Objective: To investigate the reliability of serum procalcitonin (PCT) as an early diagnostic test (within the first 12 hours of life) of neonatal sepsis in newborns with maternal or neonatal risk factors for infection.
Material and methods: We performed a prospective study of 123 newborns consecutively admitted to neonatal unit over a 2-year period with at least one risk factor for infection. We constructed a 2x2 table between the validated test (serum PCT by semi-quantitative assay, with several cut-off points: 0.5, 2 and 10 ng/ml) and the reference assay (blood culture or clinical, laboratory and microbiological confirmation of sepsis). The validity (sensitivity, specificity), safety [positive predictive value (PPV) and negative predictive value (NPV)] and likelihood ratios (LR+ and LR-) of the test were calculated.
Results: Serum PCT was measured within the first 12 hours of life in 95% of the patients (mean and median=6 hours). The best cut-off point for serum PCT was 2 ng/ml, and, taking subsequent clinical-laboratory-microbiological confirmation of sepsis as the best reference assay, showed a sensitivity of 100% (95% CI 65-100), specificity of 82% (95% CI 74-88), PPV of 25% (95% CI 13-44), NPV of 100% (95% CI 96-100), LR+ of 5.5 (95% CI 3.7-8.1), and LR- of 0.
Conclusions: Serum PCT levels<2 ng/ml within the first 6-12 hours of life in newborns with risk factors for infection are useful as a screening assay to rule out neonatal sepsis with a sensitivity of 100% (false negatives=0% and NPV=100%). However, for subsequent confirmation a more specific assay (with a low false positive rate and high PPV) should be used, such as C-reactive protein. The higher cost of the serum PCT test should be weighed against shorter admissions as a result of its use.