Cerebrospinal fluid procalcitonin as a biomarker of bacterial meningitis in neonates.

Abstract

The objective of the study was to study the performance of cerebrospinal fluid (CSF) procalcitonin as a marker for bacterial meningitis in neonates, and to determine its optimal 'cutoff' in CSF that can be called significant for the diagnosis. Neonates qualifying for lumbar puncture were prospectively studied. Procalcitonin and established CSF parameters were recorded. At a cut-off value of 0.33 ng ml-1, CSF procalcitonin had a sensitivity of 0.92, specificity of 0.87, with positive and negative likelihood ratios of 7.13 and 0.092, respectively. The area under the curve for different CSF parameters was: 0.926 (0.887 to 0.964) (P<0.001) for procalcitonin, 0.965 (0.956 to 0.974) for total leukocyte count, 0.961 (0.94 to 0.983) for neutrophil count, 0.874 (0.825 to 0.923) for protein, 0.946 (0.914 to 0.978) for sugar and 0.92 (0.955 to 0.992) for CSF:serum sugar ratio. The lumbar puncture was traumatic in 36 (21.4%) patients; out of these 15 (41.7%) had bacterial meningitis and 21 (58.3%) had no meningitis. In traumatic lumbar tap group, the median (IQR) CSF procalcitonin in patients with and without meningitis was 1.41 (0.32-3.42) ng/ml and 0.21(0.20-0.31) ng/ml respectively (p<0.05). Procalcitonin measurement has diagnostic efficiency similar to the established CSF markers. Routine assessment of procalcitonin in clean non-contaminated CSF may not yield additional information, but it may have clinical utility in situations where diagnosis of meningitis is in dilemma, as in the case of blood contamination of CSF in traumatic lumbar punctures.