• aetiology;
  • community-acquired pneumonia;
  • New Caledonia;
  • PCR;
  • influenza virus


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Objective  To describe the aetiology of community-acquired pneumonia (CAP) in hospitalized adult patients in New Caledonia, a French archipelago in the South Pacific.

Methods  Confirmed CAP patients (n = 137) were enrolled prospectively. Pathogens were detected by culture, molecular methods, serology on paired sera, immunofluorescence on nasopharyngeal swabs and antigen detection in urine.

Results  The aetiology of CAP was determined in 82 of 137 cases (59.8%), of which 31 exhibited two or more pathogens (37.8%). Hundred and seventeen pathogens were detected: Streptococcus pneumoniae was the most common one (41.0%), followed by influenza virus A (22.1%) and Haemophilus influenzae (10.2%). The frequency of atypical bacteria was low (6.0%). The most frequent and significant coinfection was S. pneumoniae with influenza A virus (P = 0.004). Influenza virus was detected from nasopharyngeal swabs in four patients (15.4% of patients tested for influenza) and by PCR from pulmonary specimens in 15 patients (57.7%).

Conclusions Streptococcus pneumoniae is the leading cause of CAP in New Caledonian adults. Viral–bacterial co-infections involving S. pneumoniae and influenza virus are very common during the winter. Such adult patients hospitalized with CAP are a clear sentinel group for surveillance of influenza. Vaccination against influenza and S. pneumoniae should be strengthened when risk factors are identified.

Etiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie 

Objectif:  Décrire l’étiologie de la pneumonie communautaire (PC) chez les patients adultes hospitalisés en Nouvelle-Calédonie, un archipel français dans le sud du Pacifique.

Méthodes:  Les patients confirmés PC (n = 137) ont été recrutés de façon prospective. Les agents pathogènes ont été détectés par culture, méthodes moléculaires, sérologie sur des paires de sérums, immunofluorescence sur des prélèvements du nasopharynx et détection d’antigène dans les urines.

Résultats:  L’étiologie de la PC a été déterminée chez 82 des 137 cas (59,8%) dont 31 présentaient deux ou plusieurs agents pathogènes (37,8%). 117 agents pathogènes ont été détectés: Streptococcus pneumoniaeétait le plus courant (41,0%), suivi par le virus influenza A (22,1%) et Haemophilus influenzae (10,2%). La fréquence des bactéries atypiques était faible (6,0%). La coinfection la plus fréquente et importante était celle avec S. pneumoniae et le virus de l’influenza A (p = 0,004). Le virus de l’influenza a été détecté dans des prélèvements nasopharyngés chez 4 patients (15,4% des patients testés pour l’influenza) et par PCR à partir de prélèvements pulmonaires chez 15 patients (57,7%).

Conclusions: S. pneumoniae est la principale cause de PC chez les adultes de Nouvelle-Calédonie. Les coinfections viro-bactériennes impliquant S. pneumoniae et le virus de l’influenza sont très fréquentes pendant l’hiver. De tels patients adultes hospitalisés avec une PC sont un groupe sentinelle claire pour la surveillance de l’influenza. La vaccination contre l’influenza et S. pneumoniae devrait être renforcée lorsque les facteurs de risque sont identifiés.

Mots-clés:  étiologie, pneumonie communautaire, Nouvelle-Calédonie, virus influenza, PCR.

Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia

Objetivo:  Describir la etiología de la neumonía adquirida en la comunidad (NAC) en pacientes adultos hospitalizados en Nueva Caledonia, el archipiélago francés del Pacífico Sur.

Métodos:  Se incluyeron, de forma prospectiva, pacientes con confirmación de NAC (n = 137.) Los patógenos se detectaron mediante cultivo, métodos moleculares, serología en sueros pareados, inmunofluorescencia en frotis nasofaríngeo y detección de antígenos en orina.

Resultados:  Se determinó la etiología de NAC en 82 de 137 casos (59.8%), de los cuales 31 exhibían dos o más patógenos (37.8%). Se detectaron 117 patógenos: Streptococcus pneumoniae era el más común (41.0%), seguido por el virus de influenza A (22.1%) y Haemophilus influenzae (10.2%). La frecuencia de bacterias atípicas era baja (6.0%). La coinfección más frecuente y significativa era S. pneumoniae con el virus de influenza tipo A (p = 0.004). Se detectó el virus de influenza en frotis nasofaríngeo de 4 pacientes (15.4% de los pacientes examinados para influenza) y por PCR de muestras pulmonares en 15 pacientes (57.7%).

