OUP user menu

The evolving epidemiology of invasive meningococcal disease: a two-year prospective, population-based study in children in the area of Athens

Maria N. Tsolia, Maria Theodoridou, Georgina Tzanakaki, Panayotis Kalabalikis, Evangelia Urani, Glykeria Mostrou, Anastasia Pangalis, Anthi Zafiropoulou, Corina Kassiou, Dimitris A. Kafetzis, C. Caroline Blackwell, Jenny Kremastinou, Th. E. Karpathios
DOI: http://dx.doi.org/10.1016/S0928-8244(03)00083-X 87-94 First published online: 1 May 2003


In response to an increase in the incidence in invasive meningococcal disease (IMD) due to Neisseria meningitidis, a system of hospital- and laboratory-based surveillance was used in a prospective epidemiological and clinical assessment of IMD in children 0–13 years of age hospitalized in the Athens area between 1 January 1999 and 31 December 2000. The annual incidence of laboratory-confirmed disease was 10.2/100,000. Serogroup B strains were predominant. There was a sharp decrease in serogroup C from 19% of cases in 1999 to 3% in 2000 (P=0.013). Of note was the emergence of serogroup A responsible for 7% of the cases. The overall case fatality rate was 4.5%, but 2.8% for microbiologically confirmed cases. A remarkable decrease in disease severity assessed by admissions to intensive care units was noted during the second study year. Polymerase chain reaction-based methods for detection of meningococcal DNA were the sole positive laboratory test in 45% of the cases and the only test on which serogroup determination was based in 52% of groupable cases. The epidemiological and clinical profile of meningococcal disease appears to be rapidly evolving and close monitoring is required particularly for input into decisions regarding use of meningococcal vaccines.

  • Neisseria meningitidis
  • Epidemiology
  • Greece
  • Serogroup C
  • Serogroup A
  • Polymerase chain reaction

1 Introduction

Invasive meningococcal disease (IMD) due to Neisseria meningitidis remains a serious public health problem causing significant morbidity and mortality worldwide, mainly in children and young adults. Considerable changes have been recorded during the past decade in the epidemiology of meningococcal infections. An increase in the overall incidence as well as in the number of local outbreaks due to serogroup C have been noted in many European countries and in Canada [15]. In the United States, an increase in the frequency of serogroup C outbreaks and occasional outbreaks caused by serogroup B have been observed; and, a notable increase in the proportion of serogroup Y cases was reported recently for the first time [6,7]. The first outbreak of serogroup W135 meningococcal disease was recorded during the year 2000 among pilgrims returning from the Hajj in Mecca and their close contacts in different countries [8]. In many countries a disproportionate increase in disease incidence has been noted among adolescents and young adults [4,9,10]. In parallel to these epidemiological changes, new conjugate meningococcal vaccines have been developed and immunization with a monovalent serogroup C vaccine was introduced in 1999 in the United Kingdom [11].

Serogroup B was predominant in Greece in the early 1990s but after 1995 a sharp increase in serogroup C meningococcal disease was noted. This was due to the introduction of an epidemic ET-15 clone [5,13]. This clone was identical to the C:2a: P1.2(P1.5) phenotype also responsible for the increase in serogroup C disease in the Czech Republic and Canada [5]. The peak of this increase was noted in 1997 when serogroup C accounted for 64% of the cases and serogroup B for 34.7% of them. In the following year the proportion of serogroup C isolates decreased to 43% and that of serogroup B increased to 46% [13]. The emergence of serogroup C was accompanied by a shift towards older age groups [5]. A remarkable increase in the case fatality rate (CFR) was also noted which peaked in 1997 when serogroup C disease activity was highest [13].

The aim of this 2-year prospective, population-based study was to investigate the epidemiological, clinical and laboratory characteristics of meningococcal disease in the Athens area before the introduction of the new conjugate serogroup C vaccines.

2 Subjects and methods

2.1 Study population

During a 2-year period between 1/1/1999 and 31/12/2000 active surveillance of all cases of IMD among children hospitalized in the Athens area was undertaken. In each hospital, cases were identified by pediatricians or intensive care specialists participating in the study. The microbiology laboratory records were cross-checked in each hospital by a participating microbiologist. This system was added to the laboratory-based surveillance established since 1994 by the National Meningitis Reference Laboratory (National School of Public Health, Athens, Greece) and the study was conducted in collaboration with this center.

