ARTICLE
Epidemiologic Features of Invasive Pneumococcal
Disease in Belgian Children: Passive Surveillance Is
Not Enough
Anne Vergison, MDa, David Tuerlinckx, MD, PhDb, Jan Verhaegen, MD, PhDc, Anne Malfroot, MD, PhDd, for the Belgian Invasive Pneumococcal Disease Study Group
aDepartment of Pediatric Infectious Diseases, Infection Control and Hospital Epidemiology Unit, Universite´ Libre de Bruxelles, Hoˆpital Universitaire des Enfants Reine
Fabiola, Brussels, Belgium;bDepartment of Pediatrics, Universite´ Catholique de Louvain, Cliniques Universitaires Mont Godinne, Mont Godinne, Belgium;cDepartment of
Microbiology, National Reference Laboratory for Pneumococci, Katholieke Universiteit van Leuven, Leuven, Belgium;dDepartment of Pediatrics, Pediatric Respiratory
Medicine, Infectious Diseases and Cystic Fibrosis Clinic, Academisch Ziekenhuis-Vrije Universiteit, Brussels, Belgium
The authors have indicated they have no financial relationships relevant to this article to disclose.
ABSTRACT
BACKGROUND.Reliable epidemiologic surveillance of infectious diseases is important
for making rational choices for public health issues such as vaccination strategies. In Belgium, as in most European countries, surveillance relies on voluntary passive reporting from microbiology laboratories; therefore, reported incidence rates are probably inaccurate.
METHODS.We conducted national, active, laboratory-based and clinically based
sur-veillance of invasive pneumococcal disease in young children.
RESULTS.During the study period, the incidences of invasive pneumococcal disease
in children⬍2 years of age (104.4 cases per 105person-years and 16.1 cases per 105person-years for invasive pneumococcal disease and meningitis, respectively) and in children 0 to 59 months of age (59.5 cases per 105person-years for invasive pneumococcal disease and 7.7 cases per 105 person-years for meningitis) were twice those reported previously through the passive surveillance system. Overall, 67% of theStreptococcus pneumoniaestrains isolated from children⬍5 years of age belonged to 7-valent pneumococcal conjugate vaccine serotypes and 18% to vaccine-related serotypes (mainly serotype 19A). Erythromycin resistance was frequent, especially among children⬍2 years of age (59%).
CONCLUSIONS.Under-reporting can explain the reported low incidence of invasive
pneumococcal disease in countries (such as Belgium) that depend on a passive epidemiologic surveillance system, which could lead to erroneous choices in vaccination policies. There is a need for an active system of epidemiologic surveil-lance for vaccine-preventable diseases such as invasive pneumococcal disease, at the national or European level.
www.pediatrics.org/cgi/doi/10.1542/ peds.2005-3195
doi:10.1542/peds.2005-3195
Key Words
epidemiology,Streptococcus pneumoniae, surveillance
Abbreviations
IPD—invasive pneumococcal disease PCV-7—7-valent pneumococcal conjugate vaccine
MIC—minimal inhibitory concentration
Accepted for publication Mar 27, 2006
Address correspondence to Anne Vergison, MD, Department of Pediatric Infectious Diseases, Infection Control and Hospital Epidemiology Unit, Universite´ Libre de Bruxelles, Hoˆpital Universitaire des Enfants Reine Fabiola, 15, Avenue JJ Crocq, 1020 Brussels, Belgium. E-mail: anne.vergison@ulb. ac.be
K
NOWLEDGE OF THE Belgian epidemiologic features of invasive pneumococcal disease (IPD) relies on voluntary reporting and the sending ofStreptococcus pneu-moniae strains to the national reference laboratory by local microbiology laboratories. The incidence of IPD reported from 1996 to 2000 (30 cases per 105child-years among children⬍5 years of age)1was far below the USincidence (96.7 cases per 105 child-years in 1998 and 1999).2Under-reporting is assumed to explain the lower
incidence of IPD in European countries, together with a higher threshold for blood culture drawing.3,4Local
ep-idemiologic data on age, serotype distribution, and anti-biotic resistance are important for guidance of vaccina-tion strategies and for analysis of the impact of immunization programs. Universal immunization of all children ⬍2 years of age with 7-valent pneumococcal conjugate vaccine (PCV-7) has been implemented in the United States, whereas vaccination has been targeted to children with risk factors in most European countries.
In Belgium, data regarding serotype distribution were not available. Recent data suggested an association of a specific serotype with either nasopharyngeal coloniza-tion or IPD, with some serotypes being more likely to cause specific pediatric infections such as pneumonia.5
Moreover, determination of serotype distribution is im-portant for estimation of the fraction of IPD cases that could be prevented with PCV-7. We wanted to assess the real burden of IPD in children ⬍5 years of age in Bel-gium, compared with the existing national laboratory-based surveillance data, to calculate the theoretical cov-erage of PCV-7 according to our local serotype distribution, and to establish the antibiotic susceptibility profiles of theS pneumoniaestrains recovered from these pediatric IPD cases.
