OBJECTIVE: To determine whether antibiotics prevent serious bacterial infections in children at risk for occult bacteremia. METHODS: Meta-analysis of randomized controlled trials involving children aged 3 months to 36 months without a focus of infection and randomized into 2 treatment groups: 1) no antibiotic vs antibiotic or 2) IM ceftriaxone vs oral antibiotic. RESULTS: The use of either an oral antibiotic or IM ceftriaxone did trend toward a reduced risk of serious infection, although neither reached statistical significance (OR = 0.60; 95% CI 0.10, 3.49; and OR = 0.38; 95% CI 0.12, 1.17, respectively). It would be necessary to treat 414 patients to prevent 1 serious bacterial infection. When only children with proven occult bacteremia were analyzed, the use of IM ceftriaxone was statistically significant in preventing serious bacterial infections (OR = 0.25; 95% CI 0.07, 0.89). CONCLUSIONS: Clinical judgment should not be replaced by widespread antibiotic use in the approach to a child with fever. If rapid methods to identify children with occult bacteremia, such as polymerase chain reaction, could be improved and become widely available, then antibiotics could be used judiciously on initial visits. Antibiotic use in all children at risk for occult bacteremia implies the treatment of many children unlikely to benefit from such therapy.
OBJECTIVE: To determine whether antibiotics prevent serious bacterial infections in children at risk for occult bacteremia. METHODS: Meta-analysis of randomized controlled trials involving children aged 3 months to 36 months without a focus of infection and randomized into 2 treatment groups: 1) no antibiotic vs antibiotic or 2) IM ceftriaxone vs oral antibiotic. RESULTS: The use of either an oral antibiotic or IM ceftriaxone did trend toward a reduced risk of serious infection, although neither reached statistical significance (OR = 0.60; 95% CI 0.10, 3.49; and OR = 0.38; 95% CI 0.12, 1.17, respectively). It would be necessary to treat 414 patients to prevent 1 serious bacterial infection. When only children with proven occult bacteremia were analyzed, the use of IM ceftriaxone was statistically significant in preventing serious bacterial infections (OR = 0.25; 95% CI 0.07, 0.89). CONCLUSIONS: Clinical judgment should not be replaced by widespread antibiotic use in the approach to a child with fever. If rapid methods to identify children with occult bacteremia, such as polymerase chain reaction, could be improved and become widely available, then antibiotics could be used judiciously on initial visits. Antibiotic use in all children at risk for occult bacteremia implies the treatment of many children unlikely to benefit from such therapy.