Study objective: To evaluate measurement of interleukin-6 as a diagnostic test for the presence and severity of invasive bacterial disease.
Design: Prospective measurement of interleukin-6 in children with signs of sepsis. (Controls, retrospective from serum bank.)
Setting: Emergency department of an urban children's hospital.
Participants: Twenty children with clinical signs of sepsis and 50 other febrile infants and toddlers with negative blood cultures.
Results: Eleven of the 20 patients had bacteriologically documented infections: four with meningitis and two with bacteremia caused by Neisseria meningitidis, three with meningitis caused by Haemophilus influenzae type b, and one each with meningitis and bacteremia caused by Streptococcus pneumoniae. Ten of these 11 had detectable interleukin-6. The geometric mean interleukin-6 level in these culture-positive patients was 407 pg/mL (95% confidence interval, 108 to 1,545); all three children with levels of more than 300 pg/mL developed septic shock, and one died. One of nine culture-negative patients with clinical signs of sepsis had detectable serum interleukin-6 (166 pg/mL), but none of 50 other febrile children without occult bacteremia did. The detection of interleukin-6 had a sensitivity of 91% and a specificity of 98% for invasive bacterial disease.
Conclusion: High levels of interleukin-6 occur in children with septic shock, and the presence of interleukin-6 in serum is predictive for the isolation of bacteria from blood and/or spinal fluid.