Rapid antigen assay for the diagnosis of pneumococcal bacteremia in children: a preliminary study.

Abstract

We determine the sensitivity of a new Streptococcus pneumoniae antigen assay in children with S pneumoniae bacteremia and also determine the specificity of the assay in afebrile children without apparent pneumococcal infection. We enrolled 2 groups of children, 3 months to 5 years of age, presenting to an urban pediatric emergency department or hospital-based clinic between January 1, 2000, and January 1, 2001. The study group included patients with S pneumoniae bacteremia. The control group included patients without fever or identified focal bacterial infection (other than a urinary tract infection). We excluded from the control group patients with a recent diagnosis (prior 4 weeks) of bacteremia, pneumonia, meningitis, acute otitis media, or other suspected S pneumoniae infection. Ninety-six patients were enrolled: 24 children with S pneumoniae bacteremia and 72 afebrile control patients. The assay produced a positive result in 23 of 24 patients with pneumococcal bacteremia, yielding a sensitivity of 95.8% (95% confidence interval [CI] 78.9% to 99.9%). The test result was negative in 67 of 72 afebrile patients without apparent pneumococcal infection, yielding a specificity in this population of 93.0% (95% CI 83.4% to 97.5%). This S pneumoniae assay demonstrated high sensitivity for the defined bacteremic population, as well as high specificity among afebrile children without apparent pneumococcal infection. Further study in febrile children at risk of pneumococcal bacteremia is needed before the value of the test in children at risk for invasive pneumococcal disease can be determined.