Conclusiones: S. pneumoniae es la principal causa de NAC en adultos de Nueva Caledonia. Las coinfecciones que involucran a S. pneumoniae y al virus de la influenza son muy comunes durante el invierno. Estos pacientes adultos hospitalizados con NAC son un grupo centinela obvio para la vigilancia epidemiológica de la influenza. Se debería reforzar la vacunación frente a influenza y S. pneumoniae cuando se han identificado factores de riesgo.

Palabras clave:  etiología; neumonía adquirida en la comunidad; Nueva Caledonia; PCR; virus de la influenza


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Community-acquired pneumonia (CAP) is an important cause of hospitalization and death worldwide (Garau & Calbo 2008). Antimicrobial treatment must be prompt but is often given in the absence of, or before, microbiological investigations, based on specific guidelines adapted to local microbial patterns (Woodhead et al. 2005; Macfarlane & Boldy 2004; Mandell et al. 2007).

New Caledonia is a French archipelago in the South Pacific. Its population is approximately 230 800 (2004 census) divided into different ethnic groups: Melanesians (44%), Europeans (34%), Wallisians (9%), Tahitians (2.6%), Indonesians (2.5%) and others (7.9%). The climate is tropical and oceanic with two marked seasons, a warm and wet season from December to March and a cooler one from June to September. Patients with respiratory infections are usually referred to the Territory Hospital in Noumea, which has intensive care unit (ICU) facilities and specialist chest physicians. Everyone has access to the health care system.

There has been no comprehensive study of the aetiology of CAP in New Caledonia; disease management is based on French guidelines (AFSSAPS 2005). However, locally developed treatment guidelines have been effective in some countries of the Pacific region (Elliott et al. 2005). Moreover, specific aetiologies have been reported in New Caledonia (Burkholderia pseudomallei, Histoplasma capsulatum), which could be taken into account to establish treatment guidelines (Estivals et al. 2008; Noel et al. 1995).

We conducted a prospective study over a 1-year period in adult patients hospitalized in New Caledonia’s main hospital to (i) determine the aetiology of CAP, (ii) evaluate the performance of locally available laboratory tests and (iii) assess the vaccination preventable proportion of causative agents, with a special focus on Streptococcus pneumoniae and influenza.


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Study design

Adult patients, aged 16 or older, admitted to the Territory Hospital with suspected CAP (n = 173) were evaluated prospectively between December 2006 and November 2007. We used the following criteria for confirmation of CAP: presence of lower respiratory tract infection symptoms at initial presentation or within 48 h of being hospitalized associated with infiltrates on chest X-ray. Patients with HIV-related immunosuppression, tuberculosis or any other illness causing radiographic abnormalities (lung cancer, for example) were excluded. Eligibility was confirmed for each patient by a senior chest physician. Patients lacking respiratory specimens for microbiological investigations were also excluded.

Ethics, patient evaluation, data and sample collection

The New Caledonian Health Authorities approved the study protocol, and informed consent was obtained from all patients. Patient data were collected using a standardized questionnaire, including pneumococcal and influenza immunization history. Chest X-ray patterns and laboratory results at presentation and during the hospital stay were collected. The two criteria used to assess CAP severity were admission to the ICU and the pneumonia severity index (PSI), defining severe disease as risk classes IV and V (Fine et al. 1997).

Respiratory tract specimens [sputum, tracheobronchial aspirate (TBA), bronchoalveolar lavage (BAL) or protected specimen brush (PSB)], two blood cultures and, when indicated, pleural fluid were collected shortly after admission, before starting antibiotherapy. We additionally took paired sera for serology, urine for antigen testing and nasopharyngeal swabs for virus detection.

