Materials and methods
Patient selection
The patients in this study were selected from the administrative
database of the Mackay Memorial Hospital, a
medical center with around 200 pediatric beds in northern
Taiwan. The bed number had only minor changes during the
past three decades. This study was approved by the institutional
review board of the hospital. The hospitalized
patients aged 18 years or younger with a positive cerebrospinal
fluid (CSF) bacterial culture results during the period
from January 1984 to December 2012 were enrolled.
Tuberculous meningitis was excluded. Patients diagnosed as
having meningitis without bacteria isolated from the CSF
were not included, even those who had elevated CSF cell
counts, positive blood culture results, or clinical presentations
of meningitis were excluded. The CSF cultures
yielded normal skin flora (e.g., coagulase-negative staphylococci)
considered to be contaminations were excluded
unless CSF cell counts and clinical presentations fit the
diagnosis of meningitis. Patients with evidence of congenital
infections were also excluded.
Data collection
We searched the patients from both the diagnostic database
and microbiological database to find eligible patients.
The analyzed data included patients’ age and sex, meningitis
occurrence year, CSF pathogens, complications, and
outcomes. If a patient had more than one spinal tap within
an episode of meningitis, only the first CSF result was
analyzed. If a patient had more than one episode of bacterial
meningitis, i.e., recurrent meningitis, the findings
were calculated separately in each episode.