Abstract 1619
Poster Session II, Sunday, 5/2 (poster 116)
Background: Recently, Interleukin-6 (IL-6) and Interleukin-8 (IL-8) have been investigated as laboratory parameters for identifying and monitoring infection in NICU patients. However, when the automated enzyme immunoassay IMMULITE ® is used, there is a difference in the required time and plasma volume for IL-6 and IL-8. Objective: To evaluate the correlation of 'positive' and 'negative' IL-6 and IL-8 values, in order to compare both parameters for their value in infection monitoring in NICU patients. Methods: Plasma levels of IL-6 and IL-8 were determined in parallel in 100 consecutive preterm and full-term infants with the automated enzyme immunoassay IMMULITE® to rule out or follow the course of infection. 'Positive' and 'negative' results for the total of 213 samples were determined by cut-off levels for IL-6 and IL-8 of >100 pg/ml and >70 pg/ml, respectively. Patients with discrepant IL-6 / IL-8 results were assessed retrospectively for clinically relevant infection. Results: The correlation coefficient of all 213 IL-6 / IL-8 pairs was 0.50. However, 35 samples (16.4%), obtained from 23 patients, demonstrated discrepant results. Of these, 8 patients with 13 samples were infected. There was no significant difference in sensitivity (p=1.0) and specificity (p=0.134) between both methods. Considering a sensitivity of 83% for IL-6 and 80% for IL-8 and a specificity of 90% and 76%, respectively, the negative predictive value (NPV) for all samples would be 82.4% for IL-6 and 82.1% for IL-8 (p>0.5). The positive predictive value for IL-6 and IL-8 would be 79.3% and 80.0% (p>0.50), respectively. Conclusions: IL-6 and IL-8 were almost identical in their value for infection monitoring in NICU patients. However, since in this specific assay, IL-8 determination needs less time and sample volume, this cytokine might be more favorable for preterm infants.