Microbiological laboratory tests

Microbiological tests were carried out on respiratory specimens according to the French Society of Microbiology (SFM) criteria (SFM 2007). Sputum samples were Gram stained and leucocytes and epithelial cells counts recorded per ×100 microscopic field. Only good-quality specimens (>25 granulocytes and <10 epithelial cells) were retained for analysis (Murray & Washington 1975). Sputum, TBA, BAL, PSB and pleural fluid samples were cultured at 37 °C on blood agar with nalidixic acid, chocolate agar and Sabouraud agar (when positive for yeast on Gram stain). Cultures were counted, with results expressed as colony forming units per millilitre (CFU/ml) and interpreted using the SFM standards.

Blood cultures were analysed with the BacT/ALERT® 3D system (bioMérieux, France). Influenza viruses A and B, parainfluenza virus-3 (PIV-3) and adenovirus (ADV) were detected from nasopharyngeal swabs by indirect immunofluorescence assay (Monofluo® Kit, Biorad, France) and respiratory syncytial virus (RSV) by direct immunofluorescence assay (Monofluo® Screen, Biorad, France). Legionella pneumophila serotype 1 and S. pneumoniae antigens were detected in non-concentrated urine using an immunochromatographic assay (Binax NOW®, Binax Inc., USA).

PCR assay

RNA and DNA from sputum, TBA, BAL or PSB samples were extracted using a single format magnetic silica-based method automated on the EasyMag system (bioMérieux, France).

A molecular assay was used for parallel detection of Chlamydophila pneumoniae, Mycoplasma pneumoniae and Legionella spp. Real-time PCR was performed using the LightCycler® 2.0 (Roche Diagnostics, New Zealand) with primers and hybridization probes, as described elsewhere (Raggam et al. 2005). Influenza viruses A and B were detected from respiratory samples by real-time RT-PCR. We used primers and hybridization probes previously described for influenza B virus (Smith et al. 2003). Influenza A virus–specific hybridization probes were used, allowing H1N1 and H3N2 subtypes discrimination by melting temperature analysis (Stone et al. 2004).


Paired sera samples were collected when possible at an optimal interval of 3 weeks. Antibodies to influenza A and B viruses, ADV, RSV, PIV types 1, 2 and 3 (Virion Institute, Switzerland) were detected by complement fixation tests. Mycoplasma pneumoniae IgM antibodies were tested by indirect ELISA (RIDASCREEN®, r-biopharm, Germany).

Diagnostic criteria for determination of microbial aetiology

Aetiology was classified as ‘definite’, ‘probable’ or ‘unknown’ following SFM guidelines and published criteria (SFM 2007; Ruiz et al. 1999). It was considered definite when (i) a usual respiratory tract pathogen was cultured from blood or pleural fluid, (ii) bacterial growth in quantitative cultures reached >105 CFU/ml for TBA, >104 CFU/ml for BAL and >103 CFU/ml for PSB, (iii) the PCR for C. pneumoniae, M. pneumoniae, Legionella spp. or influenza viruses was positive, (iv) the urinary antigen test was positive for S. pneumoniae or L. pneumophila, (v) the immunofluorescence assay for respiratory viruses was positive, or (vi) when we observed a seroconversion, or fourfold rise in antibody titres for respiratory viruses, a single titre for influenza viruses >80 or, for M. pneumoniae, a single high IgM titre >70 UA/ml. Influenza-positive serology in patients recently vaccinated – i.e. 12 months or less prior to hospitalization – was not considered, as post-infection and vaccine-induced antibodies are indistinguishable.

The aetiology of pneumonia was classified as probable when a valid sputum sample yielded one or two predominant bacterial strains.


Stata SE 8.0 (StataCorp, Texas, USA) was used for all calculation. Continuous variables were compared using the Student’s t-test when the theoretical numbers were low. Categorical variables were compared by using the chi-squared test or Fisher’s exact test. Poisson regression was used to determine associations among continuous variables. Multivariate analysis was performed using stepwise forward logistic regression. All reported P values are two tailed. The level of significance was set at 5%. We calculated odds ratios (OR) and 95% confidence intervals (95% CI).


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References


We enrolled 173 patients in the study, 36 of whom were excluded: 23 had a normal chest X-ray, eight showed abnormalities without pneumonia, two were suspected of nosocomial pneumonia, and three had no respiratory specimen available.

The 137 included patients had a median age of 58 (16 to 96), a male/female ratio of 1.4 and an ethnic distribution as follows: 77 Melanesian (56.2%), 23 European (16.8%), 23 Wallisian (16.8%), four Tahitian (2.9%) and 10 from other minor groups (7.3%).

