Reduction of unnecessary antibiotic therapy in newborn infants using interleukin-8 and C-reactive protein as markers of bacterial infections.

OBJECTIVE: To examine whether the determination of interleukin 8 (IL-8) and C-reactive protein (CRP) in neonates with suspected nosocomial bacterial infection (NBI) is feasible and cost-effective in reducing antibiotic therapy.

METHODS: Between April 1996 and May 1997, IL-8 was measured 260 times along with blood cultures, CRP, and immature-to-total-neutrophil (IT) ratio for suspected NBI in term and preterm neonates. All infants were retrospectively analyzed for NBI. Sensitivity, specificity, positive and negative predictive values, and 95% confidence intervals were calculated for IL-8, CRP, and IT ratio. Receiver-operating characteristic curves were analyzed to determine optimal thresholds. Between June 1997 and June 1998, IL-8 was measured 215 times in newborn infants with suspected NBI and the decision to start antibiotic therapy was based on increased IL-8 and/or CRP values. A cost-effectiveness analysis was performed and sensitivity, specificity, and receiver-operating characteristic curves were reevaluated.

RESULTS: At the first suspicion of NBI, the combination of IL-8 >/= 53 pg/mL and/or CRP >10 mg/L detected culture-proven NBI with 96% sensitivity. The combined culture-proven and clinical NBI were detected with 93% sensitivity and 80% specificity. The use of IL-8 reduced unnecessary antibiotic therapy for suspected NBI by 73% and was cost-effective when compared with initiating antibiotic therapy based on clinical signs alone or based on clinical signs and an increased IT ratio and/or CRP.

CONCLUSIONS: The combination of IL-8 and/or CRP is a reliable and early test for the diagnosis of NBI in newborn infants. Using the combination of IL-8 and/or CRP to restrict antibiotic therapy to truly infected infants reduces unnecessary antibiotic therapy and is cost-effective.