Objective: To identify incidence, bacterial aetiology, outcome after treatment and risk factors for poor outcome of neonatal meningitis. Design: Retrospective survey of neonatal meningitis occurring in Australia between January 1987 and December 1989. Data were obtained from Medical Records and Microbiology Departments of hospitals with neonatal nurseries. Setting: Neonatal nurseries throughout Australia. Patients: 116 infants under 6 weeks of age with bacterial or fungal meningitis. Results: The minimum incidence was 0.17 per 1000 live births. Traditional neonatal pathogens were responsible for 60% cases (group B streptococci, 35%; Escherichia coli, 22%), childhood meningeal pathogens for 10% and opportunistic pathogens for 30%. Risk factors for meningitis, including prematurity, were more common among those with meningitis due to E. coli or opportunistic pathogens than among those with infections due to group B streptococci, Listeria monocytogenes or the childhood pathogens (46/60 v. 11/55; P < 0.0001). Meningitis was more likely to be due to Gram-negative bacteria in premature infants (< 36 weeks gestation) than in full-term infants (19/30 v. 20/86; P = 0.0002). The mortality overall was 26% but was higher in extremely premature infants (< 29 weeks) (6/9 v. 24/107; P = 0.009) and among 13 patients who were judged to have had inappropriate initial therapy (7/13 v. 21/97; P = 0.04). Long-term sequelae occurred in at least 23% of survivors, but were more common in those with Gram-negative meningitis (6/10 v. 13/76; P = 0.012). Conclusions: Initial therapy with penicillin or amoxycillin plus cefotaxime is appropriate for most infants with bacterial meningitis. Since some less common Gram-negative bacteria isolated in this survey were resistant to cefotaxime, an aminoglycoside should be added, initially, in Gram-negative meningitis.
