Abstract

Two decision analysis reports published in 1991 concluded that the strategy of routine blood culture and empiric antibiotics was the superior strategy for febrile children at risk for occult bacteremia. This report describes a decision analysis of treatment strategies for these children considering the following changes that have occurred since then: (1) Hemophilus influenzae B incidence is low because of widespread vaccine use; (2) the emergence of resistant Streptococcus pneumoniae may affect the clinical effectiveness of empiric antibiotics in the future; and (3) the negative consequences of unnecessary antibiotic treatment have yet to be well defined. A decision analysis approach, modifying the original assumptions, was carried out. Sensitivity analyses were conducted on all assumption variables. Strategies employing empiric antibiotics were found to have the best outcomes, assuming low negative treatment consequences. If a high level of negative treatment consequences is assumed, strategies using a white blood cell count (WBC) are superior. If a very high level of negative treatment consequences is assumed, the strategy of no tests and no empiric antibiotic treatment is usually superior, unless the frequency of bacteremia is 10% or higher and empiric antibiotic efficacy is high, in which case a WBC strategy is superior. This information can be used to select a treatment strategy based largely on the estimation of the negative consequences of treatment.