RESUMO
A case-control study of 34 children with occult bacteremia was conducted to test the hypothesis that nonresponse to acetaminophen (decrease less than or equal to 0.8 degrees C) is a risk factor for occult bacteremia. Febrile children visiting the emergency center from May 1986 to October 1987 were monitored for occult bacteremia. Inclusion criteria were age 2 months to 6 years, temperature greater than or equal to 38.9 degrees C, and having a blood culture. Exclusion criteria were serious acute or chronic illness, sponging for fever reduction, current therapy with antibiotics or steroids, and admission to the hospital. Records of 3892 febrile children were reviewed. Of these, 2101 (54%) had a blood culture and 1028 (26%) were eligible. All patients (positive blood culture) were matched with two control subjects (negative blood culture). Patients and control subjects had similar age, gender, ethnicity, height of initial temperature, time to second temperature, and dose of acetaminophen. The estimated risk of occult bacteremia for nonresponders was 9.2 (95% confidence interval 2.7, 32.0). We conclude that children who do not respond to acetaminophen by at least a 0.8 degrees C decrease in temperature have an increased risk of occult bacteremia. However, achieving a response to acetaminophen does not eliminate the possibility that the child has occult bacteremia.