD, Fan LL, Pinsky J, Ochoa L, Smith EO, Kaplan
SL, et al. The changing face of pleural empyemas in children: epidemiology and management. Pediatrics 2004; 113:
173540.
6. Rees JHM, Spencer DA, Parikh D, Weller P. Increase in incidence of chilhood empyema in West Midlands, UK. Lancet
1997; 349: 402.
7. Gupta R, Crowley S. Increasing paediatric empyema admissions. Thorax 2006; 61: 17980.
8. Eastham KM, Freeman R, Kearns AM, Eltringham G, Clark J,
Leeming J, et al. Clinical features, aetiology and outcome of
empyema in children in the north east of England. Thorax
2004; 59: 5225.
9. Spencer DA, Iqbal SM, Hasan A, Hamilton L. Empyema thoracis is still increasing in UK children. BMJ 2006; 332: 1333.
10. Lahti E, Peltola V, Virkki R, Alanen M, Ruuskanen O. Development of parapneumonic empyema in children. Acta Paediatr
2007; 96: 168692.
11. Ramphul N, Eastham KM, Freeman R, Eltringham G,
Kearns AM, Leeming JP, et al. Cavitory lung disease complicating empyema in children. Pediatr Pulmonol 2006; 41:
7503.
12. Desrumaux A, Francois P, Pascal C, Cans C, Croize J, Gout
JP, et al. Epidemiology and clinical characteristics of childhood parapneumonic empyemas. Arch Pediatr 2007; 14:
1298303.
13. Black SB, Shinefield HR, Hansen J, Elvin L, Laufer D,
Malinoski F. Postlicensure evaluation of the effectiveness of
sevent valent pneumococcal conjugate vaccine. Pediatr Infect
Dis J 2001; 20: 11057.
14. Hsu K, Pelton S, Karumuri S, Heisey-Grove D, Klein J. Population-based surveillance for childhood invasive pneumococcal
disease in the era of conjugate vaccine. Pediatr Infect Dis J
2005; 24: 1723.
15. Byington CL, Samore MH, Stoddard GJ, Barlow S, Daly J, Korgenski K, et al. Temporal trends of invasive disease due to
Streptococcus pneumoniae among children in the intermountain West: emergence of nonvaccine serogroups. Clin Infect Dis
2005; 41: 219.
16. Byington CL, Korgenski K, Daly J, Ampofo K, Pavia A, Mason
EO. Impact of the pneumococcal conjugate vaccine on pneumococcal parapneumonic empyema. Pediatr Infect Dis J 2006;
25: 2504.

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866

865

Suppurative complications of pneumonia

Francois et al.

17. Holder EP, Moore PT, Browne BA. Nonsteroidal anti-inflammmatoty drugs and necrotising fasciitis. An update. Drug Saf
1997; 17: 36973.
18. Choo PW, Donahue JG, Platt R. Ibuprofen and skin and soft
tissue superinfections in children with varicella. Ann Epidemiol
1997; 7: 4405.
19. Zerr DM, Alexander ER, Duchin JS, Koutsky LA, Rubens CE.
A case-control study of necrotizing fasciitis during primary varicella. Pediatrics 1999; 103: 78390.
20. Lesko SM, OBrien KL, Schwartz B, Vezina R, Mitchell AA.
Invasive group A streptococcal infection and nonsteroidal antiinflammatory drug use among children with primary varicella.
Pediatrics 2001; 107: 110815.

866

21. Kaplan SL, Mason EO, Wald E, Tan TQ, Schutze GE, Bradley
JS, et al. Six year multicenter surveillance of invasive pneumococcal infections in children. Pediatr Infect Dis J 2002; 21:
1417.
22. Haudorff WP, Bryant J, Paradiso PR, Siber GR. Which pneumococal serogroups cause the most invasive disease: implication for conjugate vaccine formulation and use, part I. Clin
Infect Dis 2000; 30: 10021.
23. Al-Kaabi N, Solh Z, Pacheco S, Murray L, Gaboury I, Le Saux
N. A comparison of group A Streptococcus versus Streptococcus pneumoniae pneumonia. Pediatr Infect Dis J 2006; 25:
100812.

2010 The Author(s)/Journal Compilation 2010 Foundation Acta Pdiatrica/Acta Pdiatrica 2010 99, pp. 861866