METHODS
Study Design
Active surveillance was established in 128 Belgian hos-pitals with pediatric wards (covering 98.5% of all pedi-atric wards) for 1 year, from March 18, 2002, to March 17, 2003. Eligible cases needed to meet the following inclusion criteria: (1) age from birth to 59 months, (2) isolation of S pneumoniae from a normally sterile body site (eg, blood or cerebrospinal fluid) by the hospital microbiology laboratory, and (3) signing of written in-formed consent forms by the child’s parents or guard-ians. The local ethics committees approved the study. Case identification was based on 2 independent, active surveillance approaches, one clinical (hospital based) and one microbiologic (laboratory based).
Clinical Data
Case report forms were completed by local pediatricians and were recorded on a Web site directly by the physi-cian or by the data manager. The data manager was
aware of all strains of pneumococci sent to the reference laboratory and eventually could confirm that clinical data had been sent for all strains; she could also check that the local laboratories had sent all strains to the national reference laboratory. Recall messages were sent monthly to all participating pediatricians and microbiol-ogists, to ensure good participation in the study. We defined the following clinical entities: (1) meningitis confirmed by positive cerebrospinal fluid culture, (2) septicemia with septic shock, (3) bacteremia without a major focus of infection, (4) pneumonia with pleural effusion with a positive blood culture and/or positive pleural fluid culture, (5) pneumonia (confirmed by chest radiograph) with a positive blood culture, and (6) other (eg, arthritis or peritonitis).
Strain Collection
All pneumococcal isolates from the local laboratories were sent to the National Reference Laboratory for Pneumococci. Identification was confirmed with stan-dard methods, capsular serogroups were determined with the quellung reaction with 46 group sera from the Staten Serum Institute (Copenhagen, Denmark), and some isolates were sent to the Staten Serum Institute for serotyping. Antimicrobial susceptibility tests for penicil-lin, erythromycin, tetracycline, and ofloxacin were per-formed with the disk diffusion method. For strains with reduced susceptibility to penicillin, minimal inhibitory concentrations (MICs) were determined with the E-test (AB Biodisk, Stockholm, Sweden) for penicillin and ce-fotaxime, according to National Committee for Clinical Laboratory Standards methods.6
Statistical Analyses
In Belgium, birth rates are fairly stable (ranging from 111 225 to 114 883 births per year for 2000 –2003). Age-specific incidences were calculated by using a birth cohort of 115 000 children per year. Incidences are pre-sented with 95% confidence intervals (calculated from counts in the Poisson distribution). Statistical analyses were performed with SPSS software (version 12.0; SPSS, Chicago, IL) (Pearson test and Fisher’s exact test for expected counts of ⬍5; median comparisons with the Mann-Whitney test, for non-Gaussian distributions). The-score test was used to compare annual incidences, with the assumption of Poisson distribution.
Role of the Funding Source
RESULTS
Clinical and Epidemiologic Data
A total of 342 patients with IPD (57% male) were re-ported during the study period, of whom 330 (96.5%) were admitted to a hospital. No cases were registered in 38 hospitals, 1 or 2 cases in 49 hospitals, 3 to 8 cases in 32 hospitals, andⱖ9 cases in 9 centers. The median age was 13.8 months (age range: 0 –59 months). Thirteen children (4%) were⬍3 months of age, 227 (67%) were 3 to 23 months of age, and 102 (30%) were 24 to 59 months of age.
The incidence of IPD in children⬍5 years of age was 59.5 cases per 105child-years (95% confidence interval: 52.6 – 66.7 cases per 105 child-years), and the cumula-tive risk of IPD in these children was 296.5 cases per 105 child-years (95% confidence interval: 260.5–332.6 cases per 105child-years). The annual incidence was greater for children 0 to 23 months of age (IPD: 104.4 cases per 105 child-years; 95% confidence interval: 91.2–117.6 cases per 105child-years; meningitis: 16.1 cases per 105 child-years; 95% confidence interval: 11.3–22.2 cases per 105child-years) than for children 24 to 59 months of age (IPD: 29.6 cases per 105child-years; 95% confidence interval: 24.1–35.9 cases per 105child-years; meningitis: 2.0 cases per 105child-years; 95% confidence interval: 0.8 – 4.2 cases per 105 child-years; P ⬍ .001 for both comparisons).
The distribution of the clinical entities of IPD cases for the different age groups is shown in Table 1. Pneumonia, with or without pleural effusion, was significantly more frequent in the 24- to 59-month-old group than in the
⬍2-year-old group (P ⬍.0001). There was a trend to-ward older age for children presenting with pneumonia and pleural effusion (median age of 36.5 months, com-pared with 19.5 months for pneumonia without pleural effusion;P⫽.12).
Data on previous antibiotic use were available for 291 patients (85%). A total of 157 patients (54%) had re-ceived systemic antibiotic therapy in the previous 6 months, and 35 (12%) had taken antibiotics within 48 hours before the IPD diagnosis. Of the 35 children who were receiving antibiotics at the time of IPD diagnosis,
15 (43%) received a macrolide, 9 (26%) amoxicillin, with or without clavulanate, and 5 (14%) cefaclor.