There was no significant association between ethnic group and any comorbid condition (Table 1), although some factors were observed more frequently in Wallisian patients: current cigarette smoker (60.9%), pulmonary disease (34.8%) and diabetes mellitus (26.1%). Immunization rates against seasonal influenza and S. pneumoniae are summarized in Figure 1. Twenty-four patients (17.5%) were treated with antibiotics (at least one beta-lactam) prior to hospitalization.

Table 1.   Main characteristics in 137 patients hospitalized with CAP in New Caledonia
Ethnic group (n)Sex ratio (M/F)ICU admission n (%)Mortality rate % (n)Comorbid conditions
Cardiovascular disease n (%)Pulmonary disease n (%)Renal disease n (%)Diabetes mellitus n (%)Malignancy n (%)Alcoholism n (%)Current cigarette smoker n (%)ID n (%)Other n (%)
  1. ICU, intensive care unit; ID, immunodepression; CAP, community-acquired pneumonia.

Melanesian (77)39/3810 (13.0)6.5 (5)19 (24.7)22 (28.6)4 (5.2)7 (9.1)6 (7.8)8 (10.4)31 (40.2)01 (1.3)
European (23)13/106 (26.1)8.7 (2)7 (30.4)6 (26.1)03 (13.0)2 (8.7)3 (13.0)13 (56.5)1 (4.3)0
Wallisian (23)15/82 (8.7)4.3 (1)3 (13.0)8 (34.8)06 (26.1)0014 (60.9)00
Tahitian (4) 3/1001 (25.0)1 (25.0)0001 (25.0)1 (25.0)00
Other ethnic group (10)10/04 (40.0)03 (30.0)4 (40.0)03 (30.0)2 (20.0)2 (20.0)3 (30.0)2 (20.0)0
Total (137)80/5722 (16.0)5.8 (8)33 (24.1)41 (29.9)4 (2.9)19 (13.9)10 (7.3)14 (10.2)62 (45.2)3 (2.2)1 (0.7)

Figure 1.  Immunization rates against seasonal influenza and S. pneumoniae (23-valent pneumococcal polysaccharide vaccine).

Download figure to PowerPoint

CAP patients main characteristics

Twenty-two patients (16.0%), 17 men and five women, were admitted to the ICU (Table 1) and had a median age of 62 years (37–81). The rate of ICU admission did not differ significantly between ethnic or age groups. Fifty-one patients (37.2%) had a PSI < III, 36 of whom (70.5%) were hospitalized for reasons other than pneumonia severity. Length of hospital stay ranged from 1 to 65 days, with a median of 8 days. Mortality rate was 5.8% (8/137).

Microbiological investigations

Only 59 of the 127 sputum samples collected (46.4%) were considered to be of good quality. Blood culture was positive in 20 patients. In four cases, it was the only positive test; the others were also positive for at least one other technique (S. pneumoniae urinary antigen test, n = 12; positive culture from a respiratory specimen, n = 6). Pleural fluid analysis detected one isolate of Staphylococcus aureus. Urinary antigen detection was positive for S. pneumoniae in 40 of 133 patients (30%), and the diagnosis rate of pneumococcal CAP was increased by a factor of 2.1 using this test systematically; no urinary samples were tested positive for L. pneumophila serotype 1. Chlamydophila pneumoniae PCR assays were performed for 132 patients, yielding one positive result. Mycoplasma pneumoniae IgM was detected for 6 of 136 patients tested.

Immunofluorescence assays were positive for five patients (influenza virus A, n = 4; PIV-3, n = 1). Paired sera were obtained for only 34 of 137 patients. In this small subset of patients, 22 were positive for ADV (n = 1), PIV-3 (n = 1), influenza virus A (n = 16), influenza virus A + RSV (n = 1) and possible persistent antibodies against influenza vaccine antigens (n = 3). Diagnosis rates of the different techniques for influenza virus detection are shown in Figure 2.


Figure 2.  Diagnosis rate of the different techniques used to detect influenza virus. IF, immunofluorescence; PCR, polymerase chain reaction.