According to data obtained from the National Statistical Service, in 1999 the midyear estimate of the population for the Athens area was 3.5 million and there were 491,334 children under the age of 14. There are two large tertiary children's hospitals in Athens, Aghia Sophia and P. and A. Kyriakou, as well as the smaller Penteli Children's Hospital. Seriously ill patients from central Greece and other parts of the country are referred to the two large children's hospitals. A small number of children with milder symptoms of IMD are treated by a few pediatric departments within public or private general hospitals.

A standard questionnaire was filled for each case including demographic data, patient's symptoms and physical findings, physician's evaluation of patient's general condition and of predominant clinical presentation, laboratory findings, complications and outcome. The study was approved by the ethics committees of the participating hospitals.

Confirmed cases were defined by the isolation of N. meningitidis from blood, cerebro-spinal fluid (CSF) or by positive polymerase chain reaction (PCR) test in blood or CSF (category I). Cases were defined as probable and were also included if Gram-negative diplococci were detected in the Gram-stain from the skin lesion or the CSF or if the antigen detection test was positive in the CSF (category II). Cases with symptoms and signs suggestive of meningococcal disease but not supported by microbiological evidence were defined as suspected and were included only if the patient had a typical hemorrhagic rash and hypotension requiring pharmacological support (category III). Although other microorganisms can rarely cause these clinical manifestations, it was decided to include these patients in the surveillance since they usually represent cases of IMD. Category III cases were analyzed separately and were not included in the incidence calculation.

2.2 Clinical laboratory evaluation

Meningitis was defined by: (1) the presence of more than 10 white blood cells (WBC) mm−3 in the CSF and/or (2) evidence of central nervous system (CNS) invasion by microorganisms such as a positive CSF culture, PCR, Gram-stain or antigen detection test. Purpura was defined by the presence of ecchymoses (petechiae >3 mm). Arthritis was considered to be present when there was pain in a joint accompanied by restriction in the range of motion and swelling. Pneumonia was diagnosed by the presence of a lobar or segmental infiltrate in the chest X-ray. A patient was considered to have hypotension if blood pressure was below the fifth percentile for age.

A latex agglutination method (Pastorex®, Sanofi Diagnostics, Pasteur, France, or Murex Wellcogen Bacterial Antigen Kit, Abbott, UK) was used for antigen detection in the CSF. The C-reactive protein (CRP) was measured with a quantitative immunoturbidimetric assay (Roche Diagnostics, Indianapolis, IN, USA)

All meningococcal strains cultured from patients in each hospital were sent to the National Meningitis Reference Laboratory for further identification and analysis. Blood and CSF specimens obtained from patients with signs and symptoms suggestive of IMD were examined for N. meningitidis DNA by the reference laboratory.

2.3 Bacterial isolates

Bacterial strains isolated from the blood or CSF of patients with meningococcal disease were cultured on chocolate agar and grown at 37°C in the presence of 5% CO2. The suspected meningococcal colonies were characterized by Gram-stain, oxidase test and rapid carbohydrate utilization test (Gallerie Pasteur; Pasteur Merieux, France).

2.4 Phenotypic characterization of meningococcal isolates

Serogroups were determined by slide agglutination with polyclonal antisera to serogroups A, B, C, W135, X, Y and Z (Wellcome Diagnostics). Serotypes and subtypes were determined by whole-cell enzyme-linked immunoassay with monoclonal antibody reagents supplied by RIMV (Bilthoven, The Netherlands) [14,15].

2.5 PCR method

The isolation of DNA in CSF samples was carried out by adding 150 µl of CSF sample obtained after centrifugation at 1600×g (Hettich, EBA-12/1412) for 4 min to 150 µl of Cheelex detergent and 650 µl of distilled water. The samples were boiled for 30 min and centrifuged again at 10,000×g for 8 min.

For the DNA isolation in whole blood samples, the nucleic acid extraction kit was used (ISOQUICK ORGA, USA).

An amount of 10 µl of the sample DNA was used for the PCR assay. For the identification of N. meningitidis the IS1106 gene was used. For serogroup prediction (A, B, C, W135 and Y) the oligonucleotides in the siaD gene (serogroups B, C, W135 and Y) and in orf-2 of a gene cassette required for the biosynthesis of the serogroup A capsule were used (Table 1) [16].

