Whole blood interleukin 8 and plasma interleukin 8 levels in newborn infants with suspected bacterial infection

Aim: To evaluate a new micro-method technique for measurement of interleukin 8 in detergent-lysed whole blood (whole blood IL-8) applicable to capillary blood sampling as a test for bacterial infections in neonates. Methods: Whole blood IL-8 was measured at the first suspicion of infection along with IL-8 determined in plasma (plasma IL-8), C-reactive protein and blood cultures in 154 consecutive newborn infants with clinical signs of bacterial infection. Only 20 mul of whole blood were required for the new assay. Results: Blood culture-proven infections were diagnosed in six infants and clinical infection (defined as a combination of clinical and laboratory signs) in 20 infants. 1000 pg/ml was determined as the suitable threshold for whole blood IL-8 by ROC-curve analysis. At that threshold, whole blood IL-8 detected blood culture-proven infections with a sensitivity of 83% and a specificity of 67%. The areas under the ROC curves were similar for whole blood IL-8 and plasma IL-8. Conclusions: Compared with plasma IL-8, whole blood IL-8 offers the advantages that measurements of whole blood IL-8 require a smaller blood sample volume and are not altered by haemolysis. The diagnostic accuracy of whole blood IL-8 remains to be studied in more detail in the future.