An Epidemiological Investigation of a Sustained High Rate of Pediatric Parapneumonic Empyema: Risk Factors and Microbiological Associations

Carrie L. Byington () 0 1 5 LaShonda Y. Spencer 0 1 5 Timothy A. Johnson 0 1 4 5 Andrew T. Pavia 0 1 5 Daniel Allen 0 1 3 5 Edward O. Mason 0 1 2 5 Sheldon Kaplan 0 1 2 5 Karen C. Carroll 0 1 5 Judy A. Daly 0 1 3 5 John C. Christenson 0 1 5 Matthew H. Samore 0 1 5 0 Received 9 July 2001; revised 5 September 2001; electronically published 3 January 2002. Presented in part: Joint meeting of the Pediatric Academic Societies and the American Academy of Pediatrics , 13 May 2000, Boston , Massachusetts . Financial support: Robert Wood Johnson Generalist Physician Faculty Scholar award (to C.L.B.) and American Academy of Pediatrics resident research award (to L.Y.S.). of Utah , 50 North Medical Dr. , Salt Lake City, UT 84132 1 Parapneumonic effusion and empyema are recognized complications of bacterial pneumonia. Empyema de- velops in 2 Department of Pediatrics and Microbiology, Baylor College of Medicine , Houston, Texas 3 Clinical Microbiology Laboratories, Primary Children's Medical Center , Salt Lake City, Utah 4 School of Medicine, University of Utah 5 in Salt Lake City, Utah , in the mid-1990s were noted to have empyema, an observation that prompted an epidemiological investigation. The objectives of the investigation were (1) to determine the incidence of parapneumonic empyema among pediatric patients hospitalized with community-acquired bacterial pneu- monia (CAP) in Utah; (2) to describe and compare the demographic characteristics and the clinical and mi- crobiological findings for hospitalized patients with CAP who did and who did not have empyema ; and (3) to determine if there are specific modifiable risk factors for the development of empyema in children We investigated the increasing incidence of pediatric empyema during the 1990s at Primary Children's Medical Center in Salt Lake City. Of 540 children hospitalized with community-acquired bacterial pneumonia (CAP) who were discharged from 1 July 1993 through 1 July 1999, 153 (28.3%) had empyema. The annual population incidence of empyema increased during the study period from 1 to 5 cases per 100,000 population aged !19 years. Streptococcus pneumoniae was identified as the most common cause of CAP with or without empyema; serotype 1 accounted for 50% of the cases of pneumococcal empyema. Patients with empyema were more likely to be 13 years old, to have 7 days of fever, to have varicella, and to have received antibiotics and ibuprofen before admission to the hospital, compared with patients without empyema (P ! .0001 for each factor). The increasing incidence of empyema was associated with infection due to S. pneumoniae serotype 1, outpatient treatment with certain antibiotics, ibuprofen use, and varicella. - PATIENTS AND METHODS Patient ascertainment. The electronic medical record system at PCMC was searched for the period from 1 July 1993 to 1 July 1999 to identify all patients dis charged with ICD-9 (International Classification of Diseases, 9th revision) codes of 481482.9 (bacterial pneumonia) and/ or 510.0510.9 (empyema, pleural effusion). The medical chart of each of these subjects was reviewed to confirm the diagnosis of CAP with or without empyema. CAP was considered to be present if the chest radiograph demonstrated a focal infiltrate and if none of the exclusion criteria were present. Patients were excluded if the pneumonia was likely to have been viral in etiology (e.g., if there was a finding of a nonfocal infiltrate in a patient with respiratory syncytial virus infection), caused by aspiration, hospital acquired, associated with cystic fibrosis, or manifest at birth (in neonates). Empyema was considered to be present if there was a finding of a pleural effusion on the chest radiograph coupled with any of the following additional findings: pleural fluid with a pH of 7.2, a lactate dehydrogenase level of 1000 IU/mL, a glucose level of 20 mg/dL, a protein level of 3000 mg/dL, and/or a WBC count of 50,000 cells/mL [3, 4]; culture or Gram staining of pleural fluid that was positive for bacteria; or need for surgical decortication. Surgical and pathology reports were reviewed to confirm the diagnosis of empyema. Patients with a pleural effusion were excluded from the empyema group if there were no pleural fluid analysis results or if there were findings consistent with a transudate; however, these patients could be classified as having bacterial pneumonia if they fulfilled the inclusion criteria. Patients who had CAP and empyema were compared with patients who had CAP alone. The study was reviewed and approved by the institutional review boards of both the University of Utah and Primary Childrens Medical Center in Salt Lake City. Demographic and clinical evaluation. Demographic and clinical information obtained included the following: age; sex; race or ethnicity; date of admission to the hospital; length of hospital stay; signs and symptoms of infection at presentation; duration of symptoms; clinical outcome; and potential risk factors for the development of empyema, including antibiotic exposure before admission, ibuprofen or acetaminophen use before admission, immunization status, history of recent viral illness, immunodeficiency, chronic illness, tobacco use or exposure, and attendance at a daycare facility. Microbiological evaluation. The computerized microbiological records at PCMC were reviewed. Bacterial isolates from blood, pleural fluid, and surgical specimens were analyzed. Antimicrobial susceptibility profiles were collected for each bacterial isolate. For Streptococcus pneumoniae, penicillin susceptibility was defined according to the MIC of penicillin, as determined by use of the Etest (AB Biodisk) and with use of MIC break points established by the National Committee for Clinical Laboratory Standards [5]. The microbiology laboratory at PCMC routinely freezes specimens of pathogens isolated from blood and other normally sterile sites at 70 C. Viable isolates of S. pneumoniae recovered from patients with and without empyema were serotyped by means of the capsular swelling method with use of commercially available antiserum samples (Statens Serum Institut or Dako). The investigators who performed the analysis (E.O.M. and S.K.) were blinded to the sources of the pathogen samples. Pulsed-field gel electrophoresis (PFGE) was done for all S. pneumoniae serotype 1 isolates by means of clamped homogeneous electric fields electrophoresis on a CHEF Mapper (BioRad), as described elsewhere [6]. Genomic DNA in 2% InCert agarose plugs (FMC BioProducts) was digested with SmaI. DNA was subjected to electrophoresis in a 1.2% agarose gel with a voltage gradient of 6 V/cm and a switch time ramped linearly from 1 to 20 s over 21 h. Band patterns were interpreted according to the criteria established by Tenover et al. [7]. Estimation of population incidence and statewide attack rate of empyema. To estimate the population incidence of empyema, only data for case patients who resided in Utah were anal (...truncated)