View this table:
Table 1

Oligonucleotides used in the present study

SequenceGene amplified (serogroup)Amplicon length (bp)

In each assay, the final reaction mixture (50 µl) contained 10 µl of each DNA sample, 75 mM Tris–HCl, 20 mM (NH4)2SO4, 0.01% (v/v) Tween 20, 1.5 mM of MgCl2, 50 µM each deoxynucleotide triphosphate, the corresponding oligonucleotides (Table 1) at 1.0 µM and 1 U of Taq Polymerase (AB gene).

The PCR assays were performed in a DNA thermal Cycler (Robocycler, Stratagene) with parameters described by Taha [16]. The amplicons were analyzed on a standard 2% agarose gel. A negative control consisting of distilled water and positive strain samples (for the genus and serogroups A, B, C, W135 and Y) were always included in each test.

The sensitivity and specificity of the PCR method used was evaluated at the Reference Laboratory based on the examination of 1622 specimens collected from 1041 patients over a four year period between 1/1/1998 and 31/12/2001 [17]. The sensitivity of the test was estimated at 98.5% based on the examination of specimens obtained from patients with IMD confirmed by culture. The specificity of the method was found around 96% based on the examination of specimens obtained from children with cultures positive for other bacteria, patients with viral meningitis who received no treatment and children with no signs and symptoms suggestive of IMD. High-performance characteristics have also been found by other centers using similar PCR methods [16,18].

2.6 Statistical analysis

Continuous variables were compared by the Student's t-test and categorical variables with the chi-square or Fisher's exact test. A two-tailed P value <0.05 was considered significant.

3 Results

3.1 Age, sex and seasonal distribution of the cases

A total of 157 children were hospitalized with IMD in the Athens area, 64 in 1999 and 93 in 2000. The male to female ratio was 1.3:1. Of all cases, 132 (84%) were category I, 13 (8.3%) category II and 12 (7.7%) category III. All category III patients were admitted to ICU with fever, purpura and hypotension; four required mechanical ventilation and three died within 45 min to 4 h from admission to ICU. Eight of these patients had been treated with antibiotics before admission. Material for PCR examination was available for only two children who had received parenteral antibiotics, and no meningococcal DNA was detected. Most (110/157; 70%) cases occurred between December and April with the peak in March and April (63/157; 40%).

3.2 Incidence of IMD by age group

Age was recorded for 151/157 (96.2%) patients and the median was 54 months (range 1–175 months). Only 11/151 (7.3%) of the patients were less than 1 year old, 67/151 (44.4%) were between 1 and 4 years of age, 37/151 (24.5%) were between 5 and 9 years of age and 32/151 (21.2%) between 10 and 13 years of age (Fig. 1).

Figure 1

Age distribution of patients hospitalized with IMD in the area of Athens.

Of the 145 confirmed cases of IMD (categories I and II), a total of 100 occurred in children 0–13 years of age who were permanent residents of the Athens area, and the annual incidence was estimated at 10.2 cases/100,000/year. The incidence was 6.5/100,000 in 1999 and much higher at 13.8/100,000 during the following year (P<0.0001, adds ratio (OR) 2.1, 95% confidence interval (CI) 1.396–3.235). The incidences in the following age ranges were: 8.3/100,000 less than 1 year; 16.6/100,000, 1–4 years, 8.14/100,000, 5–9 years; 6/100,000, 10–13 years. The highest age-specific incidence was noted in the fourth year of life when it approached 21.3 cases/100,000 children. There was a second lower peak at 11 years, 10/100,000. The overall incidence in the Athens area was 10.7/100,000 if the clinical cases without laboratory confirmation were included.

3.3 Clinical characteristics and complications

The clinical characteristics of the patients are shown in Table 2. A considerable proportion of patients (32% in 1999 and 41.5% in 2000, P=0.263) did not appear septic to the examining physician at first evaluation on admission. The occurrence of IMD was suspected — because of the presence of a hemorrhagic rash and/or meningeal signs. Of the 31 patients with meningitis and a CSF WBC count above 500 cells mm−3 whose initial evaluation was recorded 20 (67%) appeared septic, five (17%) were moderately sick and six (20%) did not have a toxic appearance.