One or more underlying medical conditions was present in 7% of the cases (24 patients). Only 3 patients (0.9%) had hemoglobinopathy (sickle cell anemia), and only 2 (0.6%) had HIV infection. Five patients (1.5%) had congenital heart defects, 5 (1.5%) had congenital immunodeficiency, 5 (1.5%) had malignant hemato-logic disease, and 4 had miscellaneous conditions (2 had chronic pulmonary disease, 1 had renal insufficiency, and 1 had a cerebrospinal fluid leak).
The case fatality proportion for IPD, at 0 to 59 months of age, was 2.3%. Of the 8 children who died, 5 had meningitis, 2 septicemia, and 1 initial peritonitis with secondary meningitis and septicemia. Six of the 8 chil-dren were ⬍2 years of age, and 3 had an underlying medical condition. Of all reported IPD cases, 11 patients (3.2%), all ⬍2 years of age and with meningitis, had sequelae at the time of hospital discharge (deafness and/or neurologic deficits). Two of these children had underlying medical conditions.
Serotypes ofS pneumoniaeRecovered From IPD
Serotypes were determined for 280 of 342 strains of S pneumoniae (Table 2). The remaining 62 strains either were not sent by the hospital laboratory (41 strains, 12%) or could not be cultured after transport (21 strains). Overall, 67% of S pneumoniae strains (188 of 280 strains) belonged to PCV-7 serotypes (14, 23F, 6B, 18C, 19F, 4, and 9V), and 18% of strains (49 of 280 strains) belonged to vaccine-related serotypes (19A, 6A, 19C, 23A, and 9N). Serotype 19A was the third most frequent serotype (10%). However, serotypes were un-evenly distributed in the different age groups (Fig 1); 25%, 75%, and 57% of strains belonged to a vaccine serotype for children⬍3 months, 3 to 23 months, and 24 to 59 months of age, respectively. Strains belonging to a vaccine-related serotype were recovered from 17%, 16%, and 21% of children in the same respective age groups. Children infected with serotype 1 were signifi-cantly older (median age: 48 months) than children with IPD with another serotype (median age: 13 months;P⬍ .0001). Serotype distribution also varied according to the infection localization (Table 3). For pneumonia, there were noticeable differences in serotypes between the age groups. For children ⬍2 years of age with pneumonia without pleural effusion, serotype 14 was the most fre-quent (17 of 40 strains, 43%, compared with 5 of 30 strains, 17%, for the 2-5-year-old group; P⫽.02). For 24- to 59-month-old children with pneumonia, strains from serotype 1 were the most commonly isolated (12 of 39 strains, 31%), whereas only 1 child⬍2 years of age had pneumonia with pleural effusion attributable to se-rotype 1S pneumoniae(P⬍.001).
TABLE 1 Distribution of Clinical IPD Cases According to Clinical Presentation in Different Age Groups
IPD Type No. of Cases (%)
⬍3 mo (n⫽13)
3–23 mo (n⫽227)
24–59 mo (n⫽102)
Total (n⫽342)
Meningitis 4 (31) 33 (15) 7 (7) 44 (13) Bacteremia 7 (54) 129 (57) 45 (44) 181 (53) Pneumonia 2 (15) 48 (21) 34 (33) 84 (25) Pleural effusion 0 6 (3) 12 (12) 18 (5)
Septicemia 0 3 (1) 2 (2) 5 (2)
Antibiotic Susceptibility of the Pneumococci Isolated From IPDs
The 280 serotyped S pneumoniaestrains were tested for antibiotic susceptibility (Table 4). No strains were
resis-tant to penicillin (MIC:⬎2 mg/L), cefotaxime (MIC:⬎2 mg/L), or ofloxacin (MIC:⬎8 mg/L). Penicillin-interme-diate strains were distributed equally among the age groups (15% and 14% for ⬍2 years and 2–5 years,
TABLE 2 Serotype Distributions and Incidences ofS pneumoniaeStrains in Different Age Groups
Serotype ⬍3 mo
(n⫽12)
3–23 mo (n⫽185)
24–59 mo (n⫽83)
0–59 mo (n⫽280)
Vaccine serotypes,n
14 0 52 14 66
6B 0 43 7 50
23F 0 11 8 19
9V 1 11 4 16
19F 2 9 1 12
18C 0 7 9 16
4 0 5 4 9
Incidence of vaccine serotypes, no. of cases per 105person-years (95% CI)
11.3 (2.3–33.2) 84.1 (70.2–98.8) 16.7 (12.4–22.4) 39.9 (34.4–46.0)
Vaccine-related serotypes,n
19A 0 20 7 27
6A 0 8 6 14
18F 0 0 3 3
9N 2 0 0 2
19C 0 1 1 2
23A 0 1 0 1
Incidence of vaccine-related serotypes, no. of cases per 105person-years (95% CI)
7.5 (0.9–27.3) 18.3 (12.3–26.0) 6.1 (3.5–9.7) 10.4 (7.7–13.7)
Nonvaccine serotypes,n
1 2 1 13 16
5 1 4 2 7
7F 2 3 1 6
3 1 1 0 2
12F 0 2 0 2
22F 0 1 1 2
24F 0 0 2 2
10A, 15B, 27, 33F, 38 0 5⫻1a 0 5⫻1a
29 1 0 0 1
Incidence of nonvaccine serotypes, no. of cases per 105person-years (95% CI)
26.4 (10.6–54.5) 10.4 (5.6–16.5) 6.8 (4.1–10.6) 9.1 (6.6–12.3)
Total incidence, no. of cases per 105
person-years (95% CI)
45.2 (23.4–79.2) 112.8 (96.5–129.5) 29.6 (23.7–36.8) 59.5 (52.6–66.7)
Distributions are given as absolute frequencies. Incidences were calculated as incidence⫽n⫻105/关(birth cohort)⫻(years of surveillance)⫻(fraction of isolates collected in each age group)兴. CI
indicates confidence interval.
aOne isolate for each listed serotype.