Download figure to PowerPoint

According to our case classification, aetiology could be established for 82 of 137 patients (59.8%), considered definite in 47 (34.3%), probable in 35 (25.5%) and unknown in 55 (40.2%) patients. Of the 117 pathogens identified (Table 2), 11 viruses and 40 bacteria were identified as single cause of the CAP. Viral–bacterial co-infections accounted for 13.1% of CAP (n = 18), while mixed bacterial and mixed viral infections accounted for 9.5% (n = 13) and 0% of CAP, respectively (Tables 3 and 4).

Table 2.   Aetiology of CAP in hospitalized adult patients in New Caledonia
PathogensSingle aetiology CAPViral–bacterial co-infectionsMixed bacterial infectionsMixed viral infectionsCAP aetiologies definite/probable/unknown
n (%)n (%)n (%)n (%)n/n/n
  1. RSV, respiratory syncytial virus; CAP, community-acquired pneumonia; NA, not applicable.

Streptococcus pneumoniae (Sp)25 (49.0)12 (30.7)11 (40.7)NA42/6/NA
Classical bacteria (non Sp)9 (17.6)4 (10.0)9 (33.3)NA3/19/NA
 Haemophilus influenzae5 (9.8)1 (2.5)6 (22.2)NA0/12/NA
 Moraxella catarrhalis2 (3.9)1 (2.5)3 (11.1)NA1/5/NA
 Staphylococcus aureus2 (3.9)2 (5.0)NA2/2/NA
Atypical bacteria1 (1.9)2 (5.0)4 (14.8)NA7/0/NA
 Mycoplasma. pneumoniae1 (1.9)2 (5.0)3 (11.1)NA6/0/NA
 Chlamydophila pneumoniae1 (3.7)NA1/0/NA
 Legionella pneumophilaNA
Gram-negative bacilli + Pseudomonas aeruginosa5 (9.8)2 (5.0)3 (11.1)NA5/5/NA
 Escherichia coli1 (1.9)1 (3.7)NA1/1/NA
 Klebsiella pneumoniae3 (6.0)1 (2.5)1 (3.7)NA3/2/NA
 Enterobacter cloacae1 (2.5)NA0/1/NA
 P. aeruginosa1 (1.9)1 (3.7)NA1/1/NA
Viruses11 (21.5)19 (48.6)NA030/0/NA
 Influenza A9 (17.6)17 (43.6)NA26/0/NA
 Influenza BNA
 Para-influenza 32 (3.9)NA2/0/NA
 Adenovirus1 (2.5)NA1/0/NA
 RSV1 (2.5)NA1/0/NA
Number of pathogens51 (100)39 (100)27 (100)087/30/NA
Number of patients511813047/35/55
Table 3.   Mixed bacterial infections responsible for CAP in New Caledonia
Mixed bacterial infections responsible for CAPn
  1. CAP, community-acquired pneumonia.

Streptococcus pneumoniae plus
 Haemophilus influenzae5
 Haemophilus influenzae and Chlamydophila pneumoniae1
 Mycoplasma pneumoniae2
 Moraxella catarrhalis2
 Pseudomonas aeruginosa1
Escherichia coli and Klebsiella pneumoniae1
Moraxella catarrhalis and Mycoplasma pneumoniae1
Table 4.   Viral–bacterial co-infections responsible for CAP in New Caledonia
Viral–bacterial co-infections responsible for CAPn
  1. CAP, community-acquired pneumonia.

Streptococcus pneumoniae plus
 Influenza virus A10
 Influenza virus A and Mycoplasma pneumoniae1
Influenza virus A plus
 Staphylococcus aureus and Klebsiella pneumoniae1
 Staphylococcus aureus and respiratory syncytial virus1
 Haemophilus influenzae1
 Mycoplasma pneumoniae1
 Moraxella catarrhalis1
 Enterobacter cloacae1

Streptococcus pneumoniae was the most common pathogen; it was present in about one-third of patients studied (48/137), representing 41.0% of identified pathogens (48/117) and 74.2% of co-infections and mixed infections (23/31). Serotypes of S. pneumoniae responsible for adults CAP in New Caledonia and concordance with pneumococcal vaccines are shown in Figure 3. Ranking second was influenza virus A (22.1%), followed by Haemophilus influenzae (10.2%), Moraxella catarrhalis (5.1%), M. pneumoniae (5.1%), Klebsiella pneumoniae (4.2%) and S. aureus (3.5%).


