Background 
Severe pneumonia remains the leading cause or mortality beyond the neonatal period in children < 5 years old worldwide, particularly severe pneumococcus pneumonia. Pneumococcus conjugate vaccine has been identified as one of the effective strategies to reduce mortality in this population. Thorough understanding of the microbiologic etiology of severe pneumonia in children will allow for design of new and improved vaccines
Methods 
We conducted a prospective surveillance (2013 – 2016) of all cases of community acquired pneumonia complicated by empyema, as a surrogate of severe pneumonia, in children (0 – 18 years old) treated at a large children’s hospital in Orange County, California, and evaluated the microbiologic diagnosis sensitivity by routine culture as well as by molecular methods. When available, pneumococcal serotype was determined
Results 
During the study period 60 cases of empyema were identified, the microbiologic diagnosis was improved from 19 (31.7%) to 39 (65%) cases. Pneumococcus was identified in 27 (69.2%) cases, of which 17 (63%) were identified by molecular techniques. During the study time there were 38 additional cases of invasive pneumococcal disease.  Among the 10 pneumococci identified by culture in subjects with empyema, 7 (70%) were current vaccine serotypes (0 in 2016) all the while only 6 (15.8%) of non-respiratory invasive pneumococcal isolates were vaccine serotype
Conclusions 
Culture is not an appropriate modality to identify pneumococcus in children with severe pneumonia. Pneumococcal serotypes identified in non-respiratory disease do not accurately reflect those responsible for severe pneumonia.  Molecular techniques are going to be necessary to identify the proportion of pneumococci and there serotype in children with severe pneumonia to improve the composition and evaluate the efficacy of new vaccines

COMPARISON OF CULTURE AND MOLECULAR TECHNIQUES TO IDENTIFY THE MICROBIOLOGICAL ETIOLOGY OF SEVERE PNEUMONIA IN CHILDREN: IMPACT ON NEW VACCINE DEVELOPMENT