FIGURE 1
respectively). Erythromycin-resistant strains were much more prevalent in the⬍2-year-old group (59%) than in the 2- to 5-year-old group (35%;P⬍ .001). Similarly, tetracycline resistance was higher in the youngest age group, with 45% resistance in strains isolated from chil-dren⬍2 years of age and 35% resistance in strains from
the 24- to 59-month-old group (P⫽.02). A total of 164 (59%) of 280 pneumococcal strains were intermediate or resistant to at least penicillin, erythromycin, or tetra-cycline. Twenty-four strains (9%) were intermediate or resistant to all 3 antibiotics and 91 (33%) were resistant to 2 antibiotics, mostly with combined erythromycin and
TABLE 3 Distribution of Serotypes for the Main Clinical Entities
Serotype Clinical Entities
Meningitis (n⫽33)
Pneumonia (n⫽71)
Pleural Effusion (n⫽14)
Bacteremia (n⫽152)
Vaccine serotypes,n
4 0 2 0 7
6B 9 7 1 29
9V 3 3 1 8
14 6 21 1 35
18C 2 2 1 11
19F 1 2 0 9
23F 2 7 0 9
Vaccine serotypes, % 69.7 62.0 28.6 71.1
Vaccine-related serotypes,n
6A 1 1 1 11
19A 2 6 4 15
Othera 0 2 0 6
Vaccine-related serotypes, % 9.1 12.7 35.7 21.1
Nonvaccine serotypes,n
1 1 9 3 3
5 0 4 1 2
7F 3 2 1 0
Otherb 3 5 0 13
Nonvaccine serotypes, % 21.2 28.2 35.7 11.8
Total percentages according to clinical presentation in each serotype group are presented.
aOther vaccine-related serotypes included 18F, 19C, 9N, and 23A.
bOther nonvaccine serotypes included 3, 24F, and 27 (each responsible for 1 case of meningitis), 15B, 12F, 22F, 10A, 33F, and 38.
TABLE 4 Distribution of Penicillin-, Erythromycin-, and Tetracycline-Nonsusceptible Strains Among Vaccine, Vaccine-Related, and Nonvaccine Serotypes
Serotype Total No. of
Isolates
No. of Isolates Showing Antibiotic Nonsusceptibility (%)a
Penicillin Erythromycin Tetracycline
Vaccine serotypes
14 66 24 (36) 54 (82) 31 (47)
6B 50 5 (10) 41 (82) 37 (74)
23F 19 6 (32) 4 (21) 1 (5)
9V 16 1 (6) 11 (69) 8 (50)
19F 12 2 (17) 6 (50) 4 (33)
18C 16 0 (0) 0 (0) 2 (13)
4 9 0 (0) 2 (22) 2 (22)
Subtotal 188 38 (20) 118 (63) 85 (45)
Vaccine-related serotypes
6A 14 1 (7) 2 (14) 2 (14)
19A 27 2 (7) 18 (67) 18 (67)
Other 8 0 (0) 3 (38) 2 (25)
Subtotal 49 3 (6) 23 (47) 22 (45)
Nonvaccine serotypes
1 16 0 (0) 1 (6) 5 (31)
Other 27 1 (4) 4 (15) 3 (11)
Subtotal 43 1 (2) 5 (12) 8 (19)
Total 280 42 (15) 146 (52) 115 (41)
Percentages were calculated as %⫽n⫻100/total number of isolates (Table 2).
tetracycline resistance (90%). Most of the nonsuscep-tible strains belonged to a vaccine serotype (91%, 81%, and 74% for penicillin, erythromycin, and tetracycline, respectively). Resistance also was distributed unequally among the serotypes. “Multiresistant” strains (resistant to penicillin, erythromycin, and tetracycline) were more likely from serotype 14 (17 of 24 strains, 71%), whereas 13 (68%) of 19 serotype 23F strains were fully suscep-tible.