Figure 3.  Serotypes of S. pneumoniae responsible for adults community-acquired pneumonia in New Caledonia and concordance with anti-pneumococcal vaccines. 7-valent pneumococcal conjugate vaccine (PCV7): 4, 6B, 9V, 14, 18C, 19F, 23F; 13-valent pneumococcal conjugate vaccine (PCV13): 1, 3, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F, 23F; 23-valent pneumococcal polysaccharide vaccine (PPV23): 1, 2, 3, 4, 5, 6B, 7F, 8, 9N, 9V, 10A, 11A, 12F, 14, 15B, 17F, 18C, 19A, 19F, 20, 22F, 23F, 33F.

Download figure to PowerPoint

The most frequent (n = 11) and significant coinfection was S. pneumoniae associated with influenza A virus (= 0.004).

Of the eight patients who died, the causative agent was identified in five: Escherichia coli Klebsiella pneumoniae, n = 1; K. pneumoniae, n = 2; S. pneumoniae, n = 1; and S. pneumoniae + influenza virus A, n = 1. We noticed a particularly high incidence of Gram-negative enteric bacilli bacteremia (K. pneumoniae = 3) in this group, but there was no significant association with CAP severity (= 0.067) or suspicion of aspiration pneumonia.

Epidemiological patterns

Community-acquired pneumonia caused by S. pneumoniae (n = 48) occurred more frequently in patients aged <45 (= 0.04), with no significant difference between ethnic groups. There was no significant association between ethnicity or comorbidity and aetiology of CAP. Streptococcus pneumoniae was the most frequent pathogen isolated in patients admitted to ICU (9/22). However, its frequency in this subgroup was similar to that observed for the whole population studied (40.9% and 41.0%, respectively). Klebsiella pneumoniae was the most frequent agent detected in fatal CAP, but the association between this bacterium and ICU admission was not significant (= 0.067). There was no relationship between viral–bacterial co-infections and ICU admission neither with a PSI > III. Incidence of CAP was significantly higher during southern winter (= 0.007). This increased incidence was associated with epidemic seasonal influenza observed during the same period in the general population (= 0.002; Figure 4). Multivariate analysis only showed a significant association between influenza-associated CAP and southern winter (= 0.03).


Figure 4.  Seasonal distribution of community-acquired pneumonia in New Caledonia.

Download figure to PowerPoint


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

This is the first study of CAP aetiology in New Caledonia. Aetiology was attributed in 59.8% of cases, similar to published data (Jennings et al. 2008). Streptococcus pneumoniae urinary antigen testing provided a good alternative to tests involving more invasive means of sample collection. Indeed, the increase in diagnosis rate was consistent with a previous study using further tests such as lung aspirate analysis after transthoracic needle aspiration (Ruiz-Gonzalez et al. 1999).

Viruses were identified in 29 of 137 patients with CAP (21.1%), of which 11 were monomicrobic (8.0%) and 18 associated with bacteria. Influenza viruses accounted for 86.6% of viruses in our patients, whereas they accounted for 64.0% of viruses detected in a similar group of patients in Spain in 2004 (de Roux et al. 2004). In New Caledonia, the influenza sentinel network was set up in 1999, including paediatric hospital wards, public and private community GPs. The case definition was based on conventional influenza-like illness (ILI) (Berlioz-Arthaud & Barr 2005). Our group of patients is clearly a relevant sentinel group for influenza surveillance: the seasonality of CAP in adults follows this of influenza as evidenced by the surveillance network, suggesting the role of influenza as leading or associated cause of hospitalized CAPs (Figure 4). Moreover, in adult patients, influenza surveillance may be more specific than in children where other viruses can induce differential diagnosis. Our results show the benefit of using molecular techniques on pulmonary specimens, e.g. sputum, for the rapid diagnosis of influenza infection in adults; PCR yielded 15 positive results and was the only test to detect infection in approximately one quarter of positive patients (7/26). Immunofluorescence analysis of nasopharyngeal swab smears was positive in only four of 134 patients. By contrast, this test shows high sensitivity in children. Hospitalization of adult patients probably occurred late after the onset of symptoms, when influenza viral load in the upper respiratory tract was low or undetectable. Influenza viral infections were mixed in 65.4% of our patients; associated bacterial infection may be attributed to superinfection.

