Low Co-infection Rate in Children With Community-Acquired Pneumonia in Spain

Abstract

Background. Areas of priority research include defining the epidemiology of community- acquired pneumonia (CAP) after the development of the molecular diagnostic tests. Previous studies have reported high rates of combined infection (8-51%). Our objective is to describe the etiology of CAP in hospitalized children, with a focus on the incidence of co-infections. Methods. From April 2012 until March 2015, hospitalized children with CAP were recruited in two hospitals of Madrid, Spain. An extensive microbiological work-up was performed, including: blood cultures, S. pneumoniae by PCR in blood, S. pneumoniae antigen in urine, paired serum for M. pneumoniae, C. pneumoniae and L. pneumophila, and PCR for 16 viruses, M. pneumoniae and C. pneumoniae in nasopharyngeal aspirate (NPA). Culture and S. pneumoniae antigen in pleural fluid was added if thoracentesis was performed. Organisms were considered the causative agents of pneumonia in the following situations: any bacteria in blood culture or S. pneumoniae by PCR in blood, urinary S. pneumoniae antigen plus C-reactive protein >100 mg/ L and/or procalcitonin >1.5 ng/mL, seroconversion to any agent, presence of nucleic acids of M. pneumoniae, C. pneumoniae, RSV, hMPV, ADV, PIV or flu virus on NPA, and any bacteria on culture or S. pneumoniae antigen detection in pleural fluid. Following recent recommendations, rhinovirus, enterovirus, bocavirus and coronavirus were excluded as causality, if detected. Results. We recruited 151 patients. They had a median age of 41 months (range 2- 201), and 53% were male. Two-thirds were under 60 months. A total of 93% were fully immunized against Hib, and 64% of them had received one or more dose of PCV13. One or more pathogens were documented in 66%: typical bacteria in 24 (16%, 23 [90%] of them S. pneumoniae), atypical bacteria in 31 (20%) (28 of them M. pneumoniae) and significant viruses in 60 (40%). Of note, half of the patients with M. pneumoniae were under 60 months. There were only 19 patients (13%) with co-infection with more than an agent. If we included non-significant viruses, 81% of patients had more than one organism identified and the co-detection rate raised up to 31%. Conclusion. Viruses are the main etiological agents of CAP in children. Considering significant etiological agents only, the co-infection rate was not as high as reported elsewhere. M. pneumoniae is not uncommon in children under 5 years.