Article Vol. 11. No. 3 217 ,--.-.,_. Eur. J. Clin. Microbiol. Infect. Dis., March 1992, p. 217-223 0934-9723/92/03 0217-07 $ 3.00/0 Pneumonia in Childhood: Etiology and Response to Antimicrobial Therapy O. Ruuskanenl,, H. Nohynek 4, T. Ziegler 2, R. Capeding 4, H. Rikalainen 3, E Huovinen 5, M. Leinonen 2 A prospective eight-month study was carried out in 50 children admitted to hospital for radiologically confirmed community-acquired pneumonia. A potential causative agent of infection was identified in 44 (88 %) cases. Using virus isolation, virus antigen detection and enzyme immunoassay serology, respiratory virus infection was diag- nosed in 30 (60 %) patients. Antibody assays for seven bacteria and antigen detection from serum and urine for Streptococcus pneumoniae produced evidence of bacterial infection in 31 (62 %) cases. Streptococcuspneumoniae (38 %), respiratory syncytial virus (30 %) and Mycoplasmapneumoniae (20 %) were the most common causative agents. A mixed infection was diagnosed in 25 (50 %) episodes. Nine patients failed to respond to antibiotics within 24 h after onset of ti'eatment. Three of them had a pure viral infection, three a mixed viral-bacterial infection, two a Mycoplasmapneumonlae infection mixed with other bacteria and one a pure Mycoplasma pneumoniae in- fection. All three Mycoplasraa pneumoniae infections were initially treated with penicillin. Determining the etiologic agent of childhood pneumonia is a difficult diagnostic problem, since appropriate specimens can rarely be obtained from the lower respiratory tract. Several recent Studies have investigated the etiology of child- hood pneumonia by demonstrating antigens of causative agents in nasopharyngeal secretions, urine and serum, and by studying antibody responses to bacteria and viruses. Evidence of respiratory virus infection has been found in 30 to 60 % and of bacterial infection in 20 to 45 % of patients with pneumonia. In all of these studies, hOWever, viral or bacterial diagnostic approaches have been restricted to a few agents and, more im- portant, no information has been reported about response to treatment (1-7). The aim of our study was to examine the clinical Value of the etiologic diagnosis of childhood pneumonia. A large number of diagnostic tests Were used. All patients received antibiotics. Treatment responses in relation to etiologic find- ings are reported. ..--..__ 1Department of Pediatrics, 2Deoartment of Virology, and aDepartment of Radiology, Turku University Hospital, 20520 Turku, Finland 4 . Nahonal Public Health Institute, 00300 Helsinki, Finland. s National Public Health Institute, 20520 Turku, Finland. Patients and Methods Patients. Between 1 February and, 30 September 1989, 50 consecutive patients with community-acquired pneu- monia were enrolled in the study. The diagnosis was based on simultaneous finding of fever (>__ 38 °C) and/or respiratory symptoms and infiltrate on the chest radio- graph. All patients were admitted to hospital and treated in the Department of Pediatrics, Turku University Hospi- tal. During the study period, pneumonia was diagnosed in 52 patients. Of these, one patient was excluded because a convalescent serum sample was not obtained. One patient who had suffered from recurrent pneumonia was also excluded. Five patients had an underlying illness; two had Down's syndrome, one mental retardation, one thalassemia minor and one congenital heart disease. The mean age of the patients was 4.4 years (range 0.5 to 14.3 years). Two children were less than 1 year of age, 27 between 1 and 3 years, 14 between 3 and 9 years, and seven between 10 and 15 years of age. An equal number of boys (n = 25) and girls (n = 25) were enrolled. Samples and Methods. Blood samples were taken for blood culture, antibody assays and bacterial antigen detection. The follow-up visit was made three to four weeks after discharge and included a clinical evaluation, a chest radiograph and a convalescent serum sample for antibody assays. Nasopharyngeal aspirate was obtained for viral studies. Influenza A and B, parainfluenza virus types l, 2 and 3, adenovirus and respiratory syncytiat virus (RSV) rapid antigen detection in nasopharyngeal aspirate was per- formed by time-resolved fluoroimmunoassays based on