Comparison of procalcitonin with interleukin 8, C-reactive protein and differential white blood cell count for the early diagnosis of bacterial infections in newborn infants

Objective. To evaluate procalcitonin (PCT) as a test for early diagnosis of bacterial infections (BI) in newborn infants and to compare the results of PCT with those of interleukin 8 (IL-8), C-reactive protein (CRP) and diffe rential white blood cell count. Study design. PCT was prospectively measured along with IL-8, CRP and diffe rential white blood cell counts and blood cultures in 197 newborn infants a t the first suspicion of bacterial infection. PCT, IL-8, CRP and differenti al white blood cell counts were analyzed for sensitivity, specificity and p ositive and negative predictive values after receiver operating characteris tic curve analysis for best thresholds. The kinetics of PCT was determined in infants with and without BI. Results. Forty-six infants were diagnosed clinically as having BI, of whom 9 had BI with positive blood cultures. At a cutoff value of 0.50 mu g/l, PC T detected combined culture-proved and clinical BI with a sensitivity of 57 % (95% confidence interval, 41%, 71%) and a specificity of 66% (95% confide nce interval, 57%, 74%). The combination of IL-8 greater than or equal to 7 0 ng/l and/or CRP >10 mg/l achieved a sensitivity of 91% (95% confidence in terval, 79%, 98%) and a specificity of 73% (95% confidence interval, 64%, 8 1%). PCT values of infected and not infected infants tended to rise for 24 h after initial evaluation and then decreased. Conclusion. The combination of IL-8 and CRP is more reliable than PCT as a test for early diagnosis of BI in newborn infants.