Community-acquired pneumonia (CAP) in the pediatric age group is caused by a myriad of bacteria and viruses. The present study aims to determine the spectrum of bacterial pathogens causing CAP in children aged below 5 years, using conventional and molecular methods.

The study population comprised of 180 cases of CAP in children below 5 years of age, admitted in the Department of Pediatrics of a medical college hospital in Dibrugarh, Assam between June 2013 and May 2014. Cases were included as per WHO criteria for pneumonia, severe pneumonia or very severe pneumonia [1], and with the presence of lung infiltrates on chest X-ray. All cases of hospital-acquired pneumonia were excluded. The study was approved by the Institutional Ethics Committee of the institute. Informed consent was obtained from the parents or the legal guardians of the study participants.

The respiratory sample in majority of the cases was oropharyngeal swab. Bronchoalveolar lavage (BAL) was collected wherever possible. Blood samples were collected, and processed in Trypticase soy broth (HiMedia Labs, Mumbai, India) for culture. Oropharyngeal swab samples were collected with sterile flocked nylon swabs.

The first respiratory sample and the blood sample were processed according to standard microbiological procedures [2]. The second respiratory sample was used for PCR analysis. Total DNA was extracted by using the QIAamp DNA Mini Kit (Qaigen, Valencia, CA) according to manufacturer’s instructions. Conventional singleplex PCR reaction was carried according to previously described techniques with slight modifications [3-6]. The PCR end products were analyzed by agarose gel electrophoresis and visualized under UV transilluminator. Cultured stocks of ATCC strains were used as controls for the cultivable strains. For the difficult to culture or unculturable organisms, in-house control DNA previously verified by sequence analysis was used.

Data entry, database management and statistical analysis were done using Epi-Info software version 7. A P-value of <0.05 was considered to be satistically significant.

Out of the 180 cases studied, at least one bacterial pathogen was detected in 116 (64.4%) cases. The respiratory sample was oropharyngeal swab in 163 and BAL in 17 children. A single pathogen was detected in 105 (58.3%) and multiple pathogens were detected in 11 (6.1%) cases. A total of 131 bacterial pathogens were detected from the 116 positive cases.

Majority of the children in whom bacterial pathogen was detected were in the 0-12 months age group (n=75) and males (n=115). The mean (SD) age was 16.9 (15.1) months (Table I). A higher frequency of bacterial pathogens were detected in the cases presenting with very severe pneumonia in comparison to severe pneumonia and pneumonia.

TABLE I	Profile of Children With Community Acquired Pneumonia       

The most commonly detected bacterial pathogen was S. pneumoniae, followed by H. influenza (Table II). None of the H. influenzae strains were of type b. All the S. aureus strains were methicillin sensitive (MSSA). Non-pathogenic viridans streptococci were detected in 19 and mixed bacterial growth suggesting normal flora of the throat were detected in 33 cases (excluded from final analysis).

TABLE II	Profile of Samples Positive for Bacterial Pathogens

The overall mortality rate was 8.9%, and a bacterial pathogen was detected in 15 of these cases; the bacterial pathogens detected were S. pneumoniae in eight cases, and K. pneumoniae in four cases.

In case of S. pneumoniae, K. pneumoniae and S. aureus, all cases detected by PCR analysis of the respiratory samples were also detected by culture. The detection rate by blood culture was low (Table II).

In this study, S. pneumoniae and H. influenzae were the most commonly detected bacterial pathogens in the under-five children with CAP in this region. The atypical bacterial pathogens (M. pneumoniae and C. pneumoniae) accounted for about 10% of cases. The atypical pathogens could be detected only by PCR whereas typical bacterial pathogens could be detected as well by conventional culture methods.

The main limitation of this study was that isolates from the oropharyngeal swab samples might represent the normal oropharyngeal flora, and might not necessarily be the cause of CAP. Moreover, the study used conventional singleplex PCR for the detection of pathogens. Although the method is highly sensitive for detection, it cannot give an estimate of the bacterial load as compared to conventional culture. Also, there was no provision of testing the antibacterial sensitivity of the bacterial isolates following a PCR reaction. Convenience sampling and lack of asymptomatic controls were other limitations.

Various other studies have found S. pneumoniae and H. influenzae to be the most common isolate from respiratory samples of children with CAP [7-9]. S. aureus and some Gram-negative bacilli like K. pneumoniae were detected as the leading cause of pneumonia by Johnson, et al. [10]. The comparable performance of conventional methods and PCR in detection of typical organisms such as S. pneumoniae, K. pneumoniae and S. aureus has been reported earlier [11]. The detection of infection by blood culture was lower in the present study as compared to few other studies [9,10].

We conclude that S. pneumoniae and H influenzae are the most common bacterial organisms associated with community acquired pneumonia. PCR with pathogen-specific primers can improve the diagnostic yield by increasing the detection of fastidious organisms such as H. influenzae, and atypical agents like M. pneumoniae and B. pertussis.

Acknowledgements: Multidisciplinary Research Unit (ICMR), Assam Medical College & Hospital, Dibrugarh, Assam for providing logistic and infrastructural support.

Contributors: All authors have contributed, designed and approved the study.

Funding: Department of Biotechnology (DBT), Government of India. Competing interests: None stated.