Pleural fluid in hospitalized pediatric patients

We reviewed 105 charts of children who were hospitalized in a major teaching hospital between 1987 and 1993 with a diagnosis of pleural effusion established by radiographic study of the chest. There were 75 males and 30 females; mean age was 7.2 years (range 1 day-18 years). Pleural fluid was secondary to trauma, renal disease, or malignancy in 31.5% of patients. Parapneumonic effusions were found in the majority of patients, 64/105 (61%). However, in only 38 of the 64 patients (59%) was an organism isolated from any source. The most common bacterial organism cultured was Haemophilus influenzae (11 patients), followed by Staphylococcus aureus (seven patients). In five of 38 patients, a viral etiology was diagnosed. Pleural fluid was examined in only 35 patients (34%) and a pathogen found in only five (three bacterial, one Candida, one respiratory syncytial virus). It was possible to classify only 17 cases as exudate (fluid/blood protein ratio > or = 0.5) and six cases as transudate (fluid/blood protein < 0.5). Of 64 patients with suspected pleural fluid secondary to infection, 25 (39%) underwent thoracentesis and only 17 (27%) required tube drainage. This study demonstrates a wide spectrum of etiologies for pleural fluid in children, as well as the reduced use of thoracentesis or chest tube drainage in suspected infection. We speculate that this is probably because of the extensive empiric use of broad-spectrum antibiotics. The biochemical criteria established for adults for distinguishing pleural fluid exudates and transudates need to be studied in children.