View this table:
Table 2

Clinical and laboratory characteristics of 157 patients with IMD

CharacteristicValue (%)
Symptoms and signs
History of fever135/135 (100)
median Tmax, median duration39°C, 20 h
duration >24 h, duration >48 h31/135 (23), 22/135 (16)
Rash117/150 (78)
hemorrhagic110/150 (73)
maculopapular7/150 (4.7)
Meningitis101/136 (74.3)
Headache81/117 (69)
Vomiting100/134 (75)
Nuchal rigidity59/136 (43.4)
Seizures5/157 (3.2)
Arthritis4/157 (2.5)
Pneumonia1/157 (0.6)
Diarrhea, abdominal pain5/157 (3.2)
Purpura18/157 (11.5)
Hypotension37/157 (23.6)
Admitted to ICU57/157 (36.3)
Mechanical ventilation21/157 (13.4)
Case ascertainment
Bacteremia42/157 (26.7)
Meningitis33/157 (21)
Bacteremia and meningitis70/157 (44.6)
Occult bacteremia7/157 (4.5)
Not evaluable–not recorded5/157 (3.2)
Clinical appearance on admission
Septic45/141 (31.9)
Moderately sick43/141 (30.5)
Not appearing ill53/141 (37.6)
Dead on arrival1/157 (0.6)
CSF cell count in patients with meningitis
>500 mm−336/136 (26.5)
100–500 mm−316/136 (11.8)
10 to <100 mm−331/136 (22.8)
<10 mm−36/136 (4.4)
Peripheral blood WBC count
Median, range12,500 (1500–50,100)
WBC count <5000 mm−313/153 (8.5)
     >15,000 mm−360/153 (39.2)
CRP median (range) (mg l−1)89 (0–338)
CRP <20 mg l−130/127 (23)
  • WBC, white blood count; CRP, C-reactive protein

  • One patient with erythema multiforme.

  • A lumbar puncture was performed in 136 cases.

  • One patient underwent laparotomy.

  • After the results of laboratory tests were available. The diagnosis of meningitis was based on the presence meningeal signs in patients who did not have a lumbar puncture.

  • Normal CSF cell count but positive culture (2), PCR (3) or Gram-stain (1).

During the 2 study years, the overall CFR was 7/157 (4.5%). Three of the fatal cases were caused by serogroup B (CFR 5%), one by serogroup A (CFR 14%) and in three the serogroup was not determined because the culture was negative and blood or CSF samples were not available for examination by PCR. CFR was 4/145 (2.8%) if only the laboratory-confirmed cases (categories I and II) were included.

Although the incidence of IMD was higher during 2000, the disease was much less severe. Six of the seven deaths occurred in 1999 and the CFR was 9.4% but only 1% during the following year (P=0.019, OR 9.5, 95% CI 1.117–81.08). In 1999, 40/64 (62.5%) patients were admitted to ICU; 27/64 (44%) had hypotension and 16/64 (25%) required mechanical ventilation. In 2000, 17/93 (18.3%) were admitted to ICU; 10/93 (10.75%) had hypotension and 5/93 (5.4%) required mechanical ventilation. All these differences were highly significant (P<0.0001). If only the laboratory-confirmed cases were taken into account, the proportions of patients who were admitted to the ICU who developed hypotension or required mechanical ventilation were still significantly higher in 1999 compared to 2000 (P≤0.001); however, the CFR did not differ significantly (P=0.146) between the 2 years.

Long-term complications recorded in patients admitted during 1999 included chronic renal failure in one and skin necrosis requiring skin grafts without amputations in two patients. During the second year hearing deficit was noted in two children of which one required cochlear transplant.

3.4 Laboratory findings

The results of simple laboratory tests are summarized in Table 2.

Overall the diagnosis was confirmed with a blood and/or CSF culture in 50/157 (32%) cases. Blood cultures were obtained from all patients and 33 (21%) were positive. The rate of blood culture positivity was 31% and 14% in 1999 and 2000, respectively. Positive CSF culture was found in 22/136 (16.2%) lumbar punctures and in 22/89 (25%) cases with any abnormality in the CSF. Peripheral blood PCR was positive in 100/123 (81.3%) samples examined. Blood or CSF PCR was the only laboratory test that confirmed the diagnosis of IMD in 65/145 (44.8%) of the cases in categories I and II. The proportion of laboratory-confirmed cases diagnosed solely based on this test was 17/54 (31%) in 1999 and 48/91 (52.7%) in 2000 (P=0.013).

