Late-onset neonatal sepsis (LOS), occurring in newborn of at least 7 days of life, is frequently observed in Neonatal Intensive Care Units (NICUs) and potentially severe (mortality, neurologic and respiratory impairments).

Despite its high prevalence, a reliable diagnostic remains difficult. Currently, nonspecific clinical signs that might be linked to other neonatal conditions, such as prematurity and birth defects are used to determine the diagnosis of LOS. Laboratory results of biological markers, such as C-Reactive Protein (CRP) and Procalcitonin (PCT) are often delayed in comparison with LOS onset. Blood culture results are too late and lack sensitivity.

Excessive antibiotic use is observed in a large proportion of NICU hospitalized newborns. This results in an increased antibiotic resistance, microbiota modification, neonatal complications (pulmonary, ophthalmologic and neurologic) and mortality.

The primary objective is to identify, on a cohort of 250 patients, the optimal biomarker combination with good diagnostic performance (i.e. with maximal Area Under the ROC Curve) to early exclude a LOS diagnostic in newborns of at least 7 days of life with suggestive signs.

This identification will be carried out, as a secondary objective, with a sub-group of pre-term neonates whose birth weight is less than 1500 grams. The diagnostic value of the clinical signs that are suggestive of LOS will also be determined (sensitivity, specificity, negative and positive predictive values).

Once identified, the biomarker combination is expected to reduce unjustified antibiotic use.

Inclusion Criteria:

• patients hospitalized in NICU;

• patients with suggestive signs of LOS including at least one of the following:

  o Fever > 38°C; tachycardia > 160bpm160 bpm; capillary refill time > 3 seconds; grey and/or pale skin complexion; apnea/ bradycardia syndrome,; bloating; rectal bleeding; hypotonia; lethargy; seizures without other obvious cause; increased ventilatory support and/or increased FiO2; cutaneous rash; inflammation at the needle-puncture site of the central venous catheter;

• patients with a standard of care blood sampling, including at least a blood culture;

• consent form signed by at least one parent/ legal representative.

Exclusion Criteria:

• patients treated with antibiotics for a bacteriologically confirmed infection at the moment of/ or 48 hours before blood sampling
• patients who underwent surgery during the 7 days prior to inclusion
• patients vaccinated during the 7 days prior to inclusion