DISCUSSION
Variations in IPD Incidences
This study stresses the impact of the quality of epidemi-ologic surveillance on IPD incidence. By conducting ac-tive surveillance for 1 year, we found that the estimated incidence rate of IPD for children⬍5 years of age dou-bled, compared with the previous estimated incidence rate calculated from data collected passively by the ref-erence laboratory (30 cases per 105child-years for IPD and 4 cases per 105 child-years for meningitis).1 Our
findings confirm the importance of under-reporting as a cause of underestimation of disease incidence for IPD in
Belgium. Disparities in reported rates of IPD were hy-pothesized to explain the variations in IPD incidence between Europe and the United States,3but such
dispar-ities had never been demonstrated. IPD incidence rates usually reported for Europe vary from⬍10 cases per 105 person-years to 24 cases per 105person-years for chil-dren⬍5 or⬍6 years of age7,8(Table 5). A few countries
(Spain, Israel, and Greece) have reported higher rates of 42 to 56 cases per 105person-years.9–12Most European
countries have reported rates from national or regional passive laboratory surveillance. With passive surveil-lance, changes in IPD incidence must be interpreted with caution, because the number of reporting laboratories varies13 (J.V., unpublished data). However, 1 German
study and 1 Austrian study found equally low incidences (9 –14 cases per 105person-years and 4 – 6 cases per 105 person-years for IPD and meningitis, respectively, in the
⬍5-year-old group) despite active prospective surveil-lance based on both laboratory and clinical reports.14,15
The meningitis incidence rate for infants⬍2 years of age (16.1 cases per 105 person-years) was comparable to rates in various European countries (Spain, Denmark, or
TABLE 5 Incidence of IPD and Meningitis in Reported European and Israeli Studies
Country or Region Reference Study
Period
Characteristics of Study
IPD Incidence, Cases per 105Person-Years
Meningitis Incidence, Cases per 105Person-Years
Type of Surveillancea
Scaleb ⬍5 y ⬍2 y ⬍5 y ⬍2 y
Finland 34 1985–1989 P N 24.2 45.3 2.1 4.7
Spain 35 2000–2001 A R 6.3 13.1
Basque and Navarre 9 1988–2001 R R 56 94 8 14
Valencia 36 1996–1998 A R 17 3.8
Switzerland 37 2000–2001 P N 13.7–21.1 23.6–30.8 1.6
38 1985–1994 R N 7.6 11.0 3.1 5.6
Germany 14 1997–2000 A N 8.9 16.3 3.7 7.2
Netherlands 13 1996–1999 P N 5.8 8.2
Denmark 19 1981–1999 P N 5.3–16.8c 4.4 12.4
Scotland 18 1988–1999 P N 28.4 11.8
39 1999–2001 P N 21–51d
England and Wales 29 1995–1997 P N 14.5 39.7e 5.5 15.7e
Nottingham 40 1990–1999 R C 22.5 42.1 6.9 15
Oxford region 32 1995–1999 P (enhanced) R 21.2 6.2 14.8e
Italy 41 1994–2002 P N 1.1
21 2001 P⫹Rf R⫹Cf 2.8–6.3 5.9–11.3 4.7–5.7
Sweden 42 1990–1999 P C 5.8 10
France 43 1991–1997 P N 40–50e
Greece 12 1995–1999 R C 43 6
Norway 17 1995–2001 P N 18.6 6.1e
Austria 15 2001–2003 A N 13.7 14.5 6 7.7
Israel 10 1988–1990 A N 42 5.4 11.1
11 1989–1998 A C 45–139g 8–23g
Belgium 1 1996–2000 P N 30 4
This study 2002–2003 A N 59.5 104.4 7.7 16.1
aType of surveillance: P indicates passive laboratory based; A, active laboratory based and clinically based; R, retrospective hospital based (medical records). bScale: N indicates nationwide; R, regional (ⱖ1 region); C, city or single hospital based.
cFor 1981 to 1999 (increase).
dLess than 1 year: 21 cases per 105person-years; 1–2 years: 51 cases per 105person-years. eLess than 1 year.
Great Britain), but it was more than twice the incidence in the United States (7.5 cases per 105 person-years) before licensure.16In contrast, the incidence rate for IPD
for children⬍5 years of age observed in this study was still much lower than the incidence reported for the United States before PCV-7 use (96.7 cases of IPD per 105 person-years).2 Along with the variation in reported
rates, differences in blood culture practice and underdi-agnosis of mild IPD have also been proposed to explain the lower IPD incidences in Europe, compared with the United States.3 When meningitis incidences are
com-pared between different countries and are comcom-pared with IPD incidences (Table 5), differences in blood cul-ture practice probably also occur among different Euro-pean countries. Some countries with relatively high meningitis incidences, compared with overall IPD inci-dences (such as Netherlands, Denmark, and England), may draw fewer blood cultures than Belgium or Spain, for example. Under-diagnosis of mild IPD cases probably occurs in Belgium, because blood cultures are not per-formed in outpatient settings. Consequently, the inci-dence rate found in this study is probably underesti-mated.
Furthermore, the incidence ofS pneumoniaeinfections varies over time; it has been increasing in past decades in several Northern European countries,17–19and changes in
dominant serotypes throughout the world have been described.5 Another explanation for a lower IPD
inci-dence, compared with the United States, was the rela-tively low rate of underlying comorbidities we found in the study (7%). In the 6-year surveillance of IPD re-ported by Kaplan et al,2025.5% of children⬍5 years of
age had underlying medical conditions. In Belgium, at-risk patients such as children with hemoglobinopathies or HIV infection undergo close follow-up monitoring; in selected centers, they were vaccinated with PCV-7, al-though it was not recommended for general use.