Peltola et al. showed that the neuraminidase activity of influenza viruses correlated with their capacity to promote secondary bacterial pneumonia. Based on these observations, a high level of neuraminidase activity may explain the particularly strong and significant association between influenza A/H3N2 virus and S. pneumoniae in our study, leading to death in two cases (Peltola et al. 2005; Wagner et al. 2000). The weak immunization rate in our patients (7.3%) may lead to strengthen annual influenza immunization campaigns among New Caledonian risk groups, i.e. elderly and chronically ill patients.

The rate of atypical pathogens was very low (6.0%), possibly explained by our inclusion criteria (hospitalization), the severity of CAP (for 58.4% of the patients, PSI was ≥III) and the older age of the study population. According to European data, the frequency of M. pneumoniae usually rises as illness severity decreases (Woodhead 2002). Mycoplasma pneumoniae IgM were detected in six patients not exhibiting a positive PCR. Different situations can explain this discordance: false positive IgM results (detection of non-specific antibodies), residual IgM in patients with a past M. pneumoniae infection or failure to detect M. pneumoniae DNA (PCR inhibition, poor quality of the respiratory sample). Sputum has been previously described as specimen of choice for detection of M. pneumoniae, giving 70% concordance with IgM serology, whereas only 40–50% concordance is observed for other specimens (Dorigo-Zetsma et al. 2001). We used PCR assays to test for C. pneumoniae. Most other studies use serology tests, therefore making it difficult to compare our results with others. Legionella infection usually represents 2–5% of CAP, correlating with severity (Woodhead 2002). However, despite using two sensitive and specific methods, we did not identify Legionella infection in our study population. Studies in Australia and New Zealand have shown L. longbeachae to account for 30.4% of cases of culture-confirmed community-acquired legionellosis (Yu et al. 2002). Urinary antigen detection tests should therefore be used and interpreted with caution in the South Pacific, as only detecting L. pneumophila serotype 1.

In New Caledonia, despite differences in traditional medicine, way of life and socio-economical criteria, we did not find any differences in CAP aetiology or severity between ethnic groups. However, S. pneumoniae accounted for 41% of CAP in Melanesians under 45 but only 26% of Melanesian patients with CAP over 45. This difference is not significant, but it may be of interest to investigate age stratification in a larger study group. Analysis of S. pneumoniae capsular serotypes showed that 79% of strains responsible for CAP could be targeted with the 23-valent pneumococcal polysaccharide vaccine, highlighting the importance of pneumococcal vaccination in adults. The current 7-valent pneumococcal conjugate vaccine used in New Caledonia since 2004 is obviously poorly effective on the strains involved in adults CAPs (serotype concordance of 10% only between CAP in New Caledonian adults and vaccine); however, we hope a better indirect (herd) protection in the future with the 13-valent conjugate vaccine since concordance will reach 73% with this new vaccine.