Serogroup determination was carried out for 130/157 (82.8%) cases and 105/130 (80.8%) were groupable. Serogroup was determined by PCR in 55/105 (52.4%), by culture in 25/105 (23.8%) and by both methods in 25/105 (23.8%) cases. Of the 105 groupable isolates, 78 (74.3%) belonged to serogroup B, 10 (9.5%) to serogroup C, seven (6.7%) to serogroup A, seven (6.7%) to serogroup W135 and three (2.8%) to serogroup Y. The serogroup distribution for each of the 2 study years is shown in Fig. 2. The proportion of serogroup C strains was much lower in 2000 (2/63, 3.2%) compared with the previous year (8/42, 19%) among the groupable strains (P=0.013). There was an increase in serogroup B disease that accounted for 31/42 (73.8%) and 47/63 (74.6%) of the groupable strains in 1999 and 2000, respectively. Serogroup A strains increased from 1/42 (2.4%) in 1999 to 6/63 (9.5%) in 2000 (P=0.150).

Figure 2

Serogroup distribution of 130 N. meningitidis strains causing IMD between 1999 and 2000. NG, non-groupable.

A total of seven out of 157 (4.5%) cases of occult meningococcal bacteremia were diagnosed by PCR and otherwise would have not been recognized since they all had negative blood cultures. The PCR was used in these febrile patients because IMD was suspected but treatment was not initiated because the index of suspicion was too low. A lumbar puncture was performed in three out of seven patients and was normal. A repeat PCR was carried out in three patients and was negative.

3.5 Serotypes and serosubtypes

The phenotypic characteristics (serogroups, serotypes, subtypes) were determined for 44 meningococcal strains sent to the National Reference Center for analysis. The results are shown in Table 3.

View this table:
Table 3

Antigenic phenotypes of 44 isolates from children with meningococcal disease

Antigenic phenotypeNo. of isolates

4 Discussion

As shown by this study remarkable changes occurred in the epidemiology of IMD over a relatively short period of time in the Athens area. These findings provide additional evidence that the epidemiology of these infections is rapidly evolving and often unpredictable. The profile of disease described in children accurately depicts the features of this infection in the entire population. About 75% of all IMD cases recorded nationally in recent years were children 0–13 years of age (National Reference Laboratory, unpublished data, 1998–2001).

After the recent increase in the incidence of serogroup C disease noted between 1995–1998 in the area [5], a sharp decrease was recorded in the proportion of these cases during the study years to well below the baseline levels recorded in the early 1990s [12]. This rapid decrease represents a spontaneous change in disease epidemiology and cannot be attributed to immunization. The unconjugated polysaccharide vaccine against serogroups A and C became available for use on an individual basis for a short period of time between September 1998 and February 1999, but the proportion of children 0–14 years that were vaccinated was not higher than 12% (personal communication, Pasteur Merieux vaccine distributors). This immunization rate is too low to have a considerable impact on herd immunity. In addition, the effect of polysaccharide vaccine on the carrier state and on the induction of herd immunity is at least controversial and has not been established [1923]. The decline in serogroup C disease observed might be attributed to the increase in herd immunity after circulation in the community of the new ET15/37 clone in previous years. An increase in herd immunity against the C:2a: P1.2(P1.5) and B:2a: P1.2(P1.5) phenotypes was shown in a recent seroepidemiological study from the Czech Republic due to the spread of the same clone [24] in this country.

Epidemic and hyperendemic serogroup C disease can last for several years and its termination cannot be predicted. As a consequence, studies of meningococcal vaccine efficacy must be controlled and immunized and unimmunized groups compared at the same time. Decrease in disease incidence caused by one serogroup over time might not be due to vaccination [25,26].

In contrast to the decrease in meningococcal C disease observed in Greece, the activity of IMD caused by this serogroup continues to rise in the UK. Although serogroup C disease decreased considerably in children and adolescents vaccinated with the newly introduced monovalent conjugate vaccine, the number of cases recorded in older unvaccinated age groups was higher during the year 2000 [11].

The decrease in serogroup C disease did not result in a decrease in the overall IMD activity. Disease incidence was much higher in the second year, however, the increased use of PCR in diagnosis might have contributed to the observed increase in incidence. Serogroup B predominated and accounted for about two thirds of the cases each year. Serosubtype determination of the isolated strains failed to reveal the predominance of a single serogroup B phenotype. An increase in the overall serogroup B disease was observed during the year 2000 in the UK, both in vaccinated and in older unvaccinated age groups [11].