Important Vaccine-Related or Nonvaccine Serotypes
Serotype 19A was the third most prevalent in this study, isolated equally from infants and from older children. It was isolated much more frequently from IPD than was serotype 19F, as described for children from several countries (Israel,11 Italy,21 and France22). With
vaccina-tion implementavaccina-tion in these countries, we fear replace-ment of the vaccine serotypes by serotype 19A, because antibodies against serotype 19F have been suspected of providing poor cross-immunity against serotype 19A.23
The estimated vaccine effectiveness against serotype 19A was 41% in the analysis published in 2003 (3 years after the licensure of PCV-7 in the United States),24but recent
US surveillance data raised concerns about both protec-tion against serotype 19F and cross-protecprotec-tion against serotype 19A, with increased relative proportions of IPD caused by these serotypes.25 Moreover,
penicillin-resis-tant or multiresispenicillin-resis-tant clones of serotype 19A are
circu-lating both in Europe (Switzerland)26and in the United
States.25,27
Serotype 1 is also worthy of attention; it has been reported commonly for many European countries, in-cluding Sweden,28 Israel,10,11 England and Wales,29 and
Norway.17 In our study, it was recovered from 29% of
blood cultures from children 24 to 59 months of age with pneumonia. Serotype 1 was associated with pneu-monia but not specifically with pleural effusion. Sero-type 1 was recovered from 22% of pleural effusion cases, compared with 50% in a recent study.30For the
pneu-monia group, however, we were not able to look at complications other than pleural effusion that might be associated with serotype 1.31
There were important effects of age. Children ⬍6 months of age with meningitis were more likely to be infected with a less-prevalent serotype, which suggests that there could be partial protection by maternal anti-bodies against the more-common serotypes (such as se-rotype 14 or 6B), at least for severe infection. Infants⬍3 months of age had a different serotype pattern distribu-tion than did the 3- to 23-month-old group; however, we did not find the serotype 1 prevalence evident for neonates in England and Wales.32
Antibiotic Susceptibility
The trend toward an increase in antibiotic resistance in “Belgian”S pneumoniaewas not confirmed for penicillin, compared with previous data from the 1994 to 2000 period.33In that study, the authors found a 17.6%
pro-portion of penicillin-nonsusceptible isolates (with 11% of strains with high-level resistance) among IPD isolates from all age groups. In our study of children⬍5 years of age, the penicillin resistance rate was slightly lower, with no strains with high-level resistance. However, erythro-mycin resistance was higher in the present study (52%)
than in 2000 among children with bacteremia
(47.6%).33 Erythromycin and tetracycline resistance
rates were significantly different between children 0 to 23 months of age and those 24 to 59 months of age. These resistant pneumococci belonged mostly to the 3 most frequently isolated IPD serotypes (serotypes 14, 19A, and 6B) recovered from children⬍2 years of age, which could suggest clonal dissemination of these strains.
CONCLUSIONS
likely to increase in proportion in the future, and this should be documented. There might also be a need to adapt the vaccine serotype composition to any changing epidemiologic features over time. This work stresses the importance of standardized, reproducible, national (or European), laboratory-based and clinically based, epide-miologic surveillance.
ACKNOWLEDGMENTS
This work was supported by Wyeth Pharmaceuticals and the Belgian Society of Pediatrics.
The Belgian Invasive Pneumococcal Disease Study Group includes Jean-Paul Buts (Universite´ Catholique de Louvain), Samy Cadranel (Universite´ Libre de Brux-elles), Frans De Baets (Universitair Ziekenhuis Ghent), Philippe Lepage (Universite´ de Lie`ge), Jack Levy (Uni-versite´ Libre de Bruxelles), Anne Malfroot (Vrije Univer-siteit Brussel), Marc Raes (Hasselt), Jose´ Ramet (Univer-sitair Ziekenhuis Antwerpen), Marijke Proesmans (Katholieke Universiteit van Leuven), Etienne Sokal (Universite´ Catholique de Louvain), David Tuerlinckx (Universite´ Catholique de Louvain), Jan Verhaegen (Katholieke Universiteit van Leuven), and Anne Vergi-son (Universite´ Libre de Bruxelles).
We thank Sophie Leyman and Patricia Slachmuylders from Wyeth Pharmaceuticals for their support, and we thank Wyeth Pharmaceuticals for financial support. We thank all Belgian pediatricians and microbiologists who participated in the study and the Belgian Society and Academy of Pediatrics for scientific support of this study. We also thank Jean Vanderpas for his help in the statis-tical analysis and Mark Fletcher for cristatis-tical proofreading of the manuscript.
