| pubmed-article:21311939 | pubmed:abstractText | This article describes a study of procalcitonin (PCT) measured in cord blood as a discriminating marker of early-onset neonatal infection. This was a monocenter retrospective study with prospective collection of data including all babies born during the study period. Those presenting infection risk factors had PCT measurement. Three groups were defined: certainly infected, probably infected, and non-infected. A total of 12,485 newborns were included, 2151 had PCT measurement, and 26 were infected. Receiver operating curves of PCT determined 0.6 ng/ml as the best cut-off, with an area under the curve of 0.96 (CI 95% 0.95-0.98). Sensitivity, specificity, positive and negative predictive value and positive and negative likelihood ratios were 0.92 (range, 0.75-0.98), 0.97 (0.96-0.98), 0.28 (0.20-0.36), 0.99 (0.99-0.99), 32 (24-41) and 0.08 (0.02-0.3), respectively. Post-test probabilities were 28% (23-33) if the test was positive, and less than 0.001% (0-1.10(-5)) if the test was negative. Gestational age between 28 and 32 weeks (OR 4.4; range, 1.2-16.2) and pH at birth?<?7.10 (OR 2.9; 1.1-7.4) were other independent factors of increasing PCT (p?<?0.05). PCT measured in umbilical cord blood is reliable to detect early infected and non-infected newborns. | lld:pubmed |