The emergence of serogroup A disease in Greece is a notable finding. Serogroup A disease is uncommon in Western European countries and the USA [4,7] and was very rare in Greece before 1997 [5,12]. Only 2/194 (1%) strains analyzed by the National Reference Laboratory between 1993 and 1997 belonged to this serogroup [12] compared to 7/105 (6.7%) strains in the 2-year period studied (P=0.01, Fisher's exact test). High carriage rates for serogroup A (up to 28% of groupable isolates) were observed a few years ago among ethnic Greek school children of recent Russian immigrants [27]. The emergence of this serogroup might be associated with the introduction of such strains by immigrant groups. Molecular analysis of these isolates will provide further evidence to support this assumption.

The shift in age distribution towards toddlers and preschoolers is remarkable and the proportion of infants affected was lower compared to recent reports in which disease incidence was highest in children less than 1 year of age [4,7]. Although IMD was overall more common (52%) in children under 5 years of age, only 7% of all patients were under 1 year of age. About 45% of the cases were among the 1–4-year age group which had the highest disease incidence.

Meningitis was diagnosed in the majority (101/136, 74%) of patients who had a lumbar puncture performed. Similar and even higher rates of CNS involvement up to 83% have been noted in previous studies of meningococcal disease in children [28,29]. The proportion of patients with rash, purpura, seizures, hypotension, respiratory failure and seizures was also similar to previous reports [28,30]. Meningococcal disease was milder in the second study year as reflected by the smaller number of patients who were admitted to ICU, developed hypotension or required assisted ventilation and by the smaller number of deaths. The increased use of PCR and the inclusion of milder cases confirmed only by this method may have contributed to the observed decrease in the proportions of patients with these complications, however, their absolute numbers were also much smaller in the second year. The decrease in disease severity could be due in part to the decline of the proportion of cases caused by serogroup C which was associated with high mortality in previous years [13]. This assumption cannot be supported by the data presented since the total number of serogroup C cases was small and the serogroup is not known for 3/7 fatal cases. The CFR of serogroup B disease has been previously reported between 5% and 6% and it was about 4% in this study [7,31].

The PCR test was particularly useful for confirmation of cases of IMD. It was the only positive laboratory test in 45% of the cases and the only test on which serogroup determination was based in 52%. In a recent report from the UK, PCR was the sole method of confirmation in about 50% of meningococcal meningitis cases while the number of cases with a serogroup identified increased by 44% [32]. The proportion of patients with positive cultures was low and this finding is in accordance with other studies in which PCR was used for identification of the microorganism [32]. This test is often positive in patients with clinically suspected meningococcal disease who have negative cultures [33,34]. A substantial number of cases were positive in the screening PCR using the IS1106 sequence but could not be serogrouped when the siaD PCR was applied. Other centers have had similar experience and it has been attributed to the higher sensitivity of the sequence used for screening for meningococci compared to the serogroup-specific one [33,34].

A considerable proportion of patients did not appear septic when first examined and the diagnosis of meningococcal disease in these cases was suspected because of the presence of the rash or meningeal signs. The simple laboratory tests such as WBC count and CRP were unremarkable in a considerable number of patients with IMD and were of limited value in patient evaluation and management. A high index of suspicion is, therefore, required for early diagnosis and treatment. The PCR test is useful in the diagnosis of atypical cases since treatment might be discontinued if a negative result is obtained.

In conclusion, the epidemiology of IMD during the 2 study years was characterized by the precipitous decrease in serogroup C, the predominance of serogroup B and the emergence of serogroup A disease in the Athens area. At the same time, a remarkable decrease in disease severity was also noted. The PCR test was particularly useful in confirmation of the diagnosis of IMD and disease epidemiology related to serogroup definition. These findings underscore the fact that the epidemiology and the clinical expression of these infections are rapidly evolving; therefore, continuous surveillance is required and the new molecular diagnostic methods are very useful tools. Decisions regarding the introduction of immunization with the new conjugate vaccines have to be made based on information such as the data presented in this report. Active surveillance will also be required to monitor the effect of vaccination on the epidemiology of this infection.


We would like to thank the following physicians and heads of the departments who shared the data of their patients with us: Dr. A. Chatzis, Dr. J. Papadatos, Dr. A. Yelesme, Dr. M. Narlioglou, Dr. G. Tsolas and Dr. E. Papadaki. This study was supported in part by grants from the Special Research Account of the University of Athens (KA70/4/5927) and the Greek Ministry of Health and Welfare.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
View Abstract