REFERENCES
1. Institut Scientifique de Sante´ Publique.Surveillance des Maladies Infectieuses par un Re´seau de Laboratoires de Microbiologie Tendances Epide´miologiques 1983–2002[Surveillance of Infectious Diseases by a Microbiology Laboratories Network: Epidemiologic Trends 1983–2002]. Brussels, Belgium: Institut Scientifique de Sante´ Publique; 2004. Report D/2004/2505/20
2. Centers for Disease Control and Prevention. Direct and indirect effects of routine vaccination of children with 7-valent pneu-mococcal conjugate vaccine on incidence of invasive pneumo-coccal disease: United States, 1998 –2003.MMWR Morb Mortal Wkly Rep.2005;54:893– 897
3. Hausdorff WP, Siber G, Paradiso PR. Geographical differences in invasive pneumococcal disease rates and serotype frequency in young children.Lancet.2001;357:950 –952
4. Feikin DR, Klugman KP. Historical changes in pneumococcal serogroup distribution: implications for the era of pneumococ-cal conjugate vaccines.Clin Infect Dis.2002;35:547–555 5. Hausdorff WP, Feikin DR, Klugman KP. Epidemiological
dif-ferences among pneumococcal serotypes. Lancet Infect Dis. 2005;5:83–93
6. National Committee for Clinical Laboratory Standards. Perfor-mance Standards for Antimicrobial Susceptibility Testing: Fifteenth Informational Supplement M100-S15. Wayne, PA: National Com-mittee for Clinical Laboratory Standards; 2005
7. Hausdorff WP, Bryant J, Paradiso PR, Siber GR. Which
pneu-mococcal serogroups cause the most invasive disease: implica-tions for conjugate vaccine formulation and use, part I. Clin Infect Dis.2000;30:100 –121
8. Hausdorff WP, Bryant J, Kloek C, Paradiso PR, Siber GR. The contribution of specific pneumococcal serogroups to different disease manifestations: implications for conjugate vaccine for-mulation and use, part II.Clin Infect Dis.2000;30:122–140 9. Bernaola Iturbe E, de Aristegui Fernandez J, Herranz Aguirre
M, Garcia Calvo C, Fernandez Perez C, Grupo de Estudio de Enfermedad Invasora Neumococica en el Pais Vasco y Navarra. Study of the incidence of invasive pneumococcal disease in neonates and children aged less than 5 years in the Basque country and Navarre (Spain) [in Spanish].An Esp Pediatr.2002; 57:301–309
10. Dagan R, Englehard D, Piccard E, Israeli Pediatric Bacteremia and Meningitis Group. Epidemiology of invasive childhood pneumococcal infections in Israel.JAMA.1992;268:3328 –3332 11. Fraser D, Givon-Lavi N, Bilenko N, Dagan R. A decade (1989 –1998) of pediatric invasive pneumococcal disease in 2 populations residing in 1 geographic location: implications for vaccine choice.Clin Infect Dis.2001;33:421– 427
12. Syriopoulou V, Daikos GL, Soulis K, et al. Epidemiology of invasive childhood pneumococcal infections in Greece.Acta Paediatr Suppl.2000;89(435):30 –34
13. Spanjaard L, van der Ende A, Ru¨mke H, Dankert J, Van Alphen L. Epidemiology of meningitis and bacteremia due to Strepto-coccus pneumoniaein the Netherlands.Acta Paediatr Suppl.2000; 89(435):22–26
14. Von Kries R, Siedler A, Schmitt HJ, Reinert RR. Proportion of invasive pneumococcal infections in German children prevent-able by pneumococcal conjugate vaccines.Clin Infect Dis.2000; 31:482– 487
15. Rendi-Wagner P, Georgopoulos A, Kundi M, et al. Prospective surveillance of incidence, serotypes and antimicrobial suscep-tibility ofStreptococcus pneumoniaeamong hospitalized children in Austria.J Antimicrob Chemother.2004;53:826 – 831 16. Robinson KA, Baughman W, Rothrock G, et al. Epidemiology
of invasive Streptococcus pneumoniae infections in the United States, 1995–1998: opportunities for prevention in the conju-gate vaccine era.JAMA.2001;285:1729 –1735
17. Pedersen MK, Hoiby EA, Froholm LO, Hasseltvedt V, Lermark G, Caugant DA. Systemic pneumococcal disease in Norway 1995–2001: capsular types and antimicrobial resistance. Epide-miol Infect.2004;132:167–175
18. Kyaw MH, Clarke S, Jones IG, Campbell H. Incidence of inva-sive pneumococcal disease in Scotland, 1988 –99. Epidemiol Infect.2002;128:139 –147
19. Kaltoft MS, Zeuthen N, Konradsen HB. Epidemiology of inva-sive pneumococcal infections in children aged 0 – 6 years in Denmark: a 19-year nationwide surveillance study.Acta Paedi-atr Suppl.2000;89(435):3–10
20. Kaplan SL, Mason EO Jr, Wald ER, et al. Six year multicenter surveillance of invasive pneumococcal infections in children. Pediatr Infect Dis.2002;21:141–147
21. D’Ancona F, Salmaso S, Barale A, et al. Incidence of invasive preventable pneumococcal invasive infections and blood cul-ture practices in Italy.Vaccine.2005;3:2492–2498
22. Doit C, Loukil C, Geslin P, Bingen E. Phenotypic and genetic diversity of invasive pneumococcal isolates recovered from French children.J Clin Microbiol.2002;40:2994 –2998 23. Nahm MH, Olander JV, Magyarlaki M. Identification of
cross-reactive antibodies with low opsonophagocytic activity for Streptococcus pneumoniae. J Infect Dis.1997;176:698 –703 24. Whitney CG, Farley MM, Hadler J, et al. Decline in invasive
