OBJECTIVE
To quantify the effect of antibiotic therapy on the probability of subsequent bacterial meningitis in children with fever without source treated as outpatients.
DESIGN
Bayesian meta-analyses. REPORTS INCLUDED: All reports of the organism-specific prevalence of occult bacteremia in children with fever without source treated as outpatients, and the organism-specific prevalence of subsequent meningitis in children with occult bacteremia initially treated as outpatients stratified by type of antibiotic therapy.
RESULTS
The mean probabilities of subsequent meningitis in children with occult bacteremia were 9.8%, 8.2%, and 0.3% in the no antibiotic, oral antibiotic, and parenteral antibiotic therapy groups, respectively. All cases of bacterial meningitis in children with occult bacteremia treated with oral antibiotics were due to Haemophilus influenzae. There were no cases of culture-positive bacterial meningitis in 139 bacteremic children treated with ceftriaxone (mean probability, 0.3%; 95% confidence interval, 0.0% to 1.5%). The mean probabilities of bacterial meningitis in a child with fever without source treated as an outpatient without antibiotics were: Streptococcus pneumoniae, 0.21%; and H influenzae, 0.06%.
CONCLUSIONS
Antibiotic therapy is effective in preventing meningitis in children at risk of occult bacteremia.