  1. Top of page
  2. SummaryEtiologie de la pneumonie communautaire chez les patients adultes hospitalisés en Nouvelle-Calédonie Etiología de la neumonía adquirida en la comunidad, entre pacientes adultos hospitalizados en Nueva Caledonia
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  • Agence Française de Sécurité Sanitaire des Produits de Santé– AFSSAPS (2005) Antibiothérapie par voie générale en pratique courante dans les infections respiratoires basses. Principaux messages des recommandations de bonne pratique. (accessed 15 March 2010).
  • Berlioz-Arthaud A & Barr IG (2005) Laboratory-based influenza surveillance in New Caledonia, 1999–2003. Transactions of the Royal Society of Tropical Medicine and Hygiene 99, 290300.
  • Dorigo-Zetsma JW, Verkooyen RP, Van Helden HP, Van Der Nat H & Van Den Bosch JM (2001) Molecular detection of Mycoplasma pneumoniae in adults with community-acquired pneumonia requiring hospitalization. Journal of Clinical Microbiology 39, 11841186.
  • Elliott JH, Anstey NM, Jacups SP, Fisher DA & Currie BJ (2005) Community-acquired pneumonia in northern Australia: low mortality in a tropical region using locally-developed treatment guidelines. International Journal of Infectious Diseases 9, 1520.
  • Estivals M, Du Couedic L, Lacassin F, Mermond S & Levenes H (2008) [Septicemia, bilateral community acquired pneumonia and empyema due to Burkholderia pseudomallei (mellioidosis) with a favorable outcome following prolonged specific antibiotic therapy]. Revue des Maladies Respiratoires 25, 14.
  • Fine MJ, Auble TE, Yealy DM et al. (1997) A prediction rule to identify low-risk patients with community-acquired pneumonia. The New England Journal of Medicine 336, 243250.
  • Garau J & Calbo E (2008) Community-acquired pneumonia. Lancet 371, 455458.
  • Jennings LC, Anderson TP, Beynon KA et al. (2008) Incidence and characteristics of viral community-acquired pneumonia in adults. Thorax 63, 4248.
  • Macfarlane JT & Boldy D (2004) 2004 update of BTS pneumonia guidelines: what’s new? Thorax 59, 364366.
  • Mandell LA, Wunderink RG, Anzueto A et al. (2007) Infectious diseases society of America/American thoracic society consensus guidelines on the management of community-acquired pneumonia in adults. Clinical Infectious Diseases 44, S27S72.
  • Murray PR & Washington JA (1975) Microscopic and bacteriologic analysis of expectorated sputum. Mayo Clinic Proceedings 50, 339344.
  • Noel M, Levenes H, Duval P, Barbe C, Ramognino P & Verhaegen F (1995) [Report of acute pulmonary histoplasmosis after visiting a cave in New Caledonia]. Santé 5, 219225.
  • Peltola VT, Murti KG & McCullers JA (2005) Influenza virus neuraminidase contributes to secondary bacterial pneumonia. The Journal of Infectious Diseases 192, 249257.
  • Raggam RB, Leitner E, Berg J, Muhlbauer G, Marth E & Kessler HH (2005) Single-run, parallel detection of DNA from three pneumonia-producing bacteria by real-time polymerase chain reaction. The Journal of Molecular Diagnostics 7, 133138.
  • De Roux A, Marcos MA, Garcia E et al. (2004) Viral community-acquired pneumonia in nonimmunocompromised adults. Chest 125, 13431351.
  • Ruiz M, Ewig S, Marcos MA et al. (1999) Etiology of community-acquired pneumonia: impact of age, comorbidity, and severity. American Journal of Respiratory and Critical Care Medicine 160, 397405.
  • Ruiz-Gonzalez A, Falguera M, Nogues A & Rubio-Caballero M (1999) Is Streptococcus pneumoniae the leading cause of pneumonia of unknown etiology? A microbiologic study of lung aspirates in consecutive patients with community-acquired pneumonia. The American Journal of Medicine 106, 385390.
  • Smith AB, Mock V, Melear R, Colarusso P & Willis DE (2003) Rapid detection of influenza a and b viruses in clinical specimens by Light Cycler real time RT-PCR. Journal of Clinical Virology 28, 5158.
  • Société Française de Microbiologie – SFM (2007) Examen cytobactériologique des sécretions bronchopulmonaires. In Rémic: Vivactis Plus Ed, 3134.
  • Stone B, Burrows J, Schepetiuk S et al. (2004) Rapid detection and simultaneous subtype differentiation of influenza a viruses by real time PCR. Journal of Virological Methods 117, 103112.
  • Wagner R, Wolff T, Herwig A, Pleschka S & Klenk HD (2000) Interdependence of hemagglutinin glycosylation and neuraminidase as regulators of influenza virus growth: a study by reverse genetics. Journal of Virology 74, 63166323.
  • Woodhead M (2002) Community-acquired pneumonia in Europe: causative pathogens and resistance patterns. The European Respiratory Journal 36, S20S27.
  • Woodhead M, Blasi F, Ewig S et al. (2005) Guidelines for the management of adult lower respiratory tract infections. The European Respiratory Journal 26, 11381180.
  • Yu VL, Plouffe JF, Pastoris MC et al. (2002) Distribution of legionella species and serogroups isolated by culture in patients with sporadic community-acquired legionellosis: an international collaborative survey. The Journal of Infectious Diseases 186, 127128.