Clinical Question: How useful is the white blood cell (WBC) count for ruling out bacterial meningitis in sick, febrile infants?

Setting: Emergency department

Study Design: Cohort (retrospective)

Synopsis: This study evaluated the role of the peripheral WBC count in ruling out bacterial meningitis in 5,353 infants between the ages of three days and 89 days. At the institution where the research was conducted, all infants presenting with a temperature of 38°C (100°F) or higher underwent a sepsis evaluation, including a peripheral and cerebrospinal WBC count. The investigators evaluated all children for whom cerebrospinal fluid samples (i.e., lumbar puncture) were obtained. Even though the peripheral WBC count was not used to determine which children would have a lumbar puncture, there probably were children for whom a peripheral WBC count was performed but a lumbar puncture was not (the authors do not indicate).

The overall prevalence of bacterial meningitis was 0.41 percent (4.1 cases per 1,000; 95 percent confidence interval [CI]: 2.6 to 6.2). The rate of meningitis was 3.2 percent (95 percent CI: 1.3 to 6.4) for a WBC count of less than 5,000 per mm³ and 0.29 percent (95 percent CI: 0.16 to 0.48) for a WBC count of more than 5,000 per mm³. Overall, 64 percent of infants with bacterial meningitis would be falsely excluded if a WBC count between 5,000 and 20,000 cells per mm³ was accepted as low risk. Using a cut-off of more than 15,000 cells per mm³ as the criterion of high risk would fail to identify 73 percent of infants with meningitis.

Bottom Line: The peripheral WBC count cannot be used to safely identify febrile infants who do not have bacterial meningitis. (Level of Evidence: 1c)