Cerebrospinal fluid analysis in systemically ill children without central nervous system disease

Experience led us to question the applicability of standards for normal cerebrospinal fluid (CSF), originally developed in healthy children, to children with systemic illness but without central nervous system (CNS) infection. The purpose of this study was to test our hypothesis that systemically ill children, in the absence of CNS infection, have an elevated CSF white blood cell count and a greater percentage of neutrophils than accepted norms.

We enrolled 345 patients in the following diagnostic categories: infants 1 month of age or younger with possible sepsis (n = 95), patients older than 1 month of age with possible sepsis (n = 155), patients with a focus of infection in close proximity to the CNS (n = 51), and patients presenting with seizures and fevers (n = 45). Sociodemographic data and results of CSF examination were abstracted from the medical records. Statistical analysis systems were used for data processing.

The CSF white blood cell count did not significantly differ from standards except for a lower mean count in the group presenting with seizures. The percent of CSF neutrophils was significantly greater than standards, however, for those patients older than 1 month of age with possible sepsis, those with a focus of infection in close proximity to the CNS, and those presenting with seizures. Data analysis by quantiles shows only 25% to 50% of patients, in each of the diagnostic categories, meeting the current definition of normal CSF neutrophil count.

Our results show that a mean of at least 5% neutrophils may be present in the CSF with a diagnosis of fever without a source, a focus of infection in close proximity to the CNS, or a seizure with fever in the absence of CNS infection. These data support tailoring treatment based on clinical assessment rather than what is considered an abnormal CSF neutrophil count by current standards.