pediatric isolates collected for antimicrobial surveillance from across the USA during 2002–2003 compared to 2000 –2001. Presented at the American Society of Microbiology 44th Inter-science Conference on Antimicrobial Agents and Chemotherapy; October 30 to November 2, 2004; Washington, DC
26. Muhlemann K, Matter HC, Tauber MG, Bodmer T. Nationwide surveillance of nasopharyngealStreptococcus pneumoniaeisolates from children with respiratory infection, Switzerland, 1998 –1999.J Infect Dis.2003;187:589 –596
27. Whitney CG. Evidence for the emergence of non-vaccine types causing invasive pneumococcal disease in the US. Presented at the American Society of Microbiology 44th Interscience Con-ference on Antimicrobial Agents and Chemotherapy; October 30 to November 2, 2004; Washington, DC
28. Hedlund J, Sorberg M, Henriques Normark B, Kronvall G. Capsular types and antibiotic susceptibility of invasive Strepto-coccus pneumoniaeamong children in Sweden.Scand J Infect Dis. 2003;35:452– 458
29. Miller E, Waight P, Efstratiou A, Brisson M, Johnson A, George R. Epidemiology of invasive and other pneumococcal disease in children in England and Wales 1996 –1998.Acta Pediatr Suppl. 2000;89(435):11–16
30. Byington CL, Spencer LY, Johnson TA, et al. An epidemiolog-ical investigation of a sustained high rate of pediatric parap-neumonic empyema: risk factors and microbiological associa-tions.Clin Infect Dis.2002;34:434 – 440
31. Tan TQ, Mason EO, Wald ER, et al. Clinical characteristics of children with complicated pneumonia caused byStreptococcus pneumoniae.Pediatrics.2002;110:1– 6
32. Sleeman K, Knox K, George R, et al. Invasive pneumococcal disease in England and Wales: vaccination implications.J Infect Dis.2001;183:239 –246
33. Flamaing J, Verhaegen J, Peetermans WE.Streptococcus pneu-moniae bacteremia in Belgium: differential characteristics in children and the elderly population and implications for vac-cine use.J Antimicrob Chemother.2002;50:43–50
34. Eskola J, Takala A, Kela E, Pekkanen E, Kalliokoski R, Leino-nen M. Epidemiology of invasive pneumococcal infections in children in Finland.JAMA.1992;268:3323–3327
35. Casado Flores J, Fenoll A, Aristegui Fernandez J, et al. Pneu-mococcal meningitis in Spanish children: incidence, serotypes and antibiotic resistance: prospective and multicentre study [in Spanish].An Esp Pediatr.2002;57:295–300
36. Diez-Domingo J, Pereiro I, Morant A, et al. Epidemiology of invasive Streptococcus pneumoniae infections in children in Spain, 1996 –1998.J Infect.2002;45:139 –143
37. Epidemiologische U¨ berwaschung der Invasiven Infektionen mit Strep-tococcus pneumoniae im Jahr 2001[Epidemiologic Surveillance of Pneumococcal Invasive Diseases in 2001]. Bern, Switzerland: Bundesamt fur Gesundheit Abtellung Epidemiologieund Infecktionkrankheiten; 2002:50 –52. Bulletin 31
38. Venetz I, Schopfer K, Mu¨hlemann K. Pediatric, invasive pneu-mococcal disease in Switzerland, 1985–1994.Int J Epidemiol. 1998;27:1101–1104
39. Kyaw MH, Christie P, Clarke SC, et al. Invasive pneumococcal disease in Scotland, 1999 –2001: use of record linkage to ex-plore associations between patients and disease in relation to future vaccination policy.Clin Infect Dis.2003;37:1283–1291 40. Ispahani P, Slack RC, Donald FE, Weston VC, Rutter N. Twenty
year surveillance of invasive pneumococcal disease in Nottingham: serogroups responsible and implications for im-munization.Arch Dis Child.2004;89:757–762
41. Pincipi N, Marchisio P. Epidemiology of Streptococcus pneu-moniaein Italian children. Acta Paediatr Suppl. 2000;89(435): 40 – 43
42. Eriksson M, Henriques B, Ekdahl K. Epidemiology of pneumo-coccal infections in Swedish children.Acta Paediatr Suppl.2000; 89(435):35–39
DOI: 10.1542/peds.2005-3195 originally published online August 7, 2006;
2006;118;e801
Pediatrics
Anne Vergison, David Tuerlinckx, Jan Verhaegen and Anne Malfroot
Passive Surveillance Is Not Enough
Epidemiologic Features of Invasive Pneumococcal Disease in Belgian Children:
Services
Updated Information &
http://pediatrics.aappublications.org/content/118/3/e801 including high resolution figures, can be found at:
References
http://pediatrics.aappublications.org/content/118/3/e801#BIBL This article cites 36 articles, 3 of which you can access for free at:
Subspecialty Collections
b
http://www.aappublications.org/cgi/collection/infectious_diseases_su
Infectious Disease
following collection(s):
This article, along with others on similar topics, appears in the
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml in its entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or
Reprints
DOI: 10.1542/peds.2005-3195 originally published online August 7, 2006;
2006;118;e801
Pediatrics
Anne Vergison, David Tuerlinckx, Jan Verhaegen and Anne Malfroot
Passive Surveillance Is Not Enough
Epidemiologic Features of Invasive Pneumococcal Disease in Belgian Children:
http://pediatrics.aappublications.org/content/118